Digital PDFs

DEC-IN-GRZA-D

December 1971

211 pages

Original

7.1MB

Document:	INDAC8 2 Handbook 197112
Order Number:	DEC-IN-GRZA-D
Revision:
Pages:	211
Original Filename:	http://bitsavers.org/pdf/dec/pdp8/industrialProducts/DEC-IN-GRZA-D_INDAC8_2_Handbook_197112.pdf

OCR Text

r=or industrial detection. analysis and control

handbook

For additional copies, order No. DEC-IN-GRZA-0 from the Program Library, Digital Equipment
Corporation, Maynard, Mass. 01754

digital equipment corporation

Price: $6.00

1st Edition December 1971

The material in this manual is for informational purposes and is subject to change
without notice.

The following are trademarks of Digital Equipment
Corporation, Maynard, Massachusetts:
DEC

PDP

FLIP CHIP

FOCAL

DIGITAL

COMPUTER LAB

CONTENTS
Page

CHAPTER 1

AN OVERVIEW

1-1
1-1
1-1

The Industrial Environment
Data Collection
Process Control
INDAC Language Capabilities

1-2

Facilities of the INDAC Software System

1-2

CHAPTER 2

BUILDING THE SAMPLE SYSTEM

2.1

Introduction

2.2

Loading the HINDAC (Tape 1) Program

2-1
2-1

2.3

Loading the Monitor System Dump and I NDAC File Tapes

2-2

2.4

Loading and Building the INDAC Compiler

2-4

2.5

Running the SG EN 8/2 Program

2-5

2.6

Loading the I NDAC Executive 8/2

2-5

2.7

Loading GENDAC to Configure the Sample System

2-6

2.8

Loading Sample Program 2 Using the Editor

2-7

2.9

Compiling Sample Program 2

2-8

2.10

Executing Sample Program 2

2-9

CHAPTER 3

PROGRAMMING THE INDAC 8/2 SYSTEM

3.1

Introduction

3-1

3.2

Programming an Industrial Algorithm

3-2

3.2.1

LET Statement (Assignment)

3-3

3.2.1.1

Arithmetic

3-3

3.2.1.2

Boolean

3-6

3.2.1.3

Statement Evaluation Rules

3-8

3.2.2
3.2.3

GOTO Statement (Branching)
IF Statement (Conditional Testing)

3-9
3-10

3.2.4

FOR/NEXT Statements (Looping)

3-12

3.2.5

DO/RETURN Statements (subroutines)

3-15

3.3

Message and Logging Capabilities

3-16

3.3.1

.HEADER Statement

3-22

3.3.2

.FORMAT (Picture) Statement

3-22

3.3.3

.FORMAT (Declarative) Statement

3-23

3.3.4

.STORAGE Statement

3-24

3.3.4.1

Development of Array Names

3-24

3.3.4.2

Specification of Array Size

3-25

3.3.4.3

Specification of Array Elements

3-25

3.3.4.4

Specification of an Array Window

3-25

3.3.4.5

Array Spanning Windows

3-26

3.3.4.6

Presetting Stored Values

3-26

3.4

Data Collection and Control

3-27

3.4.1

DAC (Digital-to-Analog Conversion Devices)

3-34

iii

CONTENTS (Cont)
Page
3.4.1.1

Hardware Device

3-34

3.4.1.2

Equipment Declaration Statement

3-34

3.4.1.3

Language Statement

3-34

3.4.2

ADC (Analog-to-Digital Conversion Devices)

3-34

3.4.2.1

Hardware Device

3-34

3.4.2.2

Equipment Declaration Statement

3-34

3.4.3

AF04 (Integrating Digital Voltmeter)

3-36

3.4.3.1

Hardware Device

3-36

3.4.3.2

Equipment Declaration Statement

3-36

3.4.3.3

Language Statement

3-37

3.4.4

UDC (Universal Digital Controller)

3-37

3.4.4.1

Hardware Device

3-37

3.4.4.2

Equipment Definition Statement

3-37

3.4.4.3

Language Statement

3-38

3.4.5

Fl LE (Pseudo-Device)

3-38

3.4.5.1

Hardware Device

3-38

3.4.5.2

Equipment Definition Statement

3-38

3.4.5.3

Language Statement

3-38

3.4.6

Equipment Statement Summary

3-39

3.5

Structure of an INDAC Job

3-40

3.5.1

Program Segmentation

3-40

3.5.1.1

Job Specification Segment

3-41

3.5.1.2

.PHASE Segment

3-42

3.5.1.2.1

Activity Statements

3-42

3.5.1.2.2

.ACTION Statement

3-42

3.5.1.3

.SNAP Segments

3-43

3.5.1.4

Subroutine

3-44

3.5.1.4.1

Internal Subroutine

3-44

3.5.1.4.2

External Subroutine

3-44

3.5.1.4.3

Implicit Subroutine

3-45

3.5.2

Scheduling Capabilities

3-46

3.5.2.1

Activity Statements

3-47

3.5.2.1.1

EVE RY Activity Statements

3-47

3.5.2.1.2

DE LAY Activity Statements

3-48

3.5.2.1.3

AT Activity Statements

3-49

3.5.2.2

Action Statements

3-49

3.5.2.2.1

Timer Action Statements

3-49

3.5.2.3

EX IT Program Control Statement

3-49

3.5.2.4

Resolving Timer Requests

3-50

3.6

The Run-Time System/Making the Pieces Work

3-50

3.6.1

Time and Priority Scheduling

3-51

3.6.1.1

Division of the Process Stack

3-51

3.6.1.1.1

Executive Command Decoder

3-51

3.6.1.1.2

Field Interrupt Processing

3-51

CONTENTS (Cont)
Page
3.6.1.1.3

Foreground Processing

3-52

3.6.1.1.4

Background Processing

3-53

3.6.2

Dynamic Core Management and Swapping

3-53

3.6.2.1

Job Specification Segment

3-53

3.6.2.2

The PHASE Segment

3-53

3.6.2.3

The SNAP Segment

3-53

3.6.2.4

External (Disk-Resident) Subroutines and Functions

3-53

3.6.2.5

Switching Priority Levels

3-53

3.6.2.6

Executive to SNAP Communication

3-54

3.6.3

Dynamic 1/0 Buffers

3-54

3.6.4

Operator Communication

3-54

3.7

Servicing Field Interrupts

3-55

3.7.1

Equipment Declaration Statement

3-55

3.7.2

Language Statements

3-55

3.7.2.1

The Standard Call

3-55

3.7.2.2

INITIALIZE Request

3-56

3.7.2.3

IDENTIFY Request

3-56
3-57

3.7 .2.4

TRANSFER Request

3.7.3

Sample Program

3-58

3.8

Handling the Consoles

3-59

3.8.1

Equipment Declaration Statement

3-60

3.8.2

Language Statements

3-60

3.8.3

Program Examples

3-61

CHAPTER 4

CONFIGURING A SYSTEM HAVING STANDARD DEC
PROCESS 1/0 DEVICES

4.1

Introduction

4-1

4.2

Loading the HINDAC Program

4-2

4.3

Loading the Monitor System Dump and INDAC File Tapes

4-3

4.4

Loading and Building the INDAC Compiler

4-5
4-6

4.5

Running the SG EN 8/2 Program

4.6

Loading the INDAC Executive 8/2

4-6

4.7

Loading GENDAC to Configure a Specific System

4-7

CHAPTER 5

PREPARING THE PROGRAM

5.1

Introduction

5-1

5.2

Monitor

5-1

5.2.1

Monitor Residence

5-1

5.2.2

Starting the Monitor

5-1

5.2.3

Bootstrapping the Monitor

5-2

5.2.4

Monitor Error Messages

5-2

5.3

System Programs

5-2

5.3.1

Command String Formrit

5-3

CONTENTS (Cont)
Page

5.3.1.1
5.3.1.2
5.3.1.3
5.3.1.4
5.3.2
5.3.3
5.3.3.1
5.3.3.2
5.3.3.3
5.3.3.4
5.3.3.5
5.3.3.6
5.3.4
5.3.4.1
5.3.4.2

Errors During Compilation

5.3.4.3
5.3.5
5.3.6
5.3.6.1
5.3.6.2

Correcting Compilation Errors

Binary Loader Error Messages

5-3
5-3
5-4
5-4
5-4
5-5
5-7
5-7
5-8
5-8
5-11
5-13
5-13
5-14
5-15
5-15
5-16
5-22
5-22
5-24

CHAPTER 6

EXECUTING THE PROGRAM

6-1

CHAPTER 7

MODIFYING THE SYSTEM

7.1
7.2
7.2.1
7.2.2
7.2.3
7.2.4
7.2.5
7.2.6
7.3
7.3.1
7.3.2
7.4
7.4.1
7.4.2

7.4.3
7.4.4
7.5
7.5.1
7.5.2
7.5.3

Device Names
Filenames
Punctuation
Special Characters
Examples of Command Strings
Editor
Modes of Operation
Input Commands
Output Commands
Editing Commands
Special Characters and Functions
Editor Error Messages
INDAC Compiler
Compiler Output

PIP
Loading Programs - Disk System Binary Loader
Binary Loader Operating Procedures

Basic Module

7-1
7-1
7-1
7-1
7-2
7-2
7-2
7-3
7-3
7-3
7-4
7-5
7-5

Extension ivioduies

..,1-u
a

Call-Up Modules

7-6
7-6
7-9
7-9
7-9
7-10

Introduction
System Communication Tables
Intrinsic Functions (IF>
External Subroutines (XS>
System Devices (SD)

Core Map (CM)
Page Zero (PZ>
Special GENDAC Table(**)
Updating the Software
System Mode of Updating
Binary Mode of Updating
Library Structure

Additiona~ Notes

Library Tape Format
Definitions
Library Tape Header
Module Header

CONTENTS (Cont)
Page
7.5.3.1

Functional Group Declaration

7-10

7.5.3.2

Module Declaration

7-11

7.5.3.2.1

Physical Descriptions of Module

7-11

7.5.3.2.2

Logical Description of Module

7-13

7.5.3.3

Module Body

7-14

7.5.3.3.1

Module Title

7-15

7.5.3.3.2

Fixup Declarations

7-15

7.5.3.4

Group and Module Termination

7-16

7.6

Coding Details

7-16

7.6.1

Function (Core-Resident in Field 0)

7-17

7.6.2

Subroutine

7-18

7.6.2.1

Subroutine Requirements and Analysis

7-18

7.6.2.2

Sample of Subroutine Coding

7-18

7.6.3

7-19

7.6.3.1

1/0 Handler
1/0 Handler Operation

7.6.3.1.1

The Control Driver Table

7-20

7.6.3.1.2

The Format Driver Table

7-20

7.6.3.1.3

Data Information and Calls

7-20

7.6.3.1.4

Executive Page Zero Parameters

7-20

7.6.3.1.5

Interrupting Devices

7-21

7.6.3.2

A Typical 1/0 Handler

7-21

7.7

Vector Code

7-22

7.7.1

Numeric Formats

7-22

7.7.2

Vector Code Example

7-22

7.7.3

Load and Store Instructions

7-23

7-19

7.7.3.1

Load a Simple Variable

7-23

7.7.3.2

Store a Simple Variable

7-24

7.7.3.3

Load an Array Element (Indexed Variable)

7-24

7.7.3.4

Store an Array Element

7-24

7.7.3.5

Load an External Argument

7-24

7.7.3.6

Store into an External Argument

7-24

7.7.3.7

Load an External Argument, Indexed

7-24

7.7.3.8

Store into an External Argument, Indexed

7-25

7.7.4

GOTO Statements

7-25

7.7.4.1

Unconditional GOTO

7-25

7.7.4.2

Computed GOTO

7-25

7.7.5

IF Statement

7-25
7-26

7.7.6

DO Statements

7.7.6.1

Internal Subroutine Cal I

7-26

7.7.6.2

External Subroutine Call

7-26

7.7.7

Return Statement

7-26

7.7.8

Loop Statements

7-27

FOR Statement

7-27

l\lr''' r"t-.L-.L---- ___ ..._

7-27

7.7.8.1
,

,.,

I , I .U.L

1'111:.A I

vldlt:lllt:lll

vii

CONTENTS (Cont)
Page
7.7.9

Arithmetic and Logical Operators

7-27

7.7.9.1

Binary Vectors

7-27

7.7.9.2

Unary Vectors

7-28

7.7.10

Special Vectors

7-28

7.7.11

Special PAL-I Code
Running GENDAC

7-29

7.8

7-28

APPENDICES
APPENDIX A SUMMARY OF INDAC STATEMENTS
A.1

Executable Statements

A-1

A.2

Non-Executable Statements

A-2

APPENDIX B

EXAMPLE PROGRAM

B-1

APPENDIX C USING THE DISK MONITOR SYSTEM
C.1

Editor Command Summary

C-1

C.2

Editor Key Function Summary

C-2

C.3

Error Messages

C-2

APPENDIX D INDAC COMPILER ERROR MESSAGES

D-1

APPENDIX E SPUT ERROR CONDITIONS

E-1

APPENDIX F

F-1

INDAC 8/2 EXECUTIVE ERROR MESSAGES

APPENDIX G GENDAC ERROR MESSAGES

G-1

APPENDIX H EXECUTIVE COMMAND DECODER OPERATION
H.1

Operator Commands

H-1

H.2

Command Decoder Error Messages

H-1

APPENDIX I

THE GENDAC LIBRARY

1-1

APPENDIX J

SYSTEM COMMUNICATION TABLES

J-1

APPENDIX K ESUP OPERATION
K.1

Introduction

K-1

K.2

Loading Procedure

K-1

K.3

Producing the Equate Tape

K-1

K.4

Producing the Core Map

K-2

viii

APPENDICES (Cont)
Page

K.5

Returning to Monitor

K-2

K.6

Restrictions

K-2

K.7

References

K-3

K.8

Examples

K-3

K.8.1

Loading and Usage

K-3

K.8.2

Sample Equate Printout

K-3

ILLUSTRATIONS
Figure No.

Title

2-1

Sample System Log Output

3-1

Sample Process Control Application

Art No.

Page

08-0727

3-29

2-i

3-2

Structure of an INDAC Job

3-41

4-1

GENDAC and 1/0 Handler Dialogue

4-11

4-2

GENDAC and Function Dialogue

7-1

System Communication Tables - Interaction

08-0657

7-2

Paper Tape Binary Image

08-0671

7-5

7-3

Internal Format of Numeric Data

08-0670

7-22

4-17

SKELETONS
Skeleton No.

Title

Page

Basic Skeleton

3-1

The Algorithm

3-3

Sample Alarm Test Routine

3-16

Header/Format Details

3-19

Storage Details

3-21

3
3

3-28
Equipment Details

3-32

CHARTS
Chart No.

Title

Page

4-1

INDAC 1/0 Device Handler Library Summary

4-9

4-2

INDAC Function Library Summary

4-10

TABLES
Title

Table No.

Page

3-1

Basic Arithmetic Notation, Algebraic vs. INDAC

3-4

3-2

Relational Operators

3-10

3-3

Mask Generation

3-13

3-4

Declarative .FORMAT Specifications

3-24

3-5

Job Specification Segment, Organization and Contents

3-41

3-6

Organization of PHASE Segment

3-42

3-7

Organization of a SNAP Segment

3-43

5-1

System Error Messages

5-2

7-1

Allocation Codes

7-12

7-2

Allocation Parameters

7-12

7-3

Table Update Information

7-13

7-4

CALL Statements

7-20

7-5

Executive Parameters

7-21

7-6

Binary Vectors

7-27

7-7

Unary Vectors

-, 'lO

7-8

ESUP Commands

7-29

G-1

Numbered GENDAC Error Messages

G-1

rL.o

G-2

Unnumbered GENDAC Error Messages

G-3

J-1

System Communication Tables

J-1

CHAPTER 1
AN OVERVIEW

THE INDUSTRIAL ENVIRONMENT

Accurate gathering and reporting of data is indispensable to the efficient operation of a modern industrial facility.
The data is needed to evaluate the overall operation, to determine its efficiency, and to calculate actual vs. theoretical yields. Data must also be displayed either permanently in printed copy or temporarily through CRT displays and indicator panels, to keep operators informed of the status, trends, and disturbances within the process
or test. As a result of calculations performed on the data, the process can be controlled and modified by operator
interaction or through feedback hardware.
Multichannel analog graphic recorders can monitor a number of variables and compare them to a known time base,
but interpretation has limited accuracy and the reduction and analysis of the data requires considerable time and
effort. Also, with graphic recorders, the possibility of automatic feedback control is lost.
The high speed, accuracy, and flexibility of a computer-based system provides data gathering techniques greatly
beyond the ability of simple displays and recorders.
DATA COLLECTION

The basic function of any computer-based data acquisition system is to measure and record the operating characteristics of various equipment or process sensors and to compare these measured values with predetermined
standards. The system controller, the computer, monitors input signals, controls the data sampling process,
evaluates input data, detects process malfunctions, calculates data necessary to define system performance, formats data for display or storage, and controls process actuators.
A computer-based system will characteristically monitor many process sensors, both analog and digital. Analog
inputs, such as those from thermocouples, potentiometers, strain gauges and flow meters, must be converted to
digital signals acceptable to a data processor. In addition to converted analog inputs, the compt.ter may accept
digital information from sources such as contact closures, limit switches, and manual entry consoles. System
outputs, both direct digital and digital converted to analog, may be used to control displays, recorders, and actuators. Data and information may also be transferred to mass storage units, alphanumeric printers and punched
paper tape.
PROCESS CONTROL

A computer controller gives the data acquisition system total flexibility. The computer program is used to determine what to measure, when to measure, and how to interpret the measurement. The system can thereby translate voltage readings into meaningful engineering units and can subject the inputs from different transducers to
the proper linearization functions. The computer program can calculate indirect measurements, such as flow,
from the measurable quantities of velocity, cross-sectional area, and time. The data acquisition and control function may need to be established as a foreground priority task. If this function does not demand all available

1-1

computer time, a background task could be implemented to perform such functions as calculating overall process
or plant efficiency or maintaining inventory control records.
In an application such as engine testing, a designer may build a computer-based data acquisition system that can
format and pre-process incoming data before placing it on tape for later analysis at a computation center. Formatting and pre-processing save considerable computation time. The computer can also set limits for critical
parameters, and provide an alarm when the limits are exceeded, thereby increasing the safety of facilities and
personnel. With feedback hardware, the computer can control fuel flow rate, air mixture, load environmental
conditions, etc.
Because of their high speed, accuracy, and flexibility, computers are being implemented in the industrial world.
Successful implementation depends largely on the availability of process interface devices and easy-to-use software. I NDAC 8/2, for example, offers an industrial language to simplify bringing the computer into the process.
INDAC LANGUAGE CAPABILITIES
The current high-level languages do not allow such functions as time and priority sequencing, handling of random
external interrupts, analog and digital input and output, and file handling - all of which may be called for in data
acquisition and control systems.
With these limitations in mind, Digital's engineers and programmers implemented statements, similar to those
used in BASIC and FORTRAN for arithmetic and logical computations and program control with additional
English statements to provide for:
• Control of the start of tasks according to a time lapse, or time-of-day.
• Segmentation of programs into several units, with each unit occupying computer storage only when it is
actually in use.
• Definition of process interface equipment, associated channels, and modes of operation.
• Easy-to-use data collection and control statements.
• Transfer of data between the computer and mass storage for record keeping or"analysis.
The result: the INDAC 8/2 language, the first compiler-level language designed for real-time data acquisition and
control in a mini-computer. With a few hours training an engineer familiar with BASIC or FORTRAN can produce simple programs in the INDAC 8/2 language for analysis of process environments.
FACILITIES OF THE INDAC SOFTWARE SYSTEM
The I NDAC Software System provides the user with facilities for:
• Rapidly loading the INDAC 8/2 Software System
• Configuring or modifying the I NDAC 8/2 System to meet the users application needs
• Creating and Editing the application source program
• Compiling the application algorithms into an efficient run-time program.
• Running under control of a real-time Executive
• Utiiity Support Functions
All INDAC 8/2 programs are on paper tapes supplied with the system. The tapes can be loaded quickly using the
detailed procedures given in Chapter 2 or 4.
Once the basic I NDAC software system is loaded it can be configured to meet specific users application needs
using GENDAC (the system configuration program).

1-2

The standard PDP-8 Disk Monitor System, which is unique in the field of small computers, controls all INDAC
8/2 program preparation. With this powerful tool the user can establish and maintain files for I NDAC 8/2 source
programs, edit, compile, and execute them using simple keyboard commands. The Disk Monitor System also
includes utility programs for loading, deleting, and transferring i i\iDAC files.
The compiler converts the source program into object code, which is executed at run time. Extensive diagnostics
help the user debug his programs.
The operating system (or Executive) provides for automatic overlays of program units and provides for program
or task scheduling based on time, priority, events, or program decisions. Tasks made of program overlays have
distinct priority levels. The tasks may be executed in response to process interrupts, events, or timers; or whenever the system is idle. The system transfers tasks, program u'nits, and data automatically from disk to core and
vice versa. For slow peripherals, such as the Teletype®, the INDAC 8/2 System provides for buffered output this allows the output operation to go on independently of the program processing.

""
""Teletype is a registered trademark of Teletype Corporation.
1-3

SAMPLE SYSTEM LOG
(Sheet 1 of 5)
Fold out and use this log for reference while
you build the sample system (see Chapter 2)

2-i

•
•
•
•
•
•
•
•

•
•

LOAD, DUMP OR VERIFY- L
ENTIRE DISK OR FILE~?-E

•
•
•
•
•
•'
•
•
•
•
•
•
•
•
•
•

LOAD, DU MP OR VER lFY- L
ENTIRE DliK OR FILEl?-F

LOAD, DUMP OR VERIFY-L
ENTIRE DISK OR FILEl?-F
LOAD, DUMP OR VERIFY.LOAD

* IN-R:
*ST=Z00
t1'

.LOAD
•

*1111-R:

*st:2
•

1't

• LOAD

*IH-R:

•

*ST=2

•

.LOAD
*IN-R:

*sr=2

1't

.LOAD
•

*IN-R:

*ST:Z
•

.SGEN
.SAVE <iD>ll400;

•

•
•

.HELD
.LELD
•GEtilDAC 427
TYPE '7' Ir CONYUiED.
4K7-N

*OPT-B
•

*IN-R:
t'tP

Figure 2-1

Sample System Log Output (Sheet 1of5)

2-ii

SAMPLE SYSTEM LOG
(Sheet 2 of 5)
Fold out and use this log for reference while
you build the sample system (see Chapter 2)

2-iii

•
•

II 0 LIBRARY IND AC 8/2

SAVED <SW>! 1400; FROM SGEN?-y
CHKS... STORED AT 10200
INIT !AL RU N?-Y
CZCl STORED AT 0117'

END

•
•

ADDED TO IO CHAIN •

DISABLE CTRL/C IN COMMAND MODE?-N

•
•

'0 HZ PDP-8/E LINE FREQ CLOCK?-Y
CLKB STORED AT 05102
CLK8 STORED AT 05045
CLKB STORED AT 0531'

•

PDP-8/ E PROGRAMMABLE CLOCK?- N

•

'0 HZ PDP-8/l OR PDP-8/L?-N

•

50 HZ PDP-8 CLOCK?-N

•

Gf::NERIC ~?- N

•

AF" 1 ?- N

•

AF03?- N

•

ANY or THE ABOVE A/D?-N

•

AFC4?- N

•

•
•

U2 IC>

•
•
•
•
•
•
•
•

50 HZ PDP-8/E LINE FREQ CLOCK?-N

50 HZ PDP-8/I OR PDP-8/L?-N
HZ PDP-8 CLOCK?- N

UDCP1- N

Ar02 ?- N

·~

•

AFC-8?-N

UDCl::?- N

•
•

PTP?-Y
PTP
ADDED TO <aD>
CPP1 STORED AT 01320
PTP
ADDED TO IO CHAIN.

Figure 2-1

Sample System Log Output (Sheet 2 of 5)

2-iv

SAMPLE SYSTEM LOG
(Sheet 3 of 5)
Fold out and use this log for reference while
you build the sample system (see Chapter 2)

2-v

•
•

•
•
•

•
•

TTY!?-Y

ADDED re <**>
ADDED TO c ID>

CT2l
TTY2
CT2J

ITORCD AT 01344

CT2l

ITORED AT 03,32

TTY2

•
•
•
•

ADDED TO IO CHAII.

TTY37-N
TTY4?-N
EUD Or LIBRARY TAPE •
*OPT-

GENDAC COMPLETED •

•
•

.EDIT

*OUT-i: TEST

**IN-R:

**OPT-B

•
•
•
• *
•

·COMP

*OUT-S: TEiT

*
*lli-S:TEST
*OPTICLK@
IEXE@

•

I NPU@

!DEV@

Ii VJ I@

111

ICTR@

IMODi

2244 224.ti
2215 2215
2150 2150
22 J 1 2211
2733 2733
02 J J 0211

•
•
• *
•
• *

21115 1

• SPUT

*Hl-S: TEST
*FILE i>TART 03'6

EXEC LOADED
'tA

tDRI 1
DO YOU WISH TO BYPASS DEiCRIPTIONS?
bl

THIS PROGRAM CONTAINi 3 EXERCiiERi. ALL THREE EXECISERS MAY BE CALLED
TO OPERATE AUTOMATICALLY BY TYPING ·Auro• AND ·c1R· <CARRIAGERETURN>. ANY IINGLE EXERCISER MAY BE CALLED TO OPERATE BY TYPING
.GOTO. FOLLOWED BY THE EXERCISER NUMBER (ONE-DIGIT> AND A
•c1R·. ANY EXERCISE MAY BE STOPPED BY TYPING ·srop• AND •c1R·.

•

Figure 2-1

Sample System Log Output (Sheet 3 of 5)

2-vi

•
•
•
•
•
•
•
•
•
•

SAMPLE SYSTEM LOG

(Sheet 4 of 5)
Fold out and use this log for reference while
you build the sample system (see Chapter 2)

2-vii

•
•
•
•
•
•
•
•

•
•
•
•
•
•

e:"

*PLEASE ENTER THE CORRECT TIME.

TO iET THE TIME-OF-DAY, PREii •tA" <CTRL A>, GET REIPOM5E,
PRESS 'tV' CCTRL V>, 'c' <NO CTRL>, AND THEN 'CIR'. THE EXECUTIVE WILL
TYPE OUT THE CURRENT TIME-OF-DAY AND WAIT. WHEN THE TYPING STOPI,
ENTER EXACTLY ' DIGITi AND TWO COLO~i Ai fGLLOWi:
'HH:MM:ss· FOR HOURS, MINUTES AND SECONDS. THEN TYPE "CIR'~
AFTER YOU HAVE INSERTED THE CORRECT TIME, PRESS 'tp• <CTRL P>
TO PROCEED IN OPERATOR M~DE, THEM PREii 'CIR' <CARRIAGE RETURN>
TO CONTINUE WITH THE PROGRA~.
tA
*tVC
CLOCK 0: 2: 7,08:00:0a
tP
I

THANK YOU I
YOU HAVE- TTY2 ? N
DO YOU HAVE- PIP ?Y
TURN ON PIP-TYPE CARRIAGE RETURN WHEN READY

l>O

ENTER 'AUTO' AND "CIR• FOR AUTOMATIC TESTING OR ENTER
GOTON WHERE N Ii THE EXERCISE NUMBER REQUIRED.
AUTO
0

EXERCISER #1- 20 SECOND DELAY BEGINS NOW
12345~7890:-QWERTYUIOPASDFGHJKL;ZXCVBNM,./

THE TIME Ii NOW

8: 1:11

12345,7890:-QWERTYUIOPASDFGHJKL;ZXCVBNM,./
TIME IS NOW
8: 1:31

THE

12345,7890:-QWERTYUlOPASDFGHJKL;ZXCVBNM,./
THE Tl ME IS NOW

l :5l

12345,7890:-QWERTYUIOPASDFGHJKL;ZXCVBNM,./
THE TIME IS NOW
8: 2:11
12345,7890:-QWERTYUIOPAiDFGHJKL;ZXCVBNM,./
THE TIME IS NOW
8: 2:31
12345•789~:-QWERTYUIOPAiOFGHJKL;ZXCVBNM,./

THE TIME IS NOW

B: 2:51

Figure 2-1

Sample System Log Output (Sheet 4 of 5)

2-viii

•

•
•
•
•
•
•

•
•
•
•
•
•
•

SAMPLE SYSTEM LOG

(Sheet 5 of 5)
Fold out and use this log for reference while
you build the sample system (see Chapter 2)

2-ix

•

EXERCISER 12
THIS EXERCISE DISPLAYS THE VALUE REPRESENTED BY THE CO~PUTER
SWITCH REGiiTER IN OCTAL, DECIMAL, SIGNED DlCIMAL, AND E-TYPE
NOTATION. THE EXERCISE REPEATi EVERY 20 iECOMDS FOR 3 MIIUTES IN
AUTOMATIC MODE •

•
•
•
•
•
•
• ""
•
020!;}

128

+ 128

+iil .128HlllEHl003

4i00

2r.t48

-2048

-0.2iJ48HE+li!J01114

(

7777

- fil. lliHll 0 SIDE+ liHHH

Hl.H0000E+H00

13,,

-13"

-0.13-.00E+liHiJli!J4

13•5

+13'5

+".13•!>00[+0884

+11. lli!Jli!J00111E+H81

1755

+1755

+e. l 7550SE+iJllt84

58•

- 58•

-0 .5B•008E+Hli!J3

Hli!J0

5252

2525
0~01

3333

•
·~

EXERCI6EA 13

•
•

•

•
•
•

THE TIME IS NOW
B: •:39
WORKINGl-INPUT ALLOWED AFTER NEXT TYPEOUT OF TIME
THE TIME Ii NOW
6: i:~l
STOP
END OF TEST
tA

NOTE
EXERCISER #3 was stopped by typing "STOP". This
exerciser will repeat every 15 minutes if allowed to continue. Following the "STOP) "command, the operator
typed CTRL/C (control key and C key together) to return
to the Disk Monitor System.

Figure 2-1 Sample System Log Output (Sheet 5 of 5)

2-x

•
•
•
•
•

CHAPTER 2
BUILDING THE SAMPLE SYSTEM

2.1 INTRODUCTION
This chapter contains step-by-step procedures for the new user to check out the basic software/hardware operation of his INDAC 8/2 System by building a sample system. In building the sample system, the new user becomes
familiar with some of the operating mechanics of the system which will be of benefit when he builds his specific
system.

NOTE
The PDP-8/E control panel differs slightly from
other PDP-8 Computers. The procedures presented
in this section detail the PDP-8/E controls. When
INDAC is implemented with a PDP-8/1 or L, use
the START switch whenever CLEAR/CONT is
specified in the procedures; use LOAD ADDR
whenever EXTD ADDR LOAD or ADDR LOAD is
specified in the procedures.
The log output from the loading and execution of the sample system is shown in Figure 2-1.

2.2 LOADING THE HINDAC (Tape 1) PROGRAM
To load the computer from a cold-core start proceed as follows.

NOTE
Verify that RUN light is off. If the light is on, press
HALT and return HALT switch to up position.
Step

Procedure
Load the switch register with 0000.

Press EXTD ADDR LOAD.

Load the switch register with 0027.

Press CLEAR.

Press ADDR LOAD.

(continued on next page)

2-1

Step

Procedure
Successively deposit the following:
Location

Instruction

0027
30
31
32
33
34
35
36
37
40

6011
5027
6016
7450
5027
7012
7010
3007
2036
5027

Load the H INDAC (Tape 1) program in the high-speed reader - begin anywhere in
the initial blank tape portion.

Load switch register with 0031.

Press ADDA LOAD.

Press Clear.

Press CONT.
NOTE
The tape should now read completely through the
reader and stop on the trailer portion (code 0200)
of the tape. The computer should also halt. At
this point, both RIM and the Binary Loader are in
core. If the RUN light does not go out, or if the
tape does not read in properly, repeat the above
procedure.

2.3 LOADING THE MONITOR SYSTEM DUMP AND INDAC FILE TAPES
After the H INDAC program is loaded successfully, load the Monitor Support program using the Binary Loader.
Step

Procedure
Place the MSUP tape (Tape 2) in the high-speed reader. Set the leader portion
(code 200) of the tape under the read lamp.

Load switch register with 7777.

Press ADDR LOAD.

Set switch register to 3777.

Set the rotary console switch to AC.

Press CLEAR, then CONT.
The tape should now read until the trailer portion (code 200) is under the read
lamp. At this point, the computer will halt with the AC containing 0, and the
Iink may be on.
(continued on next page)

2-2

Procedure

Step
7

Load switch register with 200.

Press ADDR LOAD.

Press CLEAR, then CONT.
The MSUP program will begin a series of questions to determine the operation required. Each question may be answered by a single letter followed by a carriage
return (designated ) ). After it is loaded, the first question is:
LOAD, DUMP OR VERIFY-

Type L)
The next question asked is:
ENTIRE DISK OR FILES?-

Type E)
The MSUP program will come to a halt after typing "E) ". The program is now
waiting for the user to load the Monitor System Dump (Tape 3) in the high-speed
reader.

At this point, before loading the tape, set the switch register to 0001. The LSB
switch is used to control the loading of the System Dump tape. When the switch
is in the "1" position, the program will idle after completion of loading the current block (one "block" of information). When the switch is set to "O", the
program will resume loading.
NOTE
Make certain that all checksums are torn from the
end of the Monitor System Dump tape before loading.

Place the Monitor System Dump (Tape 3) in the high-speed reader. Start tape at
leader portion (code 200).

Press CONT.
At this point, if you have correctly set the LSB switch, the program will be idling.

Set the LSB to "O" to start the loading process. At any time setting the LSB to
"1" will stop the tape at the next leader/trailer. Loading will resume whenever
the LSB is reset to "O".
NOTE
During the loading process, if the reader malfunctions, MSUP will print "CHECKSUM OR VERIFY
ERROR" and halt. Back the tape up one block
(blocks are groups of punches separated by leader/
trailer, code 200). Place the leader of the block that
failed under the reader lamp. Set LSB for continued loading and press CONT. If the block will not
load correctly, the tape has been damaged and must
be replaced.

When MSUP completes the loading of the Monitor System Dump (Tape 3), there
will be a slight pause and the program will type out:
LOAD. DUMP OR VERIFYf,..nn+in11art
nn
nav+... n<>na\
\V"''''-111 _ _ _ _
, , ,,....,,,
~-!j-1

2-3

Step

Procedure

Place the INDAC Support Programs (Tape 4) in the high-speed reader. Start tape
at leader portion (code 200).
NOTE
Tear all checksums from the end of the tape (information following the last leader/trailer punches).

Type L)
The next question asked is:
ENTIRE DISK OR FILES?-

Type F)
When MSUP completes the loading of the INDAC Support Programs (Tape 4)
there will be a slight pause and the program will type out:
LOAD, DUMP OR VERIFY-

Place the I NDAC System Tables (Tape 5) in the high-speed reader. Start tape at
leader portion (code 200).
NOTE
Tear all checksums from the end of the tape (information following the last leader/trailer punches).

Type L)
The next question asked is:
ENTIRE DISK OR FILES?-

Type F)
When MSUP completes the loading of the INDAC System Tables (Tape 5) there
will be a slight pause and the program will type out:
LOAD, DUMP OR VERIFY-

When both file tapes have been loaded, type CTR L/C (hold both CTR Land C
keys down); this will return control to the Disk Monitor System just loaded.
The Monitor will respond with a period. At this point, the disk system is established for a single disk and the Disk Monitor is resident in core.

2.4 LOADING AND BUILDING THE INDAC COMPILER
Load the MAKE 8/2 (Tape 6) program using the Disk Monitor Loader. To load MAKE 8/2 from the high-speed
reader, use the following command string:
.LOAD)
*IN-R:)

*
ST=200)
tt (User types CTRL/P after each t).
NOTE
Start the tape at leader portion (code 200).

2-4

MAKE 8/2 will execute and return to the monitor. Now load the Compiler tapes:
COMP1 (Tape 7)
COMP2 (Tape 8)
COMP3 (Tape 9)
COMP4 (Tape 10)
The loading sequence (using the high-speed reader) is the following for each tape:
.LOAD)
*IN-R:)

*
ST=2)
tt (User types CTRL/P after each t).
NOTE
Start tape at leader portion (code 200).

After the user's second CTR LIP, control returns to the Monitor after about one minute. The same sequence is
repeated for each compiler tape.
2.5 RUNNING THE SGEN 8/2 PROGRAM

SG EN 8/2 takes a single-disk system and expands the file structure into as many disks as are connected to the
system (Limit 4), and protects the Executive and data file areas that will be used later in the system. Any disk
units on the system that are not to be used by the I NDAC system should be switched to "off".
To run SG EN 8/2, type the following sequence:
.SGEN)
.SAVE (SD)! 1400;)
SG EN 8/2 expands the single-disk system to include as many as the user has DS32 expander units selected and
returns to the monitor after a short pause. The SAVE command initializes the System Devices Table in preparation for the system configurator program.
2.6 LOADING THE INDAC EXECUTIVE 8/2

Load the I NDAC Executive as follows:
Procedure

Step

Place the Executive 8/2 (Tape 11) in the high-speed reader. Start the tape at
leader portion (code 200).
2

Set the switch register to 3600.

Type HELD)
The HELD program will load the Executive and halt after loading.
NOTE 1
Since paper-tape to disk operations are taking place,
the loading may appear "jerky" or "erratic"; this is
normal.

(continued on next page)

2-5

Step

3
(cont)

Procedure
NOTE 2
At this point the AC register should be "O", indicating that the checksum comparison is correct. If
the AC is not "O", repeat the Executive loading
procedure.

Set the switch register to 7600.

Press EXTD ADDR LOAD.

Press ADDR LOAD.

Press CLEAR, then CONT.
NOTE
Control returns to the Monitor after pressing CONT.

This completes the procedures required to build the initial system. Continue with the following procedures to
build the specific sample system for checkout and experimental usage.
2.7

LOADING GENDAC TO CONFIGURE THE SAMPLE SYSTEM

The function of G ENDAC is to configure the INDAC System for the specific requirements of each installation.
To configure the sample system for checkout and initial experimentation, follow the log as indicated.
NOTE
The sample system contains the specific real-time
clock purchased for this system, the high-speed
paper-tape punch, and the TTY2. This configuration is specified regardless of the exact hardware
actually connected.
Step

Procedure
Load GENDAC (Tape 12) in the high-speed reader. Start tape at leader portion
(code 200).

Set the switch register to 3600.

Type LELD)

The LELD program will load G ENDAC and return to the monitor after loading.
NOTE
Since paper-tape to disk operations are taking place,
the loading may appear "jerky" or "erratic"; this is
normal. About a third of the way through loading
the tape the Teletype® will echo a carriage return/
line-feed.

____. 6
7

When loading is completed, LELD will return to the monitor.
Press HALT, then return switch to normal position.
Set switch register to 0200.
(continued on next page)

®reletype is a registered trademark of Teletype Corporation.

Step

Procedure

Press ADDR LOAD, CLEAR, then CONT.

GENDAC wi!! begin execution.

The initial dialogue from GENDAC types the version number and a reminder to
the operator that if he is unsure of the response to a question, type"?".

GENDAC will then ask if this is to be a 4K system. The response will be "N",
because you do not have a 4K system.

GENDAC will request the mode of operation to be used via the question "*OPT-",
the response will be "B" for binary mode.

The next request is for the input device via the question "*IN-" the response will
be "R:)" for the high-speed reader.

GENDAC then types "t", waiting for the binary tape to be loaded. Load the 1/0
Handlers (Tape 13) into the high-speed reader (start tape at leader code 200) and
reply by striking CTR LIP (hold CTR L key and strike P key). G ENDAC echos tP.

Answer all questions as in the log, except for selection of clocks; only respond to
the correct computer and line frequency for your configuration. In general, if you
do not understand what the routine in question is type" I" (inspect). The requests for TTY2 and PTP must be answered "Y" for the sample system.

When GENDAC reaches the end of the 1/0 Handlers tape it will type "END OF
LIBRARY TAPE and *OPT-". Then respond with carriage return to complete
GENDAC.

GENDAC will return to the Monitor System.

2.8 LOADING SAMPLE PROGRAM 2 USING THE EDITOR
Load the Sample Program as follows:

Step

Procedure
Place Sample Program 2 (Tape 14) in the high-speed reader. Make certain that all
punched checksums are torn from the end of the tape.

Type EDIT)

Follow the dialogue listed below:
*OUT-S:TEST)
*
*IN-R:)

*
*OPT-B (NOTE: Carriage return is not used)
*R)

At this point, the ED IT program will read one buffer of information from the
tape, stop reading, and return an asterisk ( *).

Type E)

The ED IT program will process the entire tape, close the TEST file on the disk
and return to the Monitor System.

2-7

2.9 COMPILING SAMPLE PROGRAM 2

Compile the Sample Program as follows:
Procedure

Step

Type COMP)
This command will begin execution of the Compiler. Follow the dialogue listed
below:
*OUT-S:TEST)

*
*IN-S:TEST)
*
*OPT-)
NOTE
Both the input and output files are allowed to have
the same name - this is a feature of the monitor
system to simplify bookkeeping for the user.

The Compiler will now begin processing. The Compiler is disk-bound and may

appear to be in a loop accessing the disk; this is normal.

The Compiler will type out the information as shown in the log, with pauses between typeouts as the information is processed.

If the Compiler generates any error conditions, there are two possible causes for
the errors:
a. Sample Program 2 has not been loaded correctly, or the checksum at the
end of the tape has not been torn off. If this is the case, go back to Paragraph 2.8, Loading Sample Program 2 Using the Editor.
b. The compiler does not recognize a device name. If this is the case, the
system was not configured properly with GENDAC. Refer to compiler
error messages, Appendix D. You have now gained the experience
necessary to proceed correctly again from Paragraph 2.3 or as follows:
(1) Press CTR L/C to return to Monitor.

(2i Load the MSUP tape from the high-speed reader using the following
command string:

.LOAD)
*IN-R:)

*
ST=200)
tt (User types CTR LIP after each t .)
NOTE
The MSUP program wil! begin a series of questions
to determine the operation required. Each question
may be answered by typing a single letter followed
by a carriage return (designated) ). After the
MSUP program is loaded, the first question is:

LOAD, DUMP or VERIFY(3) Proceed with Paragraph 2.3, Step 17.

2-8

2.10 EXECUTING SAMPLE PROGRAM 2
To run the Sample Program proceed as follows:

Step

Procedure
Type SPUT)
This command calls the System-Put Together program. SPUT will transform the
Test program to an absolute configuration used during run time. SPUT will also
automatically link to the Executive for execution.
NOTE

If any errors are generated in this portion of the system, refer to Appendix E.
2

Follow the dialogue listed below:
*IN-S:TEST)
*FILE START 0368
EXEC LOADED

At this point, the Executive is loaded and operating. This is the "idle" loop
portion of the Executive waiting for commands. To begin Sample Program 2,
follow the dialogue listed below:

(CTRL/A)

*tDR #1)

(CTR LID followed by an "R", an optional
space, a"#", a "1", and a")".)

The Sample Program will begin typing instructions. If this is the first time you
are running this system, answer "N) "to the first question asked. The dialogue
should follow the sample log. You may now answer correctly when asked about
the equipment you have. If you have TTY2, this device should be "ON-LINE"
and connected for the test. Refer to Figure 2-1 for Sample Program 2 dialogue.

2-9

CHAPTER 3
PROGRAMMING THE INDAC 8/2 SYSTEM

3.1

INTRODUCTION

The INDAC language is an application language for IDACS 8 Data Acquisition and Control Systems. Although
the language utilizes BASIC/FORTRAN like program statements, it also contains facilities for scheduling, for
specifying industrial interface equipment and for servicing field interrupts; making it a much more powerful industrial application tool. The facility for specifying industrial user interface equipment is also available for special
devices the user may wish to integrate into his system.
A process, whether it is:
Calibration and Testing
Material Preparation
Parts Manufacturing Control
Process Control (Batching)
Materials Handling
Machine Control
is controlled by an algorithm that can effect data collection; operate on process variables and parameters to develop control values (set points, on-off signals, open-close signals, etc.) and send these values to process control
equipment; activate schedules; and accept operator inputs or commands. The INDAC language permits the user
to create the algorithm that he needs using simple, easy to understand English statements.
If the algorithm is too large for the available computer core, I NDAC's segmented structure permits the user to
develop individual segments for each function of the algorithm. Each segment will be automatically maintained
on mass storage. The user can then schedule these segments using INDAC's scheduling capability so that each
segment is executed as though the whole algorithm were core resident. Since INDAC has a segmented structure
and certain key words must be used to identify executable code from scheduling parameters, program skeletons
are used in this chapter to place the segments into perspective for the reader.
Basic Skeleton

.PHASE
.ACTION

.SNAP
.PROCESS
.END

The above skeleton shows the two basic segments of the INDAC language: the PHASE and the SNAP. The
PH.A.SE contains the key word .ACT!O!\J; the S!'J.fl.P contains the key . .vord .PROCESS. The last statement of the
job is .END.
3-1

3.2 PROGRAMMING AN INDUSTRIAL ALGORITHM
An algorithm contains routines and/or subroutines for handling process control functions such as:

a. Conversion of process variables to engineering units
b. Conversion of calculated control values for driving process equipment

c. Decision making based on process variables and parameters
d. Initializing

e. Sampling
f.

Limit checking

g. Normalizing

h. Scaling

Optimizing Control Parameters

The I NDAC language provides the user with the necessary facilities for coding these and many other algorithmic
functions. Some of the basic facilities the I NDAC language provides are:

a. Arithmetic including:
Add
Subtract
Multiply
Divide
Exponentiation
Square Root

b. Bit isolation facilities for:
Testing
Setting
Resetting

c. Relational Comparison for process decisions including:
Greater than
Greater than or equal to
Equal to
Not equai to
Less than or equal to
Less than

d. Multilateral branching

e. Looping
f.

Subroutines

Skeleton No. 1 shows the basic structure of an INDAC program and where the algorithm must be placed. To compile an I NDAC source program without error, the program must contain a .PHASE and a .SNAP. The PHASE
must contain at least one scheduling statement following the key word .ACTION. The algorithm must appear in
the .SNAP and must be bracketed by the key word .PROCESS. The last statement of the job is .END.

3-2

Skeleton No. 1 - The Algorithm
.PHASE

LET

.ACTION

IGOTO
IF

DO SNAP #2 PRIORITY 1

FOR/NEXT

DO/RETURN

INDUSTRIAL ALGORITHM

.END
3.2.1

LET Statement (Assignment)

To resolve process parameters or process variables in developing control values or set points, "Arithmetic" will
have to be performed in the algorithm. The LET statement allows the user to do arithmetic. The LET statement
also allows the user to deal with the computer bits (Boolean type operations) as though he were opening or
closing valves, turning solenoids on or off, setting indicators high or low, etc.
3.2.1.1

Arithmetic - Consider the following:
LET HI LIMIT= 80.0
LET LOLIMIT = 60.0

Notice the use of the LET statement to assign a fixed parameter to a name. The assigned name can then be used
in a limit checking algorithm. Assigning a value to a name rather than using the value itself in the program provides the option of changing the value at a later time. This option may be required in applications where dynamic
limits prevail such as checking reject figures under varying production rates. To reduce HI LIMIT from 80.0 to
70.0 one simply writes:
LET HILIMIT = HILIMIT- 10.0
The resulting value of an arithmetic expression can also be assigned to a variable name. For example:
LET RANGE= HILIMIT- LOLIMIT
The expression, HI LIMIT - LOLIMIT, is evaluated during program execution and the result (+10) is transferred
to the assigned name RANGE.
In arithmetic assignment statements, the assigned name and the expression or fixed parameter represents either
integer or real quantities. Integer quantities require less core storage than real quantities because an integer
quantity is represented by one computer word, while a real quantity is represented by three computer words.
The criteria for assigning a quantity a real or integer representation is its size. Integer quantities are I im ited to
-2048 to +2047 inclusive. Real quantities are within the range ±1 x 101±99 l.
To distinguish integer from real variables in the assigned name, the prefix I is used for all integer representation.
Assigned names of real variables are prefixed by any alphabetic character except I.
The preceding examples of the LET statement define real variables and expressions. These statements can be rewritten to define integer variables and expressions as follows:
LET IHILIMIT = 80
LET ILOLIMIT = 60
LET I RANGE= IHILIMIT- ILOl IMIT

3-3

Notice that the integer constants, 80 and 60, do not use a period as do the real constants.
The LET statement allows the user not only to define and then operate on the data, but also to redefine and operate on process data collected from sensors. For example:
LET TEMP= REFERENCE+ SENSOR
IF TEMP GE 90.0 THEN GOTO 21
GOTO 50
LET !VALVE= 1

The LET statement in this example shows how two data items (REFERENCE and SENSOR) that have been collected from the process are added together and defined as TEMP. A decision, using the IF statement, is then
made using the variable named TEMP.
Arithmetic expressions for the LET statement may be formed using the INDAC arithmetic operators detailed in
Table 3-1. A comparison between the algebraic and the I NDAC notations is also given in the Table 3-1. The
primary differences are that the operator (symbol) used to denote arithmetic must appear, and all data must be
on the same line (that is, %notation not permitted).
Table 3-1
Basic Arithmetic Notation, Algebraic vs. INDAC
Notation
Differences in Notation

Algebraic

INDAC

Arithmetic Operation

A+B

ADDITION

NONE

A-B

SUBTRACTION

NONE

AXB
or
A·B
A
13or A 7 B

A*B

MULTIPLICATION

An asterisk is used as the I NDAC operator

A/B

DIVISION

A slash is used as the I NDAC operator

At 2

EXPONENTIATION

An up arrow is used as the INDAC operator
instead of a superscript

Two major factors that must be observed when writing arithmetic expressions in INDAC are:

a. Compatibility of data type
b. INDAC-established precedence in the execution of arithmetic operations.
Except for exponentiation, unlike data types (integer vs. real) cannot be used in the same expression. In exponentiation, the exponent is always an integer. The following examples illustrate both acceptable and unacceptable expressions based on data compatibility.
Acceptable

Comments

Unacceptable

Comments

HI LIMIT+ LOLIMIT

Both real
variables

IHI LIMIT+ LOLIMIT

Mixed types, IHILIMIT is integer,
LOLIMIT is a real variable

IHI LIMIT+ I LOLIMIT

Both integer
variables

HILIMIT + ILOLIMIT

Mixed integer I LOLIMIT and real
HI LIMIT variable
(continued on next page)

3-4

Comments

Unacceptable

Comments

Acceptable
5 +I RANGE

Both integer
types

5 +RANGE

Mixed integer number and real
variable

5 + 10

Both integer
numbers

5 + 10.2

Mixed integer and real numbers

12E5 +RANGE

Both real data
types

12 E5 +I RANGE

Mixed real number and integer
variable

1.2 t 5

Mixed types
permitted

1.2 t 5.3

Power must be expressed as an
integer number

Evaluation of arithmetic expressions is carried out according to the following precedence of execution:
Precedence

Operator

Operation

Exponentiation

Multiplication and division

Addition and subtraction

NOTE
Where operations have the same precedence of execution, they are evaluated from left to right, according to their occurrence within the expression.
The steps that are performed by I NDAC in evaluating the expression 2 t 2 * 3/2 + 1 according to the established
hierarchy are:
Step 1

Raise 2 to the 2 power

:22 = 4

Step 2

Multiply result by 3

3 x 4 = 12

Step 3

Divide product by 2

12/2 = 6

Step 4

Add 1 to the quotient

6+1=7

Parentheses are used to group terms and factors of an arithmetic expression to specify a desired sequence
(hierarchy) of execution.

Operations grouped within parentheses are always performed first. For example, in the expression:
5 * ( 1 + 3)
the grouped expression is evaluated first (that is, 1 + 3 = 4); then the remaining operation is carried out (that is,

5 * 4 = 20).
Multilevel grouping by parentheses is permitted in INDAC (for example, (((A+B) +C) *D)). In multilevel grouping, the evaluation of the expression is carried out starting with the innermost group and proceeding in sequence
to the outermost group, as illustrated in the following example:
(... ((((FIRST) SECOND) THIRD) FOURTH) .... )
The manner in which grouping affects INDAC's evaluation of an expression is illustrated by the following versions
of the expression 2 t 2 * 3/2 + 1.

3-5

Examples:

2 t (2 * 3)/2 + 1

2 t (6) /2 + 1

64/2 + 1

32 + 1

( (2 t 2) * 3) /2 + 1

( (4) * 3) /2 + 1

( 12) /2 + 1

6+1

((2t2)*3)/(2+1)

((4) *3) I (2 + 1)

(12) I (3)

3.2.1.2 Boolean - Consider the following
LET IMASK = '1000
LET IOUT = '4000
Note the use of the LET statement to assign an octal constant to a name. In INDAC, octal constants are identified by an apostrophe preceding the number. Octal constants are used in Boolean operations to isolate or set bits
in bit patterns. The assigned name must be an integer type because all INDAC bit patterns are limited to one
computer word. The reason that octal constants rather than decimal constants are used in Boolean operations is
that octal constants are easily identifiable with a specific bit pattern. A named variable may also be set to the
result of a Boolean expression. For example, to test the state of a bit (0 or 1) one may write
LET IBIT = IMASK AND !PATTERN
The expression IMASK AND !PATTERN is evaluated on a bit-by-bit basis during program execution and the result is transferred to the assigned name. Remember that the IMASK isolates the bit of interest. If bit 2 of
IPATTE RN is set then the result will be 10008 because IMASK = '1000, but if bit 2 is not set, then the result
will beO.
NOTE
Engineering conventions prevail on the PDP-8 and
bits are labeled left to right.
To set a bit to "1" one may write:
LET !PATTERN= IOUT OR !PATTERN
The expression IOUT OR !PATTERN is evaluated on a bit-by-bit basis during program execution and the result is
transferred to the assigned name. Assuming that iOUT is set to 40008 (bit 0 is set to i), the resuit iiPATTERN}
will be that bit 0 will be set and the remaining bits will not be affected.
Boolean expressions for the LET statement may be formed using the following I NDAC Boolean operators:

a. NOT
b. AND
c. OR

3-6

Some examples of Boolean expression using the above operators follow:

a. IAB OR ICD
b. IAB AND ICD
c. IAB OR '1777
d. IAB AND '1777

e. NOT IAB
f.

NOT (IAB AND ICD)

In AND operations, the resultant of the corresponding bits of the logical data being combined is a" 1" only if
both corresponding bits are "1". To illustrate, given that:

a. ICHA = 010 100 111 000
b. ICHB = 101 000 101 000
the value of the dependent variable IR ES in the expression:
LET IR ES= ICHA AND ICHB
results in
IR ES= 000 000 101 000
The AND operation is used to isolate one or more bits of a logical word for testing or to eliminate a given bit in
the pattern so that their individual states (states of the devices and/or variables they represent) may be evaluated.
Isolating data using the AND operation is performed by ANDing the data word with a mask word designed to
eliminate all unwanted data bits from the resultant. The mask word must contain a "O" corresponding to each
unwanted data bit and a "1" for each desired bit. For example, to isolate the third bit of a word (that is, bit 2
in word format):
Bit Pattern

Octal

Bit No:

0 1 2 3 4 5 6 7 8 9 10 11

Equivalent

Data Word

1 1 1 1 1 1 1 1 1 1

7777

Mask Word

0 0

1 0 0 0 0 0 0 0

1000

Resultant

0 0

1 0 0 0 0 0 0 0

1000

AND

Mask words are expressed in octal forms as constants or as integer variables.
In 0 R operation, the corresponding bits of the words being combined result in a" 1" if either or both of the combined bits are "1 ". To illustrate, given that:

a. ICHA = 010 100 111 000
b. ICHB = 101 000 101 000
the expression LET I RES= ICHA OR ICHB results in
I RES = 111 100 111 000
If either of the bits being ORed contains a "1", the result will contain a "1 ".

3-7

The NOT operator inverts or complements the value of any logical data or expression which it precedes. For
example, if
ICHA = 010 100 111 000
NOT ICHA = 101 011 000 111
The order in which the logical operators are evaluated by INDAC is:
Precedence

Operation
NOT

AND

As with arithmetic expressions, I NDAC always evaluates quantities within parentheses first, regardless of the operation involved. This feature permits the user to dictate the order of evaluation by grouping within parentheses.
In the following expression, the result is changed by the use of parentheses to alter the sequence in which the
expression is evaluated. For example, the basic expression:
ISW1 AND ISW2 OR NOT IAB
may also be represented as
(ISW1 AND ISW2) OR (NOT IAB)
The meaning of the expression is not altered by adding the parentheses as shown above because the basic order
of evaluation is not changed by the parentheses. The result of ISW1 AND ISW2 is still ORed with NOT IAB resulting in a "1" if a given bit is a "1" in both ISW1 and ISW2 or is a "O" in IAB.
Now consider the same expression with the parentheses placed as follows:
ISW1 AND (ISW2 OR (NOT IAB))
In this expression NOT IAB is still resolved first. Since the expression formed by the OR operator is in parentheses, it takes precedence over the expression formed by the AND operator, thereby changing the result of the
whole expression. Now, the result of ISW2 OR NOT IAB is ANDed with ISW1 resulting in a "1" if a given bit is
a "1" in ISW1 and the result of ISW2 OR NOT IAB is a "1". The result of ISW2 OR NOT IAB is a "1" if ISW2
is a" 1" or if IAB is a "O". As shown above, the result of the expression was changed by adding parentheses to
alter the sequence in which the expression was to be evaluated. Therefore, care must be taken in writing Boolean
expressions to ensure that they are interpreted as intended.
3.2.1.3 Statement Evaluation Rules - Since assignment statements may cause the transfer of integer, real, or
logical values into either integer or real variables or arrays, the evaluation of this type of statement is performed
according to the following rules (where LET v = e):
Assigned Name (v) if

And Expression (e) is

INDAC does the following

Integer

Transfers the result of e, unchanged, to v

Integer

Real

Truncates any fractional part of e and transfers it as an integer to v

Integer

Logical

Transfers e, unchanged, to v
(continued on next page)

3-8

Assigned Name {v) if

And Expression {e) is

INDAC does the following

Real

Integer

Transforms value of e to a real data form and
transfers it to v

Real

Transfers the result of e, unchanged, to v

Real

Logical

Not permitted

3.2.2 GOTO Statement {Branching)
The GOTO Statement provides the user with a convenient way to transfer control from one part of his program
to another. The statement can be used to link procedures in the algorithm. Control can be transferred unconditionally using the fol lowing statement:
GOTO 21
21

LET ITEMP =I REFERENCE+ ISENSOR

This statement, which references a statement label, can be used anywhere in the process algorithm to transfer
control to another part of the algorithm.
NOTE
A label is an integer number from 1 to 999 that precedes an executable statement so that it can be referenced by a GOTO statement.
Control can also be transferred conditionally using the following statement:
GOTO (21,25,29), INDEX
This statement is the computed GOTO statement that can be used to transfer control to one of many possible
alternative procedures based on the value of an integer control parameter or index (I ND EX). Statement labels
identifying the starting point of alternative procedures are enclosed in parentheses. There is virtually no limit to
the number of labels that can be referenced. If the index of the computed GOTO statement is 0, negative, or
greater than the number of referenced labels a default to the statement following the GOTO will occur. The
integer quantity INDEX referenced in the above example may be derived in a number of different ways depending on what the user wishes to do. One way to generate the index is shown below:
LET INDEX= 0

LET INDEX= INDEX+ 1
GOTO (21,25,29), INDEX
GOTO 2

These statements set INDEX to 0 and then increment INDEX by one. When used as one control parameter of
the computed GOTO statement, INDEX acts as a switch causing a transfer to the statement labeled 21 if INDEX
is 1, or to the statement labeled 25 if INDEX is 2, or to the statement labeled 29 if INDEX is 3. If INDEX is
either negative, 0 or greater than 3 the computed GOTO statement defaults to the next statement in the program.
The following program and flow chart illustrate the operation and power of a computed GOTO statement.

3-9

LET INDEX=1
GOTO 2
21

LET INDEX=
INDEX +1

1
2

LET INDEX= 1
GOTO (21,21,22,22),INDEX

default
LET INDEX= INDEX+ 1

GOTO 2
LET INDEX= INDEX+ 1

LET INDEX=
INDEX+1

GOTO 2

DEFAULT
08-0726

3.2.3 IF Statement (Conditional Testing)
In any process algorithm, decisions have to be made based on collected or calculated process data so that the
desired controls can be activated.
The IF statement is a control statement that allows the user to compare process data (quantities or expressions).
The statement uses the words IF and THEN GOTO to select another process procedure (program sequence) according to the outcome of the comparison. The statement employs a relational operator to make the comparison
and a GOTO statement to branch to the alternative procedure.
IF TEMP GR 90.0THEN GOTO 4
This version of the IF statement uses an unconditional GOTO statement. The real variable TEMP may be the
product of an arithmetic operation using the LET statement or may represent data collected from the process.
GR is a relational operator meaning greater than. Six different reiationai operators are at the disposal of the user.
These are defined in Table 3-2.
Table 3-2
Relational Operators
Operator

Symbol

Greater than

Greater than or equal to

Equal to

Not equal to

Less than or equal to

Less than

LE
LS

Meaning

3-10

Relational operators can only be used with the IF statement. The sample IF statement above compares real
quantities. Real assignments require three computer words of storage and should be used only when the quantities or expressions can not be represented in one computer word (an integer quantity between -2048 and +2047
inclusive). The ! F statement can be rewritten to compare integer quantities as fo!!m.vs:
IF ITEMP GR 90 THEN GOTO 4

YES

NEXT STATEMENT

08-0723

Notice that the integer constant (90) does not use a period but the real constant (90.0) does. Both sample IF
statements above use an unconditional GOTO statement. If the quantity represented by TEMP (ITEMP) is
greater than 90.0 (90), then program control is automatically given to the statement labeled 4; otherwise, the
next statement in the program is executed.
NOTE
A label is an integer number from 1 to 999 that precedes an executable statement so that it can be referenced by a GOTO statement.
A computed GOTO can be used in the IF statement instead of one unconditional GOTO. This capability allows
the user to resolve decisions where there are multiple alternatives in the algorithm.

IF ITEMP GR 90THEN GOTO (4,5,6),INDEX
YES

PROCEDURE 1

STATEMENT 4

PROCEDURE 2

STATEMENT 5

PROCEDURE 3

STATEMENT 6

DEFAULT
PROCEDURE

FOLLOWING
IF STATEMENT

08-0724

3-11

If the value of ITEMP is greater than 90, then the value contained in I ND EX is used as a control parameter to
branch to labeled statements. (Refer to GOTO description.)
If INDEX is neither 1, 2 nor 3, or if !TEMP was not greater than 90, then the statement following the IF statement is executed (called a default condition).
The quantities and/or expressions being compared in an IF statement may be integer or real, but not mixed; for
example:

a. ISW1 GR ISW2 (acceptable, both terms are integer)
b. ABSA GR ABSB (acceptable, both terms are real)

c. ISW1 GR ABSA (not acceptable, terms are mixed types)
The terms of a relational expression may be simple or complex. For example:

a. (ISW1-5) GR (ISW2-5)
b. (12E-05*ABSA) GE (128E05*ABSB)
3.2.4 FOR/NEXT Statements (Looping)
Repetitive operations can be implemented by using the FOR/NEXT statements to form a loop. For example:
LET IMASK = 1
FOR IPASS = 1 TO 12 STEP 1
IF !MASK AND I PATTERN NE 0 THEN GOTO 10
LET !MASK= IMASK * 2
NEXT !PASS

LET !ALARM= IPASS

LET IMASK=1

!PATTERN IS DIGITAL
PROCESS DATA (12 BITS)
WHERE EACH BIT
REPRESENTS A
DIFFERENT ALARM SENSOR

LET !ALARM=
IPA SS

LOOP
SEND ALARM
MESSAGE

08-0725

Notice that the FOR statement precedes a sequence of instructions that are to be repeated. The FOR statement
defines a control variable (!PASS) and specifies the number of times the instructions following are to be repeated.
The number of times the loop is to be repeated is defined by an initial parameter, a terminal parameter, and an
incremental parameter ( 1 TO 12 STEP 1). The NEXT statement is the last statement in the loop. This statement

3-12

signals the end of a pass through the instructions and initiates a test to determine if the number of passes specified
have been executed. After the last pass through the loop, control is automatically released to the statement following the NEXT statement. The FOR/NEXT flow chart illustrates the operation of the FOR/NEXT loop example presented on the previous page. The example illustrates how a FOR/NEXT loop can be u~d to test a
particular computer word to see which bit is set. In this example, the variable !PATTERN which was collected
from the process is to be tested. This variable is a 12-bit computer word, where each bit is associated with a different process alarm sensor. If during the test, a bit is found to be set, a unique alarm message can be issued
using IALARM as an index. Notice that the integer parameter !MASK is set to "1" before entering the loop.
IMASK is then multiplied by 2 before repeating the next pass through the loop, resulting in an !MASK variable
that has a progressive value of the powers of two. The IF statement in the loop is used to test each bit. !MASK
is used to isolate bits in !PATTERN using the AND operation; the relational operator NE is used to test if the
specific bit is set ("1 "). Table 3-3 shows the powers of two progression with the octal equivalent associated with
the alarm index.
Table 3-3
Mask Generation
IMASK
Decimal

Octal

0001

0002

0004

0010

0020

0040

!PASS/ALARM BIT
1

0100

178

0200

256

0400

512

1000

1024

2000

-2048

4000

Once a bit is found to be set, the control variable (!PASS) of the FOR/NEXT loop can be used to inform an operator of the indicator in alarm or to take programmatic action in the process.

NOTE
PDP-8 engineering conventions prevail. The bits of
a PDP-8 computer word are numbered from left to
right and not from right to left reflecting the programming convention of the binary weight of the
bits. Therefore, when the algorithm is developed,
the user must keep this characteristic in mind.
In the preceding example, the FOR statement specified an incrementing parameter of 1 (STEP 1) for the loop
control variable !PASS. This resulted in 12 passes through the Loop because after each pass IPASS was incremented by 1.

3-13

NOTE
The incrementing parameter (STEP) may be omitted
for increments of 1. However, if increments other
than 1 are desired, STEP must be included to specify
the increment.
If we change the incrementing parameter to 3 (STEP 3) then the FOR/NEXT loop would only make 4 passes and
would test alarm sensors 1, 4, 7, and 10 because the control parameters will progress from 1 to 4 to 7 to 10 and
then release control.
In addition, the preceding example used fixed parameters (integer quantities) in the FOR statement to define the
number of times the loop was to be executed. Named integer variables can also be used to define the number of
passes to be performed. For example:
LET IMASK = 1
FOR IPASS = ISTART TO !STOP STEP INCREMENT
IF !MASK AND I PATTERN NE 0 THEN GOTO 10
LET IMASK =!MASK* 2
NEXT IPASS
10

LET !ALARM= !PASS

Using named integer variables to define the number of passes the loop is to execute provides the option to change
the FOR statement under program control. This option may be helpful, as related to the example, to selectively
test alarm messages based on process variable evaluations. If the following assignment statements preceded the
FOR/NEXT loop, then the loop would make 6 passes, thereby permitting any of the 6 alarm sensors to be tested:
LET ISTART = 1
LET ISTOP = 6
LET INCREMENT= 1
The previous examples of the FOR/NEXT loop illustrated how 12 indicator alarms can be tested. Many industrial applications contain more than 12 such alarms. In these applications, more computer words would be required for computer storage. The simple FOR/NEXT loop developed in previous examples only tested the bits
of a single word. However, this single loop can be incorporated (inserted) into a larger loop to test additional
words. inserting one ioop into another loop is called nesting. In the following example, the parameter IWORD
is used to select the word for the test and the parameter IPASS is used to test the individual bits of the word.
The following example illustrates how 5 alarm indicator words can be tested:
FOR IWORD = 1 TO 5
LET !MASK= 1
FOR !PASS= 1TO12
IF IMASK AND !PATTERN (!WORD) NE 0 THEN GOTO 10
LET IMASK = IMASK * 2
NEXT IPASS
NEXT IWORD
10

LET IOFFSET = (IWORD -1) * 12
LET IALARM = IOFFSET +!PASS

3-14

In the earlier example, the alarm number was the same as the pass number. In this example, the alarm number,
which could be from 1 to 60 (5 computer words) is calculated. An offset based on the IWORD parameter is
established. This offset assumes a value that is based on the value of IWOR D ( 1 through 5) when a set bit is
found. IOFFSET can assume the foiiowing vaiues:

0 when IWORD = 1
12 when IWORD = 2

24 when IWORD = 3
36 when IWORD = 4
48 when IWORD = 5
With this offset defined, the alarm number can be determined by simply adding IOFFSET and IPASS.
Nesting is allowed as long as the range of one loop does not cross the range of another loop. The following diagrams illustrate acceptable nesting techniques.
Unacceptable Nesting Techniques

Acceptable Nesting Techniques
Two-Level Nesting

Three-Level Nesting

FOR

FOR
[NEXT

[
[

NEXT
FOR

NEXT
NEXT

FOR

FOR
[

LNEXT
NEXT
An inner (nested) loop is cycled through its entire range for each cycle of the outer (primary) loop. If, for example, the primary loop has an iteration of 5 cycles and the inner loop an iteration of 10 cycles, the nested loop
will have completed a total of 50 cycles when the outer loop reaches its limit of 5 cycles.

3.2.5 DO/RETURN Statements (subroutines)
When particular process variables or parameters are evaluated several times, or a sequence of instructions must be
executed several times in an algorithm, the user may code these op~rations as a subroutine that can be called by
the algorithm when needed. Two types of subroutines are permitted in INDAC:

a. Internal
b. External
Internal subroutines are an integral part of the algorithm, while external subroutines are stored on the disk separately. External subroutines are discussed later in this chapter. An internal subroutine is called into operation
by the DO statement referencing a label. For example:
DO 21
The statement labeled 21, which should be the first instruction in the internal subroutine, and all statements until
a RETURN statement is encountered are then executed. Executing the RETURN statement terminates the subroutine and returns control to the statement directly following the call DO 21. The following example shows the
techniques for installing a subroutine into the algorithm.

3-15

DO 21
GOTO 50
(first subroutine statement)

Other uses of the DO/RETURN statements are discussed later in this chapter.

3.3 MESSAGE AND LOGGING CAPABILITIES
Messages, Reports, and Logs may be required to inform the operator of alarm conditions, process states, process
variables, and parameters or any other information that may be significant in implementing and keeping the
process running smoothly. To illustrate the reporting and logging capabilities of the INDAC language, the examples for testing alarm indicators developed in programming an industrial algorithm (Paragraph 3.2) will be expanded.
NOTE
Skeleton No. 2 shows the basic structure of an
INDAC program and shows where the algorithm
and information pertaining to the Message, Report
or Log is to be placed. A sample alarm test routine
is included in the algorithm section.
Skeleton No. 2 - Sample Alarm Test Routine

.PHASE
.ACTION
DO SNAP #2 PRIORITY 1

.SNAP

I.PROCESS

. ______,,,_,_,,,,_,,.,,_~,,_,.,._,,,,_,. , __~,:)-111(-------~1 GET /SEND
LET !PATTERN= '0100
LET IMASK = 1

FOR !PASS= 1 TO 12
IF (!MASK AND I PATTERN) NE 0 THEN GOTO 21
GOT010
21

LET !ALARM = IPASS

. '1----------~

f
10

r..i::T/C::.i::l\ln

l'-''-1/V'-1''-'

LET IMASK = IMASK*2
NEXT IPASS
.END

MESSAGE AND LOGGING

3-16

Messages, Reports, or Logs can be sent to the Teletype, the paper-tape punch, or other ASCII type output devices
using a simple SEND statement in the algorithm. The SEND statement must identify the device name, the message to be sent, and a data list if data is to be sent with the message.
SEND (TTY,#101)
This statement sends the message tagged #101 to the Teletype. Notice the device name and the tag are separated
by a comma and are enclosed in parentheses. A message without accompanying data is identified by the
.HEADER statement as follows:
#101

.HEADER
ALARM INDICATOR REPORT

The .HEADER statement must always be tagged so that the message can be referenced in the SEND statement.
The message declared in the .HEADER statement can occupy more than one line.
To send the same message to another ASCII type output device, it is only necessary to change the device name in
the SEND statement. For example, to send to the paper tape punch:
SEND (PTP, #101)
Operators consoles may not be attended so when a message is sent, it may be desirable to report the time of day
the message was sent. For this reason and for scheduling purposes, the I NDAC System contains a real-time clock
(CLOCK) that maintains the time of day. The time of day is maintained as integer quantities in three computer
words of storage; one for hours, one for minutes, and one for seconds. Before the time of day can be reported
it must first be obtained. The time of day can be obtained using a simple GET statement in the algorithm. The
GET statement must identify the device name (pseudo-device) and a data Iist of three integer variables.
GET (CLOCK) ICLK1, ICLK2, ICLK3
This statement gets the time of day from the real-time clock and stores it in the specified data list. Notice each
integer variable specified in the list is separated by a comma. Now the time of day may be sent with the alarm
message using a simple SEND statement as before. This time, however, a data list must be included in the statement to sehd the time of day with the message
SEND (TTY,#102) ICLK1, ICLK2, ICLK3
This statement sends the message tagged #102 and the data list to the Teletype. Notice the use of parentheses
and commas as described earlier.
A message with accompanying data is identified by a FORMAT statement. INDAC offers two types of .FORMAT
statements: picture type and declarative type. A picture .FORMAT statement for the alarm message and accompanying data is written as follows:
#102

.FORMAT
"ALARM INDICATOR REPORT AT: "XX": "XX": "XX

The .FORMAT statement must always be tagged so that the message can be referenced in the SEND statement.
Notice the use of quotation marks and Xs in the message. The quotation marks are used to bracket characters,
thereby identifying the characters of the message to be printed. Each digit of the data to be printed is represented
by an X. Military time representation is used in INDAC.

3-17

0 - 23 for hours
0 - 59 for minutes
0 - 59 for seconds
If the time is 11 :30 a.m., the above .FORMAT statement causes the time to be printed as follows:
ALARM INDICATOR REPORT AT: 11 :30:00
Notice digits for hours, minutes, and seconds are printed where the Xs were placed in the .FORMAT statement.
Since the colons (:) are declared as printable characters of the message, the colons are also printed. To send the
same message to another ASCII type output device, it is only necessary to change the device name in the SEND
statement. For example:
SEND (PTP,#102) ICLK-1, ICLK2, ICLK3
The message declared in the picture .FORMAT statement cannot exceed one line of a printing device. If the line
length of the input device (device program is prepared on) is shorter than the output device (device to which the
message is sent), then a continuation line can be used to prepare a message that will extend the full line length of
the output device. A continuation line is indicated by typing an up-arrow (t). Another limitation of the picture
type .FORMAT statement is that it can only be used to provide formats for integer or real decimal data having a
maximum of 15 digits including sign.
The declarative .FORMAT statement is more powerful than the picture type .FORMAT statement. The declarative .FORMAT statement permits the use of a virtually unlimited number of continuation lines (indicated by an
up-arrow, t) and permits the user to specify the data to be output as fixed point, floating point, signed, unsigned,
octal, binary, exponential, or ASCII representation. The statement also permits the user to specify special characters for CR/LF Editing and for control of the operator console modes. A declarative .FORMAT statement for
the Alarm message and accompanying data is written as follows:
#103

.FORMAT ("ALARM INDICATOR REPORT

AT: ",R3.0":",R3.0":",R3.0)

The ASCII string to be printed is bracketed by quotation marks as in the picture FORMAT statement. The output data format however, is represented differently. The commas indicate the start of a data format declaration.
Also, notice that the entire message and data declaration is enclosed in parentheses. In the above example, R3.0

means the data is to be output as 2 unsigned digits with no decimal point R3.1 means that the data is to be output as 2 unsigned digits with the decimal point between the two digits. The SEND statement is the same whether
the message is defined by a picture or a declarative.FORMAT statement, that is:
SEND (TTY,#103) ICLK 1, ICLK2, ICLK3
The sample algorithm in the skeleton (starting with LET !PATTERN= '0100) is similar to those developed in
programming an industrial algorithm, Paragraph 3.2. Notice that the sample algorithm in the skeleton is structured to continue testing after an alarm is found.
When an alarm is found (a bit is found to be set) by the IF statement in the loop, the statement labeled 21 is then
executed to assign the IPASS parameter to I ALARM. The value of !ALARM will then be the number of the
sensor that caused the alarm. The following SEND statement can then be used to report the sensor number that
caused the alarm.
SEND (TTY,#104) !ALARM

3-18

As explained earlier, the tag should reference a .FORMAT statement since data is to accompany the report. The
statement may be of the picture or declarative type. For example:
Picture Type
#104

.FORMAT
"SENSOR NO." XX

Declarative Type
#104

.FORMAT ("SENSOR NO.", R3.0)
NOTE
If no FORMAT is declared in the SEND statement,
the data list will be output as real data in exponential type notation of the following form ( 16 data
items per line).
±0.XXXXXXE±XXXX

Some of the statements developed in this paragraph have been placed in perspective for the reader in the following update to skeleton No. 2 for instructive purposes. Notice that all message declarations .HEADER and
.FORMAT statements are located between the key words .SNAP and .PROCESS, and all executable statements
are located in the algorithm section between .PROCESS and .END.
Skeleton No. 2 - Header/Format Details

SEND (TTY, #104) IALARM

,· · Ui~::, ·,:·::::::··.::.~:~>'.;~~~:1•~~'ii.Ms1'~!~·:.····'····
NIXTt1>AS$

·. . :. ,; ·:;~~·~IMJ

3-19

In the previous examples illustrating the message and logging techniques of the INDAC language, the parameters
IMASK and !PATTERN, and the storage location for the time of day (ICLK1, ICLK2 and ICLK3), were defined
in the algorithm. Parameters (or constants) and storage locations can also be defined with the .STORAGE statement.
The following .STORAGE statement defines the integer parameters !PATTERN, IMASK and integer storage locations for the time of day .
.STORAGE IPATTERN/1/,
ti MASK ( 1, 12)/'0001,'0002,'0004,'0010,'0020,'0040,
t'0100,'0200,'0400,'1000,'2000,'4000/,ITIME ( 1,3)
Continuation lines, indicated by an up-arrow (t), are permitted in the STORAGE statement.
Notice the use of names, numbers in parentheses and numbers between slashes. The name identifies the parameter or storage location so that it can be referenced in the algorithm. The two numbers separated by a comma
in parentheses define the number of storage locations (commonly referred to as a dimensioned array) to be defined for the named variable. When the number of words are not defined, as in !PATTERN, only one storage
location is set aside. A storage location is one computer word for integer variables and three computer words for
real variables. The number enclosed by slashes are the values to which the storage locations of the array are to be
set. For real arrays the preset values must contain a period. If no preset values are specified (as in IT IM E) the
locations will be automatically set to 0. Notice that the integer storage locations defined in the .STORAGE statement can be preset with decimal or octal quantities. Having the parameters and storage locations defined in the
.STORAGE statement, a simpler code can be developed for the algorithm as follows:

GET (CLOCK) ITIME
SEND (TTY, #102) ITIME
FOR IPASS = 1TO12
IF IMASK (IPASS) AND !PATTERN NE 0 THEN GOTO 21
GOTO 10
21
10

LET !ALARM= IPASS
NEXT IPASS
.END

The statement
GET (CLOCK) ITIME
obtains the time of day from the real-time clock as before but now stores the three words in the predefined array
called ITIME. Therefore, to report the time of the alarm message the following SEND statement is used.
SEND (TTY, #103) ITIME
No change is required in the .FORMAT statement itvhen stornge locations are predefined. All three 1,.vords of
ITI ME are automatically transferred since the whole array is declared in the data list of the SEND statement.
Notice that the statements that defined !PATTERN and IMASK and the statement that generated the powers of
two progression in the earlier algorithm are no longer needed. Removing these statements makes the Algorithm
faster to operate since the multiply operation is no longer used. Also notice that the IF statement has been
changed slightly. The statement now reads:

3-20

IF !MASK (!PASS) AND I PATTERN NE 0 THEN GOTO 21
I MASK is now subscripted with !PASS to step through the IMASK array defined in the .STORAGE statement.
Subscripted variables provide the user with additional computing capabilities for dealing with lists, tables, matrices,
or any set of related variables defined in the .STORAGE statement. In I NDAC, variables are allowed one subscript. A named variable followed by a constant or a named integer variable in parentheses indicates the location
of the variable in the list. This facility allows the user to reference any data item or storage location defined in
the .STORAGE statement. In the previous example, IMASK is subscripted with IPASS and IPASS assumes the
values 1 through 12 as the loop is executed. Therefore, for each pass through the loop, the corresponding IMASK
parameter in the list defined in the .STORAGE statement is used in the IF statement.
The .STORAGE statement and Alarm test algorithm developed in the previous paragraph have been placed in
perspective for the reader in the following update to Skeleton No. 2 for instructive purposes. Notice that the
.STORAGE statement is located between the key words .SNAP and .PROCESS along with the.FORMAT and
.HEADER statements; the executable statements are located in the algorithm section between the key words
.PROCESS and .END.
Skeleton No. 2 - Storage Details

SEND (TTY,#102)1TIME

FOR IPASS = 1TO12
IF IMASK (IPASS) AND !PATTERN NE 0 THEN GOTO 21
GOT010

LET IA LA RM= IPASS
SEND (TTY, #104)1ALARM

The following paragraphs serve as a summary and provide additional details of the .HEADER, .FORMAT, and
.STORAGE statements.

3-21

3.3.1 .HEADER Statement
The specification of messages or headings to be printed for tabular and other types of reported information is
accomplished by .HEADER statements. This type of statement, referenced by a SEND statement, consists of
two or more lines with the initial line having the form:
.HEADER

where t presents a mandatory reference tag. The following lines contain alphabetic and alphanumeric information
arranged exactly as they are to appear when printed.
The .HEADER statement is terminated by any line whose first nonblank character is a period.
The following example illustrates the use of .HEADER statements:
.HEADER

COMPUTER COOLING ANALYSIS
SAMPLE

CLOCK

BANK

TEMP

TIME

DEG

3.3.2 . FORMAT (Picture) Statement
This .FORMAT statement is used to specify message and/or the format in which designated output data is to be
printed. These statements are referenced by SEND statements for formatting purposes.
The picture .FORMAT statement consists of two lines, with the initial line having the form:
.FORMAT

Where tis a mandatory reference tag. The second line of the .FORMAT statement contains a message, if any,
and an actual representation of the data as it is to appear in its printed form. The representation will consist of:

a. Declaration of space characters (blanks). Spaces are regarded as meaningful elements in a
picture .FORMAT statement; they specify the actual spacing to be established between
nonblank format elements. Tabs are retained as TAB characters and are effective only on
Teletypes with TAB capability.
b. Declaration of number formats. Both real and integer numbers are represented by:
1)

A letter S for the sign (if used)

A letter X for each digit ( 15 digit maximum - INDAC maintains 6 digits
of accuracy), and

A decimal point which cannot precede the sign (letter S).

For example, the format for a 6-digit signed integer would be represented by:

sxxxxxx
A 6-digit real number with a 3-digit integer portion is represented by:

sxxx.xxx
c. Declaration of Text. Titles, comments, etc., may be specified for printout of a .FORMAT
statement by enclosing each text line segment within quote(") marks. These quote marks
are not printed; they identify and delimit text units.

3-22

The following are examples of text declarations:
"FURNACE"

"VALUE"

"T 104"

"NO."

Thus the information contained in the second line of a declared .FORMAT statement consists of number and
text declarations separated by spaces. For example:

#10

.FORMAT
"TEMP" L......JSXXX.XXL......J "DEGREES CENTIGRADE"
NOTE
In the examples given, the symbol L......J represents unit
spaces that would be entered into systems using the
INDAC keyboard space bar.

The .FORMAT statement may contain continuation lines that extend the number of characters to be printed on
a single line. For example:

.FORMAT
"SENSOR NO. 1"

"TEMP" SXXX.XX "SENSOR NO. 2 TEMP"

SXXX.XX

A format statement that specifies a group repeated in a serial manner throughout the line (for example, SXXXX
SXXXX SXXXX .... ) may be written using a shorthand technique. This technique consists of placing the format group to be repeated (including interposing space units) within parentheses and preceding the enclosed group
with an integer that specifies the number of times the group is to be repeated.
For example, the line:
SXXX.XX SXXX.XX SXXX.XX
may be expressed:
3 (SXXX.XX
The shorthand technique may be used to reduce statements which specify mixed repetitive and nonrepetitive
format groups. For example, the statement:
"TEMP" SXX SXX SXX "PRESSURE" XX.XX XX.XX XX.XX
may be written:
"TEMP" 3(SXX ) "PRESSURE" 3(XX.XX )

3.3.3 .FORMAT (Declarative) Statement
A declarative form of .FORMAT statement is available with expanded capability. The statement has the form:
#t
Where #tis a mandatory tag, the left parenthesis is the first non-blank character following the ".FORMAT" and
Sis a FORMAT element containing either an ASCII string, a data field specification, or a special FORMAT character. Blanks may be used for readability but will be ignored by the Compiler (except in text strings). The declarative format string may contain continuation lines, and repeat groups as in the picture type format. Repetition of
format groups is restricted as in picture type format to a single level of nesting. Declarative format statements are
terminated by the first right parenthesis not terminating a repeat group. A synopsis of the declarative format is in
the following table.

3-23

Table 3-4
Declarative .FORMAT Specifications
Control

Function

Sample

Sample Output

Output Specifications
Fw.d

Output a real or integer variable as a signed field, w characters
wide, including sign and period, with d decimal places.

F7.3

+99.999

Rw.d

Same as "Fw.d" except that field is unsigned.

R7.3

999.000

Output real data in exponential notation. Requires 15 spaces on
output.

+O. 123456E-0003

Output integer data as octal number

7654

Output n blanks, if n not specified one ( 1) assumed

Transmit n words, each as a twelve bit word - no conversions

Special Format Characters
II

As in picture format used to delimit an ASCII string to be generated.

Delimits FORMAT elements.

Generate a CR/LF in format.

Generate a CR in format. Does not generate LF.

Suppress CR/LF at termination of format. Must be followed by
terminating parenthesis of statement.

Suppress CR/LF at termination of format but leaves console in
"Operator" mode. Must be followed by terminating parenthesis
of statement.
NOTE
Integer data is limited to four digits (-2048 to +2047).
Real data is accurate to six digits; data output greater
than those limits should be viewed with these restrictions in mind. Data output greater than 15 digits will
generate a compiler error.

3.3.4 .STORAGE Statement
In many applications, I NDAC is required to accumulate and store large amounts of identifiable numeric data.
Acquired and computed data as well as constants and other items entered by the user must be stored in lists or
tables (arrays) that are defined by the user. In addition to defining arrays, the most important function of
.STORAGE is to supply common areas for data communication.
To define required arrays, the user must write a .STORAGE statement and follow it with a listing of symbolic
names unique to each required array. Once defined, an array name may be referenced by a statement (GET,
SEND, IF, LET) each time the data is to be entered into or read from that array.
3.3.4.1 Development of Array Names - Observe the following rules in developing an array symbolic name:

a. It must be unique within the program.
b. It may consist of an unlimited number of alphabetic or alphanumeric characters.

(continued on next page)

3-24

c. The first character must be alphabetic and must identify the type of data or constant (real
or integer) which the array is to store. The first character must be:

I for an integer array. (For example, IAMP, IH10, 1256, etc.)

Any letter but I for a real array. (For example, AMPS, 8526, ZZZZ, etc.)

3.3.4.2 Specification of Array Size - The use of a symbolic name in a .STORAGE statement reserves only one
storage element (one word for integers, three words for real numbers). When multielement storage is required,
the user must specify the size of the array by adding an enclosed range of elements having the form ( 1, N) to the
array's symbolic name. The subscript is interpreted as:
1 = The first element in the array
N = An integer representing the last element in the array ( N > 1)
For example, an integer array for the storage of 50 sampled voltage levels could be specified as:
ISVOLTAGE (1,50)
3.3.4.3 Specification of Array Elements - Individual elements of an array can be specified by using the symbolic
array name followed by an enclosed integer that represents the number of the element within the array. For example, given the designation of a 50-element array:
IABC (1,50)
The 30th word element is specified by the expression:
IABC (30)
The following examples illustrate the manner in which the .STORAGE statement is written:
Meaning

Statement
.STORAGE ISUM

Single-element array .ISUM

.STORAGE ISUM (1,20)

20-element array ISUM

.STORAGE ISUM (1,20),120,C,D

20-element integer array ISUM,
single-element integer array 120, and
single-element real arrays C and D

3.3.4.4 Specification of an Array Window - INDAC provides a unique technique that permits the user to specify
a subset of an array or a superset spanning a number of arrays.
Dimensioned arrays consist of a sequential group (list) of word locations within the processor memory; to specify
a particular subset of locations within an array, the programmer:

a. Lists the name of the array, (for example, IAMP)
b. Indicates the first element of the subset (for example, IAMP (5) or IAMP ( IJ))

c. Specifies the number of elements in the subset. For example, the statement:
IAMP (5) * 4
establishes a window 4 elements long (4 word locations for integer variable elements and 12
word locations for real variable elements) beginning with the fifth element of array IAMP
(that is, elements 5, 6, 7, and 8) of array IAMP. Note that the number of elements must
be an integer constant.

3-25

3.3.4.5 Array Spanning Windows - When arrays are defined in the same .STORAGE statement, they are established in core as a continuous series of word locations in the order listed. For example:
.STORAGE IAMP (1,6), ISTAT (1,10), ILOG (1,5)
establishes the list:
IAMP (1)
IAMP (6)

ISTAT ( 1)

ISTAT (10)

ILOG (1)

ILOG (5)

In situations such as described in the preceding example, the user may specify a window that can include portions
of two or more sequentially defined arrays by using the same technique described in Paragraph 3.3.4.4.
For example, the following statement:

IAMP (5) * 14
specifies a window that spans portions of arrays IAMP, ISTAT, and ILOG of the preceding example. The following array elements are included in the window:
IAMP (5)
IAMP (6)

ISTAT (1)

ISTAT (10)

ILOG (1)

ILOG (2)

The array window permits the user to specify a consecutive series of small arrays that simplifies gaining access to
data; moreover, it considerably reduces the code developed by the Compiler by having only one window bringing
in or sending out data to and from aii the defined arrays.
3.3.4.6 Presetting Stored Values - In INDAC, the user can preset the value of stored data whenever the array to
contain the data is defined. This feature of the.STORAGE statement is useful for establishing constants or initial
data values. The required statement formats and procedures for presetting value in .STORAGE statements are as
follows:

a. To preset a single defined item the format is:
.STORAGE NAME/value/
NOTE
a. The value must be preceded by a slash(/).
b. A value preset is terminated by a slash (/).
c. To specify an octal value, precede the number by
an apostrophe ('),e.g.
d. Commas are used to delimit each item specified.

3-26

a. (cont)

Examples:
.STORAGE ISUM/50/
.STORAGE ISUM/'0050/

b. To write a statement which presets a series of values into an array, the required format is:
.STORAGE NAME (ARRAY SIZE)/n 1 ,n 2 , ... ,n/
For example:
.STORAGE ISUM (1,4)/50,60,70,80/

NOTE

a. The number of values specified in the preset list
may not exceed the defined size of an array.
b. Any elements of an array which are not preset
are set to 0.
c. If the same value is to be preset into more than one array element, the statement has the
following format:
.STORAGE NAME (array size)/No. of Elements (Value)/
Example:
.STORAGE ISUM (1,20)/5 (50)/
This statement establishes the 20-element array ISUM and presets the first five elements to
the decimal value 50; the remaining 15 elements are set to 0.

3.4 DATA COLLECTION AND CONTROL
In automating any industrial process, process variables must be collected and control must be effected. Since
there are usually many sensors and controls involved in a process, multiplexing devices are used to interface the
computer with the process. It is through individual channels of these devices that process data is collected and
control is effected. DEC provides a full line of such multiplex devices to service the sensors and control equipment of industrial plants.
DEC multiplex devices (IDACS) include:
Digital-to-Analog Converters
Analog-to-Digital Converters
Digital Output
Digital Input
Integrating Digital Voltmeter
The following names have been assigned to the I NDAC System multiplex devices:
ADC (Analog-to-Digital Converter)
DAC (Digital-to-Analog Converter)
UDC (Digital 1/0)
AF04 (IDVM)
To illustrate the data collection and control capability of the I NDAC language, a simple example of a process
control application is detailed in the following paragraphs.

3-27

NOTE
Refer to Skeleton No. 3 to place the program statements of the example into perspective.
The example demonstrates maintaining the bath temperature of a solution, including the following operations:
a.

Open hot and cold water valves

b. Open drain and turn BATH OK lamp off.
c.

Measure water temperature. If water temperature is 68 ±0.5 degress, turn BATH OK lamp on.
If water temperature is not 68 ±0.5 degrees turn BATH OK lamp off and adjust hot water valve.

d. Repeat c.
Skeleton No. 3
.EQUIP~ENT

.
........
________________,

+1-1I

device
channel
control

.PHASE
.ACTION
DO SNAP #2 PRIORITY 1

.SNAP
.PROCESS
_ _ _ _ _ _ _ _ _ _ _____,GET /SE ND
.END
DATA COLLECTION AND CONTROL

An illustration of the process and sensor device and channel assignments is shown in Figure 3-1. Since the
process interface devices are multiplexed (that is, more than one channel can be accessed in each device) the
channels of interest must be specified in the program to send data to and to get data from the process elements.
The .EQUIPMENT statement permits the user to declare the channels of interest. The .EQUIPMENT statement
must be the first non-comment line in the program. A series of continuation lines (indicated by an up-arrow, t)
are used to declare the channels of interest for each multiplex device. Control options and any associated signalconditioning subroutines can also be declared in the .EQUIPMENT statement. Details relating to control option
and subroutine declaration are discussed later in this chapter. The following statement declares two DAC channels, IHOT and ICOLD:
.EQUIPMENT

*DAC
CHAN (1) IHOT
CHAN (2) ICOLD

Notice that devices are identified by a preceding asterisk and channel declarations follow the name of the associated device. Also notice that both channel declarations contain the key word CHAN, a number in parentheses,
and a name. The name declares both channels as integer variables. The I HOT variable is assigned channel one and
the ICO LO variable is assigned channel two. Notice that when a DAC channel of interest is declared it is just like
declaring the process control element itself. Once the above .EQUIPMENT declaration is made, the HOT and
CO LO valves can be positioned by writing:

3-28

LET I HOT = '2000
LET ICOLD = '2000
SEND (DAC) IHOT, ICOLD
Note that the . EOU I PM ENT statement does not declare the data to be sent to the device; but, only the channel
associated with the variable name. The name can then be used for assignment of data using a LET statement.
The octal number 2000 has been chosen to represent a value setting for a half-open valve.
IHOT

HOT
WATER
ID RAIN
COLD
WATER

FILM-PROCESSING
SOLUTION
BATH

~
L__Y_J

IBATHOK
TEMP

MUX DEVICE

CHANNEL

HOT valve

DAC

COLD valve

DAC

Temperature sensor

ADC

Drain valve

UDC

15*

BATH OK indicator

UDC

16*

PROCESS ELEMENT

*NOTE
Drain valve and BATH OK indicator are connected
to the screw terminals for UDC channels 15 and 16
as follows:
CHANNEL15

1---0_0
_ _....J}

CHANNEL 16

OB-0727

Figure 3-1 Sample Process Control Application
The next step in the sample program is to open the drain and turn off the BATH 0 K lamp. As defined earlier,
the drain valve and the BATH OK lamp are connected to the UDC. There are two methods of declaring UDC
channels: Explicit and Point methods. Only the explicit method is discussed at this time. The following
.EQUIPMENT statement declares the channels for the drain valve and the BATH OK lamp.

3-29

.EQUIPMENT

t
t
t

* UDCE
CHAN (15) IDRAIN
CHAN (16) IBATHOK

The device name UDC is modified with an E to denote explicit channel declaration. Both channels are declared
as integer variables. IDRAIN is assigned channel 15 and IBATHOK is assigned channel 16.
NOTE
Octal constants are generally used to set or reset
bits of interest in a particular channel. The bits of
interest and the associated constants for setting these
bits in this example are as follows:
CHAN

Bit

Constant

Drain Valve

BATH OK lamp

8
11

'0001

Process Element

'0010

When bit 8 of channel 15 is set the drain valve is open; when bit 11 of channel 16 is set the BATH OK lamp lights.
The following program statements cause the drain to open and the lamp to go out.
LET IDRAIN = '0010
LET IBATHOK = 0
SEND (UDCE) IDRAIN, IBATHOK
The LET statements assign octal 10 (bit 8 is set) to ID RAIN and 0 (bit 11 and all other bits are reset) to
IBATHOK. The SEND statement sends these two bit patterns to the UDC channels declared in the .EQUIPMENT
statement.
The next step in controlling the sample process is to measure the bath temperature, adjust the HOT water to
obtain a bath temperature of 68 ±0.5 degrees, and turn on the BATH OK lamp. As defined earlier, the temperature sensor is connected to channel 37 of the ADC. In addition to declaring the channel to which the sensor is
connected, ADC's with variable gain capability also require a control option declaration (see Paragraph 3.4.2).
The following .EQUIPMENT statement declares the temperature sensor and a control option
.EQUIPMENT

t
t

CHAN (37) TEMP

t#901

X200 DO TCONV

*ADC

The temperature sensor is declared on channel 37 as a real variable named TEMP. By declaring a real variable for
the temperature channel, temperature may be maintained in 0.5 degree increments. An integer variable would
restrict the temperatures to full degree increments. The control option declared is X200 with a tag of #901.
Control options must always be tagged so that the option can be referenced in the GET statement. A control
option declaration is required when the ADC of the system has scaling capabilities. Some DEC ADCs do not have
scaiing capabiiities (fixed gain) and therefore do not require a controi option deciaration. To get and test the
temperature and to control the water temperature and indicator lamp, the following statements are used.

3-30

GET (ADC, #901) TEMP
IF TEMP GR 68.5 THEN GOTO 12
IF TEMP LS 67.5 THEN GOTO 13
LET IBATHOK = '0001
GOTO 10

LET IHOT = IHOT - 1
GOTO 9

LET IHOT = IHOT +1
LET IBATHOK = '0000
SEND (DAC) IHOT

SEND (UDCE) IBATH OK
GOTO 11

Notice the structure of the GET statement for the temperature (TEMP); the device name and the control option
tag are separated by a comma and enclosed in parentheses. Since TEMP was declared to be associated with channel 37 in the. EOU IPM ENT statement, the GET statement transfers the quantity in channel 37 to the storage
location TEMP while the ADC is operating with a gain of 200 (X200). The two IF statements test the quantity
that was stored in TEMP by the GET statement. If the sensor supplies a value that must be converted or I inearized before a meaningful measurement can be made, the required conversion routine can be declared with the
control option (see Paragraph 3.4.2). In the example, a conversion routine is declared (DO TCONV). The routine is performed during the GET statement and must be coded to perform the following: convert the 12-bit
ADC value to a real variable, linearize the value and store it in the location named TEMP. If the temperature is
not within 0.5 of 68 degrees, the statement labeled 12 or the statement labeled 13, depending on the temperature,
will be executed. These statements increase or decrease the variable named I HOT. In either case, the statement
labeled 9 is executed next to set IBATHOK to 0, thereby turning the light out. The two SEND statements that
follow transmit the new position value to the HOT \lljater valve and turn off the power to the !BATHO K !amp.
Program control is then transferred to the statement labeled 11 to repeat the sequence. When the bath temperature reaches 68 ±0.5 degrees, then the statement following the IF statements is executed to set bit 11 of
IBATHOK to 1. The statement labeled 10 is then executed to turn on the BATH OK lamp and the sequence is
repeated.
All the statements related to data collection and control that were developed in the previous paragraphs have
been placed in perspective in the following update to Skeleton No. 3. Slight changes have been made to the program to illustrate how algorithm code can be minimized. Notice that the channel and option declarations for
each device appear on continuation lines following the . EOU I PM ENT statement. All executable statements are
located in the algorithm section between the key words .PROCESS and .END.
In the previous example, 2 UDC channels (words) were used; one for the drain valve and another for the BATH
OK lamp. Since each UDC channel contains 12 bits, the valve and lamp can both be connected to one channel,
in addition to other controls and indicators.
NOTE

A given UDC channel is limited to input or output
applications, not both.

3-31

Skeleton No. 3 - Equipment Details
.EQUIPMENT

t
t

CHAN (1) IHOT

CHAN (2) ICOLD

t
t
t
t
t
t#901

*UDCE

*DAC

CHAN (15) !DRAIN
CHAN (16) IBATHOK
*ADC
CHAN (37) TEMP
X200 DO TCONV
~PMA.li::···

...... ACtlO:N:;.,.,
•®. ~'
'.~~::.:.:.;;:
· .fRQ~t;~.
LET IHOT = '2000
LET ICOLD = '2000
LET IDRAIN = '0010

LET IBATHOK = 0
SEND (DAC) IHOT, ICOLD

SEND (UDCE) !DRAIN, IBATHOK
GET (ADC,#901) TEMP
IF TEMP GR 68.5 THEN GOTO 12
IF TEMP LS 67.5 THEN GOTO 13
LET IBATHOK = '0001
GOTO 10

LET IHOT = IHOT - 1
GOT09

LET IHOT = IHOT + 1
GOT09

When more than one control, indicator, or other field element is connected to one UDC channel, then Boolean
operations must be used to isolate the bit of interest to test, set, or reset the bit. The following example illustrates how Boolean operations can be used to operate on specific bits of a UDC word. Assume that the drain
valve and the BATH OK lamp are connected to one UDC channel as follows:

CHANNEL15
0

TO DRAIN
VALVE

TO BATH OK
LAMP
08-9728

3-32

With the valve and lamp connected to one channel as shown above the .EQUIPMENT statement should be
changed to reflect that only one UDC channel is used. The following .EQUIPMENT statement defines UDC
channel 15.
.EQUIPMENT

t
t

*UDCE
CHAN (15) IOUT

Notice that a different integer variable name is used in declaring the channel. The algorithm for the sample
process can now be rewritten as follows:
LET I HOT = '2000
LET ICOLD = '2000
LET IOUT = '4000
LET IOUT = IOUT AND NOT '4
SEND (DAC) IHOT, ICOLD

SEND (UDCE) IOUT
GET (ADC, #901) TEMP
IF TEMP GR 68.5 THEN GOTO 12
IF TEMP LS 67.5 THEN GOTO 13
LET IOUT = IOUT OR '4
GOT010

LET IHOT = IHOT -1
GOTO 9
LET IHOT = IHOT +1

GOT09
Notice that octal constant 4000 is assigned to IOUT. The LET statement labeled 9, using Boolean operations,
sets up the bit pattern required for opening the drain and turning off the lamp. The LET statement following
the IF statements, using another Boolean operation, sets up the bit pattern required to turn on the lamp. Otherwise, the algorithm is the same as the one presented earlier.

The reader was introduced to the following three process 1/0 devices in the preceding example:
DAC
ADC
UDC
All three devices require channel declarations in that they are all multiplex devices. Only the ADCs require control
option specifications. The following paragraphs provide additional details for each of the above devices and introduce the following two devices:
AF04
FILE

3-33

3.4.1

DAC (Digital-to-Analog· Conversion Devices)

3.4.1.1

Hardware Device - DEC offers two digital-to-analog conversion devices: the AA01 and the AA50. Both

devices are assigned the name DAC. GENDAC declares this name to the INDAC software system when the 1/0
handler of either device is selected from the G ENDAC library. Therefore, only one of these devices can be
declared in the I NDAC System.
3.4.1.2 Equipment Declaration Statement - The DAC and associated information is declared in the INDAC program as follows:

t
t

*DAC
CHAN (m,n) v

Where m, n are, respectively, the first and last channels associated with the integer array. If v is a single integer
variable, then only a single channef is declarec:J. There may be as many repetitions of the channel multiplexer line
as there are channels to declare. The DAC has no control options to be declared.
3.4.1.3 Language Statement - The INDAC language statement is:
SEND (DAC) (data list>
Where (data list> is the list of the items to be output. Note that every variable specified in the list must also appear
in the equipment definition to establish the correlation between the variable and the associated channel address.
3.4.2 ADC (Analog-to-Digital Conversion Devices)
3.4.2.1

Hardware Device - DEC offers five analog-to-digital converter devices: the AF01, AF02, AF03, AFC8

and AD01. All of these devices are assigned the name ADC. GENDAC declares this name to the INDAC software
system when the 1/0 handler of any one of the above devices is selected from the G END AC library. Therefore,
only one of these devices can be declared in the I NDAC System.
3.4.2.2 Equipment Declaration Statement - The ADC and associated information is declared in the INDAC program as follows:

t
t
t#t
t#t
t#t

*ADC
CHAN (m,n)v
DO s
Xn DO s
DO AREF

Where m, n are, respectively, the first and last channels associated with the integer or real array v. If v is a single
variable, then only a single channel is declared. There may be as many repetitions of the channel multiplexer line
as there are channels to declare. For the AF03, AFCB and AD01 the control line contains a tag reference #t and
a gain factor (Xn) where n may be:

3-34

AF01

AF02

AFC8

AD01A

x
x

x
x
x

x
x

x
x
x
x
x
x

1
2

AF03

(Fixed Gain)

10
20
50

x
x
x

100
200
1000

•

The control line may also contain a subroutine call DO s, wheres is a linearizing or signal conditioning subroutine.
There may be as many control lines of this type as there are variable gains or signal conditioning subroutines to
declare. In addition to the optional control lines, one line must contain the statement DO AREF. This subroutine establishes the location of the reference junction or offset variable required by some of the conversion routines. The AF01 and AF02 are fixed gain devices and do not require a gain specification on the control line, but
may require specification of the same subroutines described above.
Language Statement - The I NDAC language statements are:
GET (ADC) (data list)
GET (ADC, #t) (data list>
Where #tis a tag reference to the control line and (data list) is the list of the items (v) to be converted. Note that
every variable specified in the list must also appear in the equipment definition to establish the correlation between the variable and the associated channel.
If the signal conditioning subroutines created by the user are called in a control 1ine (DO s) and used in a GET
statement, then the first GET statement for the ADC executed in the job must define the variable used as the
reference junction or offset variable by referencing the "AR EF" subroutine call on the control line in the GET
statement. Then the GET statement for the data channel of interest can be executed with the reference to the
conversion routine call. For example:
.EQUIPMENT
t

*ADC

CHAN (0) IREF

CHAN ( 1,6) I LEV

t#10

X200 DO AREF

t#11

X100 DO CONV
GET (ADC, #10) I REF
GET (ADC, #11) ILEV

3-35

NOTE
The 1/0 handler for the ADC returns a 12-bit integer
count for the conversion. Normally the variable in
the data list is an integer variable. If a real variable
input is required, it is the responsibility of the subroutine called through the control line to convert
the integer variable to a real variable. If a channel
declaration is made using a real variable name, then
that channel may only be called by referencing a
control line with a subroutine declaration.
3.4.3 AF04 (Integrating Digital Voltmeter)
3.4.3.1

Hardware Device - DEC offers one Integrating Digital Volt Meter (IDVM) designated and named AF04.

G ENDAC declares this name to the I NDAC software system when the 1/0 handler of the I DVM is selected from
the GENDAC library.
3.4.3.2 Equipment Declaration Statement - The AF04 and associated information is declared in the I NDAC
program as follows:

*AF04

t
t#t
t#t

CHAN (m,n) v
a1 , a2 , a3

DO s

DO AREF

When m, n are, respectively, the first and last channels associated with the array v, the array must be a real array.
If v is a single variable, then only a single channel is declared. There may be as many repetitions of the channel
multiplexer line as there are channels to declare. The control line contains the item specifying the selectable
function, range, and resolution required of the AF04. One each from the following three lists may be selected
(a 1 , a2 , a3 )

a,
FUNCTION

a2
RANGE

a3
RESOLUTION

10MV

FREQ

100MV

.1
.01

PERIOD

1000MV

.001

OHMS

10V

100V
AUTO

Note that the selectable voltage ranges provide for increased conversion accuracy. The array or variable v is always supplied in volts (with the proper exponent) regardless of the selected range.
The control line may also contain a subroutine call DO s, wheres is a linearizing or signal conditioning subroutine.
There may be as many control lines of this type as there are variable gains or signal conditioning subroutines to
declare. In addition to the optional control lines, there must exist one line containing the statement DO AR EF.
This subroutine establishes the location of the reference junction or offset variable required by some of the conversion routines.

3-36

3.4.3.3 Language Statement - The IN DAC language statement is:
GET (AF04, #t) (data list>
where #tis a tag reference to the controi iine and (data list> is the list (real variables) of the items to be converted.
Note that every variable specified in the list must also appear in the equipment definition to establish the correlation between the variable and the associated channel.

If the signal conditioning subroutines created by the user are called in a control line (DO s) and used in a GET
statement, then the first GET statement for the AF04 executed in the job must define the variable used as the
reference junction or offset variable by referencing the" AR EF" subroutine control line in the GET statement.
For example:
.EQUIPMENT

*AF04

CHAN(O) REF

t
t
t

#10 DC, 100 MV, .01 DO AREF

CHAN ( 1,6) LEV
#11 DC, 1000M, .01 DO CONV
GET (AF04, #10) REF
GET (AF04, #11) LEV

NOTE
The 1/0 handler for the AF04 returns a real variable
for the conversion. Normally the data list contains
only real variables. If an integer output is required,
it is the responsibility of the subroutine called
through the control line to convert the real variable
into an integer variable name, then that channel may
only be called by referencing a control line with a
subroutine declaration.
3.4.4 UDC (Universal Digital Controller)
3.4.4.1

Hardware Device - DEC offers one highly flexible digital 1/0 device designated UDC8. Two 1/0

handlers have been developed for this device: an explicit handler called UDCE and a point table handler called
UDCP. G ENDAC declares these names to the I NDAC software system if the handlers are chosen from the
GENDAC library. Either or both handlers may be selected. The following descriptions cf the equipment declaration and language statements deal with the UDCE handler only. The UDCP handler is discussed later in this
chapter.
3.4.4.2 Equipment Definition Statement - The UDCE and associated information is declared in the INDAC
program as follows:

t
t

*UDCE
CHAN (m,n) v

3-37

where m, n are, respectively, the first and last channels associated with the variable v. If v is a single variable, then
only a single channel is declared. There may be as many such multiplexer lines as are required to define the input
and output channels.
3.4.4.3 Language Statement - The INDAC language statements are detailed below:

a. Each set of 12 input lines, connected to a digital input channel, is available to the system on
demand through the statement
GET (UDCE) (data list>
where (data list> is the list of items to be input. Note that every variable specified in the list
must also appear in the equipment declaration to establish the correlation between the variable
and the associated channel.
b. Each set of 12 output lines, connected to an output channel, may be exercised by the system
on demand, through the statement

SEND (UDCE) (data list)
where (data Iist) is the Iist of the items to be output. Note that every variable specified in the
list must also appear in the equipment definition to establish the correlation between the
variable and the associated channel.
3.4.5 FILE (Pseudo-Device)
3.4.5.1

Hardware Device - The FI LE is maintained on the system storage device, Mass Storage Disk D F32.

3.4.5.2 Equipment Definition Statement - The Fl LE and associated information is declared in the INDAC program as follows:

t
t#t

*FILE
k

where #tis the tag reference of the control line and k may be one of thefollowing three identifiers: 1, 2, or 3.
3.4.5.3 Language Statement - The INDAC language statements are:
GET (FILE, #t) la, (data list)
SE!\JD ( F ! LE, #t) !a, (data !ist)
where #tis a tag reference to the control line specifying the FI LE identification, la is an integer variable specifying the record (page) within the FI LE, and (data list) is the list of the items contained within the record.
NOTE
The total word count of the data-list items may not
exceed one page (128 decimal words). The total number of records available is dependent upon the size of
the user job and the number of disks in the user's system.
The totai fiie is organized as one sequential set of records from the base of FI LE 1, for as many. records as are
used by the program. FI LE 2 and 3 are simply convenient locations within the sequential set of records for reference purposes. There is no protection within the Fl LE, since the overlap feature has a useful function. The
starting location of each of the FI LEs may be altered via the Executive Command Decoder at run-time.

3-38

3.4.6 Equipment Statement Summary
The .EQUIPMENT statement must be the first non-comment line of any program that uses multiplex devices or
devices that require specification of control options. This statement uses a series of continuation lines to specify
each device, its channel declarations, control option, and any associated-conditioning subroutines.
The .EQUIPMENT statement has the following general form:
.EQUIPMENT

t
t
t

(device code line)
(multiplexer line)
(control line)

Each specified device (code line) must be followed immediately by applicable multiplexer lines, then applicable
control I ines.
The .EQUIPMENT statement is used to declare symbolic names of devices, device control options, and subroutines.
Standard device declarations (TTY, KEYS, STATUS, PRIORITY and FILE) are a permanent part of the compiler
tables. These tables can be modified by G ENDAC (Generate INDAC) to add 1/0 handlers for additional devices.
The description and format of the line types which make up the. EQUIPMENT statement are as follows:

a. Device Code Line
This continuation line contains a device symbolic name and has the form:

t*u
where u is the assigned symbolic name for the device to be specified. The asterisk (*) indicates
that all information presented until the detection of the next asterisk or(.) statement refers to
the specified device.
The symbolic name used must correspond exactly to that supplied to the system during initialization (refer to Appendix I).

b. Multiplexer Line
When the specified device has multiplexed input or output lines, the code line must be followed
by a multiplexer line which has the form:
tCHAN mux v
where
1.

CHAN identifies the type of information presented.

mux represents the multiplexed channel or channels of the device to be associated with v.

v represents a name of a variable array to which the specified channel or channels are
assigned and defines the data buffers for the 1/0 operations.

When more than one device channel is to be specified, the following form is used:
tCHAN (m, n) v
where m is the first channel and n is the last channel in the desired range. For example, the
list (0, 19) specifies a range of 20 channels identified as 0 through 19.
The array named is automatically dimensioned to the number of elements specified by m and
n. The array name is the data buffer where the digitized values of input signals connected to
multiplexer channels will be stored.
When the specified input or output channel or channels of a multiplexed device are to be
assigned to more than one variable or array, one multiplexer line is required for each different
assignment.

(continued on next page)

3-39

c. Control Line
One or more control lines may be used to list the symbolic names assigned to the device-control
options available for the specified device. If a signal-conditioning subroutine is required for the
device input or output operations, it is specified in the control line.

t #t

SCN 1 , SCN 2 , ... , SCNn

t #t

SCN 1 , SCN 2 , .•. , SCNn

or
DOS

where
1.

t is a mandatory reference tag.

SCN represents symbolic control names assigned to control options available for the specified
device.

s represents the assigned symbolic name of a conversion subroutine.

Control statements for each device must be listed immediately after the last multiplexer line for
that device.
Control lines must be tagged since they are referenced by input/output statements (GET, SEND)
to specify the manner in which the manipulated data is to be handled.
Control lines are used only to specify a particular control option or set of options required to
carry out the GET or SEND statement.
A separate, individually tagged control line is required for each specified set of options used.
3.5 STRUCTURE OF AN INDAC JOB
Usually in an on-line, real-time acquisition and control application, there are many tasks that must time-share core
memory for economy of core storage. To attain such a goal in INDAC 8/2, all programs (or program segments)
are disk resident; they are brought into core by the operating system in response to process interrupts, elapsed
timers, or task requests, see Figure 3-2. To allow this kind of time-sharing, an INDAC program is not made of a
string of statements such as a BASIC or FORTRAN program which is executed as a single program unit. Rather,
it is a collection of different tasks or program units separated by appropriate segmentation statements. Process
interrupt handling is covered later in this chapter.
3.5.1

Program Segmentation

a. Job Specification - This segment of the program includes the .EQUIPMENT, .INTERRUPT,
.STORAGE, .HEADER, and .FORMAT statements. These statements define the Global
parameters of the job. This information is transferred to core during the initial call for the
job and remains in core throughout the life of the job. This area is never swapped.

b. The PHASE - This segment contains the timer parameters for calling SNAPS or other
PHASEs. The PHASE segment exists in core only when the PHASE is operational. If a
PHASE is released and then recalled, that PHASE will be completely initialized on recall.

c. The SNAP - This segment (task) contains the actual algorithms used to perform the job.
SNAPs may also call other SNAPs or PHASEs as required to service the job.

d. The SUBROUTINE - This unit contains code for common operations to avoid duplication
of effort. When included in a job, the SUBROUTINE becomes an integral part of the job.
The INDAC Executive and the INDAC language permit a variety of sequencing and scheduling functions. Some of
the functions are programmed to occur at a specific interval of time or at the next available opportunity. Other
functions are dependent upon the operation of a peripheral device. Due to the random operation of such devices,
the Executive is prepared to resolve conflicts between the different requests to operate parts of the system.

3-40

Job
Device channel and Control Declaration, System Parameters (Globals)/messages,
Logging format, storage
PHASE
Activity

Schedule

SNAPS

Action
Local Parameters
Algorithm

PHASE
Activity

Schedule

SNAPS

Action
Local Parameters
Algorithm
Local Parameters

SUBROUTINE

Code
Return

.END
Figure 3-2 Structure of an INDAC Job
3.5.1.1 Job Specification Segment - This is the first segment in a program. It presents any parameters required
for systems-level program activities (Global). A detailed breakdown of the organization and contents of the Job
Specification segment is given in Table 3-5. The segment components introduced in the table are described in
detail in the following paragraphs.
Table 3-5
Job Specification Segment, Organization and Contents
Hierarchy of Possible

Statement Type and Use

Job Statements
.EQUIPMENT

Specifications of logical device units, channels, control modes, and linearization
subroutines associated with the particular device

.INTERRUPT

Specification of field interrupt device

.STORAGE

Specification of system-level data storage areas

.HEADER

Specification of messages, titles, or other types of information which is to be output (printed or displayed) on demand from other segments within the program

.FORMAT

Specifications of exact format(s) for the output of data common to the system

3-41

3.5.1.2 .PHASE Segment - The phase segments consist of:

a. The key word .PHASE preceeded by a tag.
b. A group of statements that specify the scheduling of the data-handling operations to be
performed in the PHASE.

c. One or more .SNAP program segments that contain the actual operational instructions for
carrying out the task assigned the PHASE segment.
The PHASE identifier has the following form:
#t

.PHASE

where #tis a mandatory reference tag.
The basic organization and content of a .PHASE segment are shown in Table 3-6.
Table 3-6
Organization of PHASE Segment
Hierarchy of Possible PHASE Statements

Statement Type and Use

#10

.PHASE

PHASE segment identifier

#14

DO SNAP #50 EVERY 5 SEC PRIORITY 2

Example of an activity statement

.ACTION

Compiler directive statement which introduces
a list of PHASE action and timer statements

TIMER (START, #14)

TIMER statement initializing previously defined activity statement

DO SNAP #55 PRIORITY 1

Example of a SNAP call statement
Identifies first SNAP to operate in PHASE

3.5.1.2.1

Activity Statements - The activity statements define the scheduling parameters of the actions listed

under the .ACTION statement. The key words EVERY, DELAY and AT, in conjunction with the DO statement,
are used in the activity statements. For example:
#10

.PHASE

#14

DO SNAP #500 EVERY 5 SEC PRIORITY 2

#15

DO SNAP #501DELAY10 MIN PRIORITY 3

#16

DO SNAP #502 AT 13: 15 PRIORITY 1

3.5.1.2.2 .ACTION Statement - .ACTION is a statement which is used in PHASE segments to introduce a list
of specified actions to take place in the PHASE. The .ACTION statement occurs only once in each PHASE.
The action list consists of executable DO and TIMER statements. The list is started by the .ACTION statement
and is terminated by the occurrence of the first .SNAP segment. For example:
.ACTION
DO SNAP #500 PR 10 R ITY 2
TIMER (START, #14)
#500

.SNAP

3-42

3.5.1.3 .SNAP Segments - The snap segments consist of:

a. The key word .SNAP preceeded by a tag.
b. SNAP specification which may include ,STORAGE, .HEADER, and .FORMAT statements.
These statements define the local parameters of the task.

c. A list of executable program statements which are introduced by a SNAP level .PROCESS
statement.
The SNAP identifier has the following form:

.SNAP

where t is a mandatory reference tag.
The SNAP specification statement is a list of statements written immediately after the SNAP identifier that
specifies SNAP-level (local) data storage and output header and format requirements. The organization and contents of a .SNAP are shown in Table 3-7.
A .PROCESS compiler directive is employed in each SNAP to introduce the executable statements that form the
operational portion of this type of program segment.
Table 3-7
Organization of a SNAP Segment
Hierarchy of Possible SNAP Statements

Statement Type and Use

.SNAP

SNAP segment identifier

.STORAGE

Statement for specification of SNAP-level
storage

#12

.HEADER

Statement for specification of SNAP-level
header outputs

#13

.FORMAT

Statement for specification of SNAP-level output data formats

.PROCESS

Compiler directive, identifying executable
portion of SNAP

#10

List of executable statements
EXIT or

Program control statements

EXIT THEN DO SNAP#_ PRIORITY - or
EXIT THEN DO PHASE#_ PRIORITY_
NOTE
Comment lines may appear anywhere in the INDAC
program to simplify readability. A comment line is
written as follows:
·L_I - · - - - - -

Notice that a space delimits the period and the first
word of the comment.

3-43

3.5. 1.4 Subroutine - Often certain operations are used repeatedly throughout the procedures specified in the
program segments. In order to avoid duplication of effort, such common operations may be specified as subroutines. Subroutine call statements are inserted into the program list at each point where the subroutine operation
is required. During the execution of the program, each call causes the processor to execute the specified subroutine before proceeding to the next listed statement.
There are two types of INDAC subroutines.

a. Internal
b. External
The manner in which each is written and used is described in the following paragraphs.
3.5. 1.4. 1 Internal Subroutine - Subroutines of this type are written as part of a standard .SNAP segment. The
repeatable group of statements that may be referenced (called) as a subroutine unit is defined by:

a. A labeled executable statement used as the first statement in the subroutine group.
b. The use of a program control RETURN statement.
The call statement for an internal subroutine has the form:
DO k
where k represents the label of the first subroutine statement.
For example, the internal subroutine:
LET N = N + 1

401

RETURN
would be called by the statement:
DO 401
Internal subroutines are defined only for the .PROCESS segment in which they appear. More than one RETURN
statement may be used within a subroutine to facilitate alternative operations.
3.5. 1.4.2 External Subroutine - Subroutines of this type may be included as part of a complete job. When included in a job, external subroutines may be called only from the segments of that job. When entered into
INDAC separately, an external subroutine may be called from within any INDAC system program. External subroutine identifiers have the form:
.SUBROUTINE s (v1 , v2 , ... , vn)
wheres is a unique symbolic name assigned to the subroutine, and the parenthetical list contains dummy variables
for input and output data.
The identifier dummy variables represent those used throughout the subroutine only; their equivalent variables
in the program segments that call the subroutine must be specified in the call statement. The list v1 , .•. vn may
contain variable or array names. If a variable is an array name, it represents the base of the array.

a. Calling the External Subroutine - Call statements for external subroutines have the following
form:

(continued on next page)

3-44

a. (cont)

wheres represents the name of the called subroutine, and the call list contains a one-for-one
equivalent variable for each dummy variable in the subroutine identifier. For example, given
the subroutine identifier:
.SUBROUTINE TCON (X, Y, Z)
one example of an acceptable call would be:
DO TCON (A, B, C)
An example of a subroutine call is given in a listing shown in Appendix B. Note the way in
which arrays I K and A are made available to the subroutine.

b. Returning Control to the Caller - External subroutines return control when the following
statement is encountered:
RETURN s
where the RETURN statement specifies that the processor return to and execute the statement
immediately following the subroutine call. The external subroutine is organized in a manner
similar to that of a SNAP segment, that is,
.SUBROUTINE s (v1 , v2 , ••• , vn)
.STORAGE
.PROCESS
(List of Executable Statements)
RETURN s
As indicated above, the only nonexecutable statement which may be used in the body of
the external subroutine are.STORAGE and .PROCESS. Executable 1/0 statements GET
and SEND, plus TIMER and activity statements are not permitted.
NOTE
There may be more than one RETURN s statement
in the external subroutine.
3.5.1.4.3 Implicit Subroutine - Implicit subroutine calls are defined in the device control of the .EQUIPMENT
definition statement, such as:
.EQUIPMENT

t
t

CHAN (0, 10) VALUE

t#10

X200 DO IRONC

*ADC

Whenever a GET (ADC, #10) VALUE statement accesses the device ADC, the converted value from ADC is given
as input to the implicit subroutine I RONC; the output value of the subroutine is then deposited in VALUE.
Three parameters are implicit in this subroutine call. The first is the raw value read from the device; the second is
the output value of the routine, and the third is an argument to supply a reference or offset value for conversions
or linearizations.
The following is an example of a temperature scanning program using iron/constantan thermocouples in channels
1 through 10. The RTD of the isothermal reference box is connected to channel 0.

3-45

.EQUIPMENT

t
t
t
t #10
t #30

*ADC
CHAN (0) REF
CHAN (1, 10) TEMPERATURE
X200 DO IRONC
DO AREF
.STORAGE REF

.PHASE
.ACTION

#2
#3

DO SNAP #200 EVERY 5 MIN PRIORITY 1
DO SNAP #300 EVERY 40 SEC PRIORITY 1
TIMER (START, #2)
TIMER (START, #3)
DO SNAP #100 PRIORITY 1

#100

.SNAP
.PROCESS
GET (ADC, #30) REF
LET REF= 10.94 *(REF+ 57.5)
EXIT

#200

.SNAP
.PROCESS
GET(ADC)REF
EXIT

#300

.SNAP
.PROCESS
GET(ADC,#10)TEMPERATURE
EXIT
.SUBROUTINE IRONC (IVAL, CONV, REFDUM)
.PROCESS
LET COUNT 1 = IVAL
LET COUNT 2 = K1 * REFDUM + K2
LET CONV =COUNT 1 +COUNT 2
LET CONV =CO+ (C1 * CONV + (C2 * CONV + (C3 * CONV)))
RETURNIRONC
.END

The implicit subroutine AREF called at initialization establishes the location of the reference variable required.
The variable "REF" contains the reference value, in this case, reference junction temperature.
3.5.2 Scheduling Capabilities
The activity and action sections of a .PHASE segment define the initial scheduling parameters of a job. One or
more phase segments may be used depending on the size of the job or the scheduling requirements of the job.
The scheduling statements that may be included in the phase segment are:

3-46

.PHASE

#_
#_
#_
#_
#_

DO SNAP#_ EVERY

PRIORITY

DO SNAP#_ DELAY ___ PRIORITY_
DO SNAP#_ AT

PRIORITY

DO PHASE#_ DELAY ___ PRIORITY_
DO PHASE#_ AT __ PRIORITY_
.ACTION
TIMER (START,#_)
DO SNAP#_ PRIORITY

Note that a DO PHASE#_ EVERY ___ PRIOR ITV_ is not permitted. This statement is not permitted because the scheduling parameters of the calling phase are overlayed at call time.
Scheduling requests may also be specified in the .SNAP segments intermingled with executable code. The
scheduling requests that may be included in a .SNAP segment are:
.SNAP

#10

DO SNAP#_ PRIOR ITV
TIMER (STOP,#_)
DO PHASE#_ PR 10 R ITV
TIMER (START,#_)
EXIT THEN DO SNAP#_ PRIORITY
EXIT THEN DO PHASE#_ PRIORITY
EXIT
When a SNAP terminates (exits) with an EXIT THEN DO_ request, the request is stacked on the scheduling
queue for execution at the priority specified. This type of request is placed in front of all other requests at that
priority level. A SNAP terminated with a simple EXIT statement releases control to the INDAC Executive.

3.5.2.1 Activity Statements - Activity statements operate in conjunction with TIMER statements to establish
incremental control timers that schedule the performance of specified SNAP and PHASE program segments.
These statements can appear only in the activity section of a phase.
There are three activity statements:

a. EVERY
b. DELAY

c. AT
All activity statements must be listed in the PHASE before a .ACTION statement.

3.5.2.1.1 EVERY Activity Statements - EVERY activity statements are used in the formation of incremental
timers to control the repetitive execution of program SNAP segments. EVERY statements must be assigned a
tag reference (that is, #xxx) and can be referenced only by TIMER statements located in the ACTION or in the
SNAP program segments.

3-47

The EVERY statement has the form:
DO SNAP #t 2 EVERY c PRIORITY k
where

a. t 1 is a mandatory tag reference
b. t 2 is the tag reference assigned to the SNAP to be executed

c. c is a time unit declaration which has the form
1.

i SEC

i MIN

i HR

where i in an integer variable or integer constant

d. k is the assigned priority level 1 through 11
For example:
DO SNAP #300 EVERY 5 MIN PRIORITY 1

#200

This execution of this statement causes the SNAP tagged #300 to be executed at five-minute intervals.

3.5.2.1.2

DE~AY Activity Statements -

DELAY activity statements are used in the formation of elapsed time-

control statements that specify the time at which designated SNAPs or PHASEs are to be performed.
DE LAY activity statements must be assigned a tag reference. Th is type of statement has the form:
DO SNAP #t 2 DELAY c PRIORITY k
or
DO PHASE #t2 DELAY c PRIORITY k
where:

a. t 1 is a mandatory tag reference
b. t 2 is the tag reference assigned to the SNAP or PHASE to be executed

c. c is a time unit declaration which specifies a time interval of delay after which the designated
program segment is to be executed. This declaration unit may have the form:

i SEC

i MIN

i HR

where i represents an integer variable or integer constant
d. k is the assigned priority level 1 through 11

For example, the execution of:
#101

DO SNAP #300 DELAY 10 MIN PRIORITY 1

causes the SNAP tagged #300 to be scheduled for operation 10 minutes after the TIMER (START, #101) statement is executed.

3-48

DELAY statements are single-shot in operation; they specify an action to be carried out after one specified
interim of elapsed time.
3.5.2.1.3 AT Activity Statements - AT activity statements are used to specify the time-of-day program segment
execution.
The AT statement must be assigned a tag reference. This statement has the form:
DO SNAP #t. 2 AT HH:MM PRIORITY k
where HH is the hour (0-23) and MM is minutes (0-59) of the day. For example:
#101

DO SNAP #1AT18:30 PRIORITY 1 or

#102

DO SNAP #2 AT 01 :09 PRIORITY 1

will, when system clock has been set, execute SNAP #1 at 6:30 p.m. or SNAP #2 at 1 :09 a.m. This statement is
re-initialized every time it is executed and will take place at the same time every day unless the timer is stopped
with a TIMER (STOP,#_) statement.
3.5.2.2 Action Statements - The action statements are executable DO and TIMER statements. These statements may appear in the .ACTION section of a PHASE or the PROCESS section of a SNAP.
DO action statements are used for executing program segments without regard to system incremental timers. DO
action statements have the form:
DO SNAP #t.1 PRIORITY k
where t 1 is a SNAP tag in this PHASE segment and k is the priority level 1 through 11.
DO PHASE #t2 PRIORITY k
where t 2 is a tag of the current or another PHASE.
3.5.2.2. 1 Timer Action Statements - TIMER statements control the operation of the EVERY, DE LAY, and AT
activity statements. This statement has the forms:

a. TIMER (STOP, #t)
Where tis the reference tag of an EVERY, DELAY, or AT statement. Execution of this
statement halts the timing actions associated with the designated activity statement.
b. TIMER (START, #t)

Where t is the reference tag of an EVERY, DE LAY, or AT statement. Execution of this
statement starts the timing action associated with the designated activity statement.
3.5.2.3 EXIT Program Control Statement - This program control statement is used to terminate the program
.SNAP segments; its execution halts the operation of the SNAP being processed, and either returns control to the
Executive or specifies a new program segment to be executed.
The EXIT statement has two forms:

a. As a stand-alone statement
EXIT
which, when executed, terminates the current SNAP, or

3-49

b. As a combined statement
EXIT THEN DO SNAP #t PRIORITY k
or
EXIT THEN DO PHASE #t PRIORITY k
which, when executed, terminates the current SNAP operation, then directs the processor
to the next SNAP or PHASE to be executed.
3.5.2.4 Resolving Timer Requests - Timed requests are resolved using the single system incremental counter and
the process queue. As each second is recorded by the hardware clock, the system clock is incremented by one.
Each phase contains the individual timer parameters established by the user under the activity list of the phase.
A timer parameter is established by one of the following five activity statements:
DO SNAP #t. 2 EVERY c PRIORITY k
DO SNAP #t 2 DELAY c PRIORITY k
DO PHASE #t 2 DELAY c PRIORITY k
DO SNAP #t 2 AT HH:MM PRIORITY k
DO PHASE #t 2 AT HH:MM PRIORITY k
These language statements generate timer parameters examined by the timer scan operation. The parameters
are initially inactive and remain inactive until the statement: TIMER (START, #t.) in the .ACTION section of the
PHASE or the PROCESS section of the SNAP is executed. At this point, the timer is considered active in the
scan. Correspondingly, an active timer may be disabled by executing the statement: TIMER (STOP, #t) in the
.PROCESS section of a SNAP.
A phase may contain more than one active timer at any point. The timer scan resolves any conflict by selecting
the timer with the highest priority and time-due. If there is no single, highest time-due and the priorities are the
same, the scanner resolves the conflict by selecting the timers in the order in which they appear in the .ACTION
list. A DO or an EXIT TH EN DO statement from a SNAP is resolved by the process queue. The highest priority
SNAP or PHASE is executed first. If a new PHASE is initialized all subsequent SNAPs in the calling PHASE are
terminated.
An arbitrary division is made in the assignable priority levels: priority 0 (Interrupt) and priorities 1 through 7 are
considered "foreground" levels that must be run to completion, once initiated. This guarantees completely dependable sequential operation. Each SNAP within each priority level must run to completion before the next
priority may operate. Priority 0 is discussed later in this chapter. Priority levels 8 through 11 are considered
interruptable levels that may be suspended. Each SNAP called to operate at these levels will be stacked in the
process queue and called in priority sequence. When a background SNAP begins operation, no other background
priority SNAPS, higher or lower, can intervene. The operating background SNAP can be suspended, however, if
a foreground SNAP is called.
3.6 THE RUN-TIME SYSTEM/MAKING THE PIECES WORK
The I NDAC 8/2 Executive executes the compiled program. Services provided to the user by the I NDAC Executive include:
a.

Time and Priority Scheduling of SNAPs

b. Dynamic core management and swapping
c. Dynamic 1/0 buffers

d. Operator communications (Command Decoder)

3-50

3.6.1

Time and Priority Scheduling

The INDAC Executive and the INDAC language permit a variety of sequencing and scheduling functions. Some
of the functions are dependent upon the operation of a peripheral device; other functions are programmed to
occur at specific times, intervais of time, or the next available opportunity. Due to the random operation of
such devices as the operator's console, or the field interrupt device, the Executive must be prepared to resolve
conflicts between the different requests to operate parts of the system.
The Executive resolves requests to operate by building a process stack. Following the successful operation of any
member of the stack, the executive returns to the stack and executes the next scheduled process call. All process
calls scheduled in the stack have an associated priority level ranging from 0 (highest) to 11 (lowest). If any priority level is inactive, the Executive proceeds to the next lower level. If no level is active, the Executive is considered idle and returns to the highest priority level to begin again.
3.6.1.1

Division of the Process Stack - The process stack is considered to have three primary scheduling divisions:

a. Interrupt Priority (0) - This level is available only to the Executive Command Decoder and
the Interrupt Device Handlers.

b. Foreground Priority (1-7) - These levels are used for SNAPs that cannot be suspended in
their operation.

c. Background Priority (8-11) - A SNAP running at this priority will be suspended in its operation if a Foreground or Interrupt call is scheduled for operation. Only one such background
SNAP may be held in suspension at any moment of system operation.
3.6.1.1.1

Executive Command Decoder - The Executive Command Decoder may be called from the command

Teletype. Once called, the following job requests can be made:

a. Display and/or modify system parameters
b. Stop or suspend operation of job

c. Activate a specific PHASE of the job
Operator communication through the Teletype is usually a time-consuming typing process compared with the
speed of the computer. To avoid lengthy system tie-ups, the operator input is buffered and the Command Decoder is not activated until the entire request is completely entered. The Command Decoder is scheduled at
priority level 0 when a completed process request is received.
3.6.1.1.2

Field Interrupt Processing - Field interrupt devices (the 1/0 handlers) are declared by the INDAC

language statement:
.INTERRUPT (device names)
This statement must appear with the job specification statements following the. EOU IPMENT statement. The
statement must declare all device handlers expected to call the INTERRUPT SNAP. For example:
.INTERRUPT (UDCP)
The name UDCP identifies the point-table 1/0 handler for the UDC. This name is declared to the INDAC software system by GENDAC if the point-table handler is selected. Programming details employing the UDCP
handler are covered in Paragraph 3.7.
The actual INTERRUPT SNAP that will service the interrupt is declared by the INDAC language statement:
DO SNAP #t PRIORITY INTERRUPT

3-51

This statement must appear as part of the PHASE scheduling parameters immediately following the .ACTION
statement of a phase. Only one PRIORITY INTERRUPT SNAP call is permitted within a phase. For example:
.ACTION
DO SNAP #10 PRIORITY INTERRUPT
This statement declares the INTERRUPT SNAP to the interrupt device handlers. The INTERRUPT SNAP request will be placed on the process stack at priority level 0 (first-in, first-out) whenever called. Since only one
PRIORITY INTERRUPT SNAP call is permitted, the INTERRUPT SNAP algorithm must first determine which
device interrupted if more than one interrupt device is declared by the . INTERRUPT statement. The pseudodevice STATUS will contain the index to the device specified in the INTERRUPT statement when a field interrupt occurs. Therefore, the following statements may be used to segment the SNAP for handling more than one
interrupt device:
GET (STATUS) IDEX
GOTO ( 1,2,3), IDEX
Code for
device 1
EXIT
Code for

device 2
EXIT
Code for

device 3
EXIT

The INTERRUPT SNAP is activated by a hardware interrupt-level 1/0 handler. Once activated, the SNAP will
run until it releases control through an EXIT or EXIT THEN DO statement. The INTERRUPT SNAP itself cannot be interrupted.
3.6.1.1.3 Foiegrnund PiOcessing - SNAPs assigned priority !eve!s 1- 7 are considered foreground segments.
Once a foreground priority SNAP is activated, system control is retained until that SNAP executes one of the
following statements:

a. EXIT THEN DO SNAP #t PRIORITY k
b. EXIT THEN DO PHASE #t PRIORITY k
c. EXIT
Statements a. and b. above allow the SNAP to release control, but they establish a process request at the stated
priority level. Statement c. releases control unconditionally. In all three cases, the Executive interprets the
EXIT key word, terminates the SNAP, and returns to the top of the process queue for the next request. Statement b. schedules an action that will unconditionally release the operating phase, eliminate all associated timers
of that phase, and establish a new phase for operation. A foreground SNAP may also execute the following
statement calling a background priority SNAP.
DO SNAP #t PRIORITY k
In this case, the background priority SNAP is scheduled and the current SNAP proceeds with the next language
statement.

3-52

3.6.1.1.4 Background Processing - SNAPs assigned priority levels 8-11 are considered background segments.
Any background SNAP may be interrupted in its operation at the following points:

a. While the SNAP is processing a source language statement.
b. Any bid for the teletype core buffer that cannot be serviced {a!! bids are broken dO\'Vn to no
more than 20 characters).

c. An execution of an EXIT statement.
A background priority SNAP may also execute the following statement calling a foreground priority SNAP:
DO SNAP#t PRIORITY k
In this case, the foreground SNAP is scheduled and the background SNAP is suspended. The Executive then returns to the process queue to activate the scheduled SNAP.
3.6.2 Dynamic Core Management and Swapping
3.6.2.1 Job Specification Segment - This segment of the program is the .EQUIPMENT, .INTERRUPT,
.STORAGE, .HEADER, and .FORMAT information that the user defined at system level in his job. This information is transferred to core during the initial call for the job and remains in core throughout the life of the job.
This area is never swapped.
3.6.2.2 The PHASE Segment - This segment contains the scheduling of the SNAPS that may be run under that
PHASE. The PHASE segment exists in core only when that .PHASE is operational. If a PHASE is released and
then recalled, that PHASE will be completely initialized on recall.
3.6.2.3 The SNAP Segment - The base core location of all of these units, and the address referencing within each
unit to the other unit, are all precalculated during compile time. The relative disk addresses of each of the units
are all converted to absolute addresses by SPUT - the System Put-Together Program - before the Executive is
activated. Thus, all calls during the operation of the Executive are absolute; there is no time lost in table look-up,
relocating, or link loading.
3.6.2.4 External (Disk-Resident) Subroutines and Functions - The INDAC 8/2 System allows the user to develop external subroutines and functions that may be called at run time. The system also allows the user to call
mathematical functions supplied with the library. Calls for these routines direct the Executive to transfer the
requested code from the disk to allocated call pages, in core. The Executive maintains two subroutine call areas:
one reserved for interrupt or foreground priority SNAPs and the other reserved for background priority SNAPs.
Because there is only one reserved area for each priority group, subroutine nesting is not permitted (that is, one
external subroutine may not call another external subroutine).
3.6.2.5 Switching Priority Levels - The priority level at which a SNAP operates is determined by the process
call scheduling the SNAP, rather than by some pre-assigned value associated with the actual SNAP. It is possible,
therefore, to schedule a single SNAP within a PHASE to operate at different levels of priority. The Executive is
structured to take advantage of this as follows:

a. Common routines are not re-entrant but, maintain linkage parameters in page zero. This
reduces the overhead time and core of re-entrancy (significant on the PDP-8), yet allows
the linkage parameters to be saved on demand.
b. A save area is reserved in core for the linkage parameters. This save area is used whea a
SNAP operating at a background priority level is suspended.
I, rnntin11i::>rl
n::tnQ\
__ ....... , ___ nn
- · · nQvt
•• _,, .. .....
-:1-1

3-53

c. A swap area is reserved on the disk. This area is used when a suspended SNAP must be moved
out of core.
The suspension of a SNAP, therefore, takes place in two stages. Initially, the Executive reacts to the suspend
command by transferring the linkage parameters for the background,ipriority to a reserved core area. If the next
scheduled request is to opt:;rate that same SNAP at a foreground priority level, then the SNAP is directly executed
and no disk transfer is required. If the next scheduled request is to operate a new SNAP, then the suspended
SNAP is swapped to the disk to make room for the new request.
With this system, it is practical to develop a program of one SNAP, run that SNAP at different priority levels,
and never access the disk again after the startup procedures.
3.6.2.6 Executive to SNAP Communication - Since SNAPs may operate at different priority levels, the Executive must provide some technique of communicating to the SNAP the reason why it was scheduled. Communication between the Executive and source-level code is provided generally by the STATUS item; a one word,
integer item available through the pseudo-device command: GET (STATUS) la. When a SNAP is called into
execution by the Executive, the STATUS item is set to one of three values, depending upon the original function
that scheduled the SNAP to operate.

a. SNAP scheduled by a timer - STATUS contains the index value of the timer in the activity
list (for example, the third timer statement would have index value 3).

b. SNAP scheduled by a device interrupt handler - STATUS contains the index value of the
interrupting device in the .INTERRUPT (. .. ) statement; if only one device is declared in the
statement, then it is not necessary to test for the interrupting device.

c. SNAP scheduled by a DO SNAP ... statement - STATUS contains a 0.
To differentiate which value of a. orb. STATUS contains (only required for SNAPS that run at both interrupt
and foreground priority), the pseudo-call GET (PRIORITY) la will set la to the current priority level of the SNAP
running.
3.6.3 Dynamic 1/0 Buffers
The Executive provides a number of 1/0 handlers to service the different devices supported by INDAC 8/2. Those

handlers that require core buffers (F ! LE, TTY, etc.) request a!!ocated core from the Executive. Such allocations
are on a dynamic basis by job; thus, for each job only the core required to support the 1/0 handlers for that job
is allocated. This is opposed to many systems where the user defines the total set of 1/0 devices supported by the
system and must relinquish the core whether the specific job uses all those devices or not. The BK Executive allocates buffers in one page segments. The total number of words allocated, subtracted from the highest location
allowed (7000) will yield the top limit of User Area. This information, combined with the top limit of SNAPs
supplied by the Compiler, is sufficient to calculate the core requirements of the job.
3.6.4 Operator Communication
INDAC 8/2 applications vary from almost completely automatic operations to highly interactive, almost manual
control of systems and experiments. A wide variety of operator communication is available, ranging from the
Teletype console (using the Command Decoder supplied with the Executive) to more sophisticated operator consoles supplied by the user and interfaced to the digital 1/0 subsystems.
The standard INDAC 8/2 System is supplied with a Model 33 (optionally 35 or 37) Teletype. This Teletype is
the Command Console and is required to initiate the different jobs. The Executive's Command Decoder recognizes this console and provides the user with a means of inspecting and modifying parameters during the execution

3-54

of a job. The Executive can be configured to support additional Teletypes and may have any or all Teletypes in
Command mode. The command instructions recognized by the Executive are covered in Chapter 6, Executing
the Program. The essential information is that the Command Decoder can access only permanently core-resident
parameters. Such parameters are defined as system-level storage items in each job. It is possible to access a
greater amount of data, but, this involves code which the user must create within his own job.
3.7 SERVICING FIELD INTERRUPTS
The contact interrupt modules of the UDC provide the necessary hardware interface for field devices capable of
issuing interrupts. The point-table 1/0 handler (UDCP) provides a convenient method for obtaining the interrupt
information. In addition, the UDCP handler allows the user to access his UDC by channel selection rather than
by scanning a predefined set of channels.
A consistent method of accessing the UDC is held throughout the handler by using a point-table to define the
channels required. The UDCE handler requires that the user define to the Compiler the explicit channels that
will be used during his running program. The Compiler then allows a data item to be associated with the defined
channels. While this has the advantage of coding simplicity, there is no provision for altering the number of channels to be scanned, or the channels that an operator or program would like to select at any given moment. The
UDCP handler allows the user to develop a "point-table" defining the channels (or points) to be selected for a
given scan. Any point-table may be associated with any data item (or array). This allows the input or output of
information without pre-defining channels and data items for the Compiler.
3.7.1

Equipment Declaration Statement

The UDCP and associated information is declared in the INDAC program as follows:

*UDCP

t#1
t#2
t#3

IDENTIFY

INITIALIZE
TRANSFER

where the equipment control lines define the requests that may be made on the point-table 1/0 handler to supply
information.
3. 7 .2

Language Statements

The I NDAC language statements are detailed in the following paragraphs.

3.7.2.1

The Standard Call - The standard calls for accessing digital input and output functional modules are:
GET (UDCP) (v1 , v2 )
SEND (UDCP) (v1 , v2 )

where v1 and v2 are the two components of the data list. The first component is the point-table. This item may
be a simple variable, an indexed variable, or an array defining the channels to be accessed. The second component
specifies the data items that either contain the information to be output, or will receive the information to be
input. Generally, the point-table and the data item contain the same number of elements; however, this is not
required. The UDCP handler will sequentially process through both components until one component is exhausted.
The first component to be exhausted terminates the 1/0 operation.

3-55

NOTE
The control lines declared in the .EQUIPMENT
statement are not referenced in the standard call.
Reference to these control options are required only
to service contact interrupt modules.
3.7.2.2 INITIALIZE Request -The INITIALIZE request is required only for those modules that require the
Executive to buffer information, that is, information relating to the previous status of the device. The initialize
statement has the following form:
SEND (UDCP, #t) (data list>
where #tis a tag reference to the device control line specifying the INITIALIZE command, and the (data list> is
constructed as follows:

a. The first item of the data list must be a system-level storage array containing two (2) words
of storage for every channel of Contact Interrupt (Generic Code 2).
NOTE
This array must be provided for the exclusive use of
the Executive.

b. The second item of the data list, and all succeeding items, are point-table declarations. Each
item is a simple variable or an array containing the channel numbers of those interrupting
modules that require special buffering by the Executive (like the Contact Interrupt Module).
Each point-table must be a system-level storage item. The Executive will reference these
tables whenever interrupts occur, and the tables must be core-resident. Each array declared
must contain channel numbers of the same generic type modules. The first channel declared
in each array will be sampled for generic type by the Executive; the remaining channels of
that array will be assumed to have the same generic code.
3.7.2.3 IDENTIFY Request-The IDENTIFY request is required for all modules that create a "PRIORITY
INTERRUPT" in the job (that is, call the INTERRUPT SNAP). The identify statement has the following form:
GET (UDCP, #t) Iv
where #tis a tag reference to the device controi iine specifying the iDENTlFY command, and Iv is an integer
variable. The command will return the generic code of the module forcing the system interrupt in the item Iv.
This statement is normally followed by a computed GOTO statement using the item Iv. By this technique, the
user can branch to the correct access statement for the specific generic code of the module.
A typical coding scheme for servicing a UDC with several channels of interrupt modules is to create a loop using
the IDENTIFY request and the computed GOTO to control the loop. The GET statement is executed repetitively
until a 0 is returned in the integer item. The following GOTO statement is then executed with a 0 value, causing
a default condition. For example:

3-56

t
t#101
900

*UDCP
IDENTIFY

GET (UDCP,#101) !GEN
GOTO (901, 902, 903, xxx), IGEN
(default condition - no more interrupts outstanding)

901

(service UDC device error)

902

(service Contact Interrupt)

903

(temp, reserved)

xxx

(etc.)
GOTO 900

3.7.2.4 TRANSFER Request -The TRANSFER request is used in conjunction with the IDENTIFY request.
TRANSFER and IDENTIFY are a matched set and must occur in pairs or the Executive will generate an error
condition. Any GET or SEND statement separating an IDENTIFY and TRANSFER pair is prohibited. The
transfer statement has the following form:
GET (UDCP, #t) (data list)
where #tis a tag reference to the device control line specifying the TRANSFER command, and the (data list> is
formatted according to the specific generic type module being accessed.
In all cases where the TRANSFER command is used, the Executive will allow only one channel of data to be
accessed, regardless of whether that channel is directly scanned from the UDC or buffered in core by an interruptlevel 1/0 handler.
The following UDC access request is available:

a. For Contact Interrupt Module (Generic Code 2)
GET (UDCP,#t) IP 1, lv1, lv2• lv3
IP 1 is the name of the point-table containing all the channel numbers that hold a Contact
Interrupt board already specified under INITIALIZE.
lv 1 will be supplied by the Executive with an index into the point-table IP 1 , of the interrupting channel. If it is required, the statement:
LET I CHAN = IP 1 ( lv 1 )
will place the interrupting channel number in ICHAN.
lv 2 is a logical item (may be array element) that will be supplied with a 12-bit identification
of the interrupting channel status. The 12-bit word will contain a "1" bit for every relay of
the Contact Interrupt board that has changed state (COS), in the direction wired by the user,
since the last time that the board was sampled.
lv 3 is a logical item (may be an array element) that will be supplied with the last scanned
value ( LSV) of the Contact Interrupt channel. This value will contain a "1" bit for every
relay that is "closed". Note that "closed" refers to the state of the isolation relay on the
mcdu!c.

3-57

3. 7 .3 Sample Program
The following "program" was created to illustrate the usage of the UDCP handler. The UDC in the example has
the following configuration:

a. A BM804 Fljp-Flop Relay board in channel 2.
b. A BW732 Contact Interrupt board in channel 4.

c. A BM804 Flip-Flop Relay board in channel 5.
The function of the program is to indicate that an operator has pressed a push-button and to display the total
number of buttons or switches closed, as well as the last button(s) or switch(es) closed. The BW732 board has
been wired as Close-Only (Pulse-Close). The information is displayed on both the Console Teletype and the UDC.
Both BM804 boards are connected to lamp indicators and the BW732 board is connected to either 12 switches or
12 push-buttons. Channel 5 will display the last change and channel 2 will display the current status of the input.
There are some features of the program that require explanation:

a. The array I RAY( 1,6). .. has been set aside for exclusive use of the Executive according to the
requirements of INITIALIZE.

b. The array ICON IS( 1,3)/4, 104, 105/... provides for the current channel 4 of Contact Interrupt
and two expansion channels (containing Contact interrupt modules) that the user would like
to implement in the future. Since the Executive will not "see" any interrupts on these channels, there is no loss in time by including these channels in the table. If the empty channels
had been specified first in the table, there would be a time loss in lookup and match for the
active channel.

c. The array ID IGS( 1,3)/5,2,-1 / ... contains a negative channel designation and also demonstrates
that channels do not have to be specified in ascending sequence. The negative channel is another
method of reserving a channel slot; but, for a totally different purpose. The Executive will
bypass all negative channel declarations on input and output; however, the Executive will also
bypass the data item associated with that channel. The user, for example, may code an output
transmission to six channels using six items of data and dynamically modify the actual list of
channels to receive the information, as well as the number of channels to be output. Conversely,
the user may code the system to allow an operator to add or delete sample or display points to
the process scanner.
d. The item IDIG2 is reserved in system-level storage and used in #100.SNAP(102+1 ). The item is
initially associated with the negative channel and will be bypassed by the Executive; however,
if the operator, or another .SNAP entered a legal channel value in place of the "-1 ",the Executive
would output IDIG2 to that channel.

3-58

.EQUIPMENT
t

*UDCP

t#lO
t#11

INITIALIZE
IDENTIFY

t#12

TRANSFER

.STORAGE
IRAY(1,6). ICONIS(1,3)/4,104,105/,
tlDEX, ICOS, ILSV, IDIGS(1,3)/5,2,-1/, IDIG2
.INTERRUPT (UDCP)
.PHASE

.ACTION
DO SNAP #100 PRIORITY INTERRUPT
DO SNAP #2 PRIORITY 2.
#100

.SNAP

#121
.HEADER
UDC FAULT
#122

.FORMAT

"CON I INPUT" XX XXXX XXXX
.PROCESS
100

GET (UDCP,#11) IA
GOTO ( 101, 102), IA
EXIT

101

SEND (TTY,#121)
GOTO 100

102

GET (UDCP,#12) ICONIS, IDEX, ICOS, ILSV
SEND (UDCP) IDIGS, ICOS, ILSV, IDIG2
SEND (TTY,#122) ICONIS (IDEX), ICOS, ILSV
GOTO 100

.SNAP
.PROCESS
SEND (UDCP,#10) IRAY, ICONIS

EXIT
.END
3.8 HANDLING THE CONSOLES
INDAC 8/2 is capable of handling up to four ASCII compatible (TTY type) consoles such as the ASR33, KSR35,
VT05, VT06, LP30, etc. One console, which is standard in the INDAC 8/2 System, is permanently defined in the
compiler tables as TTY. Three additional consoles, TTY2, TTY3, and TTY 4, can be configured through
GENDAC. The TTY console has been discussed in terms of the system output (message and logging) device in
previous paragraphs. The use of the console in this application is termed the system mode. The standard TTY
console and any additional consoles can also be used in the "operator mode" for operator guidance of the job.
Two options are available in the operator mode:
System input
Command functions

3-59

The ability to perform command functions can be allowed through program control. Thus, consoles primarily
designated for operator interface can be restricted for system input functions, that is, operator responses to programmed questions. The console designated for the designer or supervisor can be opened for command functions
as well, so that he can inspect and modify system parameters, stop a job, or start another job. After a job is
started (by calling SPUT and commanding the Executive to start the job) all consoles are placed in the system
mode unless specifically programmed otherwise. If any console is placed in the operator input mode, and the
operator responds with an answer, the console is automatically reverted back to the system mode. If the operator
does not respond, the console remains in the operator input mode unless the program cancels the request.
3.8.1

Equipment Declaration Statement

Consoles and associated information are declared in the INDAC program as follows:

t
t#1
t#2

*TTY
COMMAND
CANCEL

where the equipment control lines define the requests that may be made of the 1/0 handler to handle information.
3.8.2 Language Statements
The language statements for accessing the consoles vary in format depending on the mode in which the user wishes
to access the console.

a. System Mode - The following statements are acceptable when the console is in the system
mode:
1.

SEND (TTY, #t)
This statement is used to send the .HEADER or a .FORMAT statement tagged #t to the
console. Normally, a .FORMAT statement would be referenced when the user wishes to
place the console in the operator mode without having a data list to be sent. The declarative .FORMAT statement terminated with a question mark (?) should be used for
this purpose.

SEND (TTY) (data list)
This statement is used to send unformatted data {in exponential notation) to the console
where (data list) is the list of items to be sent.

SEND (TTY,#t) (data list>
This statement is used to send formatted data to the console where (data list> is the list of
items to be sent and #tis the tag reference to the .FORMAT statement. The .FORMAT
statement may be of the picture or declarative type. The declarative .FORMAT statement
should not be terminated with a question mark control character(?) unless the user
wishes to place the console in the operator mode.

SEND (TTY,#t1 , #t 2 ) (data list)
This statement is used to send formatted data to the console and place the console in the
operator mode with the command option permitted. The (data list> is the list of items to
be sent, #t1 is the tag reference to the control line containing the key word COMMAND,
and #t2 is the tag reference to a Declarative .FORMAT statement that is terminated with
a question mark control character(?).

To access the Command Decoder of the Executive the user simply types CTR LIA; to release the Command Decoder the user types CTRL/P.

3-60

NOTE
Once a console is placed in the operator mode, it
will remain in this mode until released. This may be
accomplished by programming a CANCEL command
or by the operator completing an input message.

b. 0 perator Modes
1.

System Input Option
The console may be placed in this mode to allow the user to supply the job with additional information. Up to one line of information terminated with a CR/LF may be
typed. The following statement must be used in the program to retrieve the information.
GET (TTY) (data I ist)
where (data list) are the locations where the information in ASCII representation (one character
per word) is to be stored. After this statement is executed, the program may examine the
operator message.

Command Option
The console may be placed in this mode to allow the user to call the Executive Command
Decoder to inspect and modify system parameters, to stop the job, or to start a new phase.

Cancel
The operator mode can be cancelled to revert the console to the system mode by using
the following statement
SEND (TTY,#t)
where #tis the tag reference to the control line containing the key word CANCEL.

3.8.3 Program Examples
The following program examples are included to illustrate some techniques in programming the consoles. Example 1 illustrates two methods of placing a console in the operator mode. The coder has the option to allow any
console access to the command mode by simply referencing the control line containing the key word COMMAND.
Example 1 - Placing console in operator mode
.EQUIPMENT

t
t #100
#1

*TTY
COMMAND
.PHASE
.ACTION
DO SNAP #2 PRIORITY 2

.SNAP

.FORMAT ("ENTER VALUE-",?)

.FORMAT ("COMMAND MODE AVAILABLE",?)
.PROCESS
SEND (TTY, #3)
or

SEND (TTY, #100, #4)
EXIT

3-61

The INDAC output command SEND (TTY ... ), assumes that the specified console is in the system mode, not
the operator mode. If the console happened to be in the operator mode because it was not released by the user,
the output command will default to the next statement. If this conflict is possible, the "STATUS" item may be
tested to determine if the output was successful.
Examp~e 2 presents an INDAC code that waits for the console to be released from the operator mode.

The

STATUS item will be set to "O" when the output is successful.
Example 2 - Waiting for a console
101

SEND (TTY,#t)
GET (STATUS) Ix
GOTO (101), Ix

Any console placed in the operator mode may be cancelled to release the console under program control. The
system designer can then take appropriate action. Example 3 illustrates some programming techniques for
cancelling a non-responding console.
Example 3 - Canceling a non-responding console
.EQUIPMENT

t
t #101

CANCEL

.PHASE

#10

DO SNAP #3 DELAY 20 SEC PRIORITY 11

#11

*TTY

DO SNAP #4 DELAY 40 SEC PRIORITY 2
.ACTION
DO SNAP #2 PRIORITY 2

.SNAP

#21

.FORMAT ("ENTER VALUE-",?)
.PROCESS
SEND (TTY, #21)
TIMER (START, #10)
TIMER (START, #11)
EXIT

.SNAP
.PROCESS
GET (TTY) I VALUE
EXIT

.SNAP
.PROCESS
SEND (TTY, #101)
EXIT

3-62

The SNAP that typed the output message and placed the console in operator mode was at a "foreground"
priority level. This is usual though not required; however, notice that the SNAP containing the ... GET(TTY ...
is operated at a "run-interruptable" level. This operation guarantees that if the SNAP must be suspended, the

entire system is not suspended also. By this technique, the Executive suspends the input rnquests and allows
the "cancel" to take place.

a. The user may:
1.

Output information to a console in system mode

Change a console to operator mode

Allow a console to access command mode

Cancel the operator mode of a console, placing it in system mode

Input information from a console ... GET (TTY ...

Test if an output request was successful

b. The Executive will:
1.

Initialize the primary console in operator mode (command mode allowed) to access
the Command Decoder

Initialize all other consoles in system mode to prevent unauthorized access to the
system

Allocate a single page of dynamic buffer for each console, to be used for input or
output of data respectively

Collapse the input or output buffer if a CANCEL command is received

Bypass the output request if a ... SEND command is received for a console in operator
mode

Set the STATUS word to a "O" if an output request is successful or to "1" if the request cannot be honored. GET (STATUS) la will transfer the status word into la

Revert any console that has initiated a "run" command (tD R# ... ).back to system
output mode

Revert any console that has completed a GET (TTY) ... back to system output mode

c. System Considerations:
1.

Once the primary console initiates the first phase run command, all consoles will be
in system output mode. It is the prerogative of the user to place any console in operator
mode and to permit any operator mode console access to the Command Decoder and
system-level storage.

Any SNAP requesting a ... GET (TTY ... will be suspended until the operator has completed the input. If the suspended SNAP is a "foreground" priority then, by definition,
the entire system is suspended. It is therefore, adviseable to interrogate consoles from
run-interruptable SNAPs.

3-63

CHAPTER 4
CONFIGURING A SYSTEM
HAVING STANDARD DEC PROCESS 1/0 DEVICES

4.1

INTRODUCTION

A specific system can be built and configured by the guidelines and procedures contained in this chapter. The
cold-start procedure for building and configuring the system is as follows:

Step

Procedure
Load the following programs:
Manually enter the bootstrap loader for HINDAC
HINDAC
MSUP
Monitor System Dump
I NDAC Support Programs
INDAC System Tables
MAKE 8/2
COMP1
COMP2
COMP3
COMP4

Run SGEN.

Load following programs:
Executive 8/2
GENDAC

Run GENDAC to configure system.

After all programs are loaded and SG EN is run, the specific system can be configured using G END AC and its
library tapes. Currently, the GENDAC library tapes contain 1/0 handlers for standard DEC 1/0 devices and trig
functions. As new 1/0 devices are developed and new applications or requirements encountered, the G ENDAC
library tapes will be updated accordingly by DEC. GENDAC, in conjunction with the library tape, generates a
dialogue to enable the operator to select the desired 1/0 handlers and functions. Handlers for the TTY, pseudodevices: KEYS, STATUS, PRIOR ITV and FI LE are not included in the library since they are a permanent part
of the Executive.
Once a library routine is selected for inclusion in the system, the routine becomes a permanent part of the system
and cannot be deleted by GENDAC. If a routine is inadvertently selected, or if a routine is no longer desired,
the complete system must be rebuilt starting with loading the Monitor System Dump (Paragraph 4.3). Then, when

4-1

routine, he can simply run G END AC under the non initial run option and select only that routine he wishes to add.
If the user answers "Y" (yes) to a question pertaining to a routine already in the system, G END AC will abort and
print "CANNOT CONTINUE". At this point, the syst-em communications tables are incomplete and the system
must be rebuilt starting with loading the Monitor System Dump (Paragraph 4.3).
The following paragraphs contain the procedures for building and configuring the software system to complement
the user's hardware. If a sample system was built (see Chapter 2), the system communications tables are configured to reflect the sample system. Since these tables cannot be modified, the system must be rebuilt starting with
loading the Monitor System Dump (Paragraph 4.3).
NOTE
The PDP-8/E control panel differs slightly from other
PDP-8 Computers. The procedures presented in this
chapter detail the PDP-8/E controls. When INDAC
is implemented with a PDP-8/1 or L, use the ST ART
switch whenever CLEAR/CONT is specified in the
procedures; use LOAD ADDR whenever EXTD
ADDR LOAD or ADDR LOAD is specified in the
procedures.
Paragraphs 4.2 and 4.3 contain procedures that are designed for starting the building process from a cold start;
that is, the monitor head is not in core or disk resident (refer to Chapter 5 for the procedure to bootstrap the
Monitor into core). If the monitor head is in core (the monitor period is typed when CTRL/C is pressed), the
MSUP program required for loading the Monitor System Dump (Paragraph 4.3) can be loaded from the highspeed reader with the following command string, instead of with the cold-start procedure described in Paragraph

4.3.
.LOAD)
*IN-R:)

*
ST=200

tt (User types CTR LIP after each t)
Then continue with Step 10 of Paragraph 4.3.

4.2 LOADING THE HINDAC PROGRAM
To load the computer from a cold-core start proceed as follows:
NOTE
Verify that RUN light is off. If the light is on, press
HALT and return HALT switch to up position.
Step

Procedure
Load the switch register with 0000.

Press EXTD ADDA LOAD.

Load the switch register with 0027.

Press CLEAR.

Press ADDA LOAD.
(continued on next page}

4-2

Procedure

Step

Successively deposit the following:
Location

Instruction

0027
0030
0031
0032
0033
0034
0035
0036
0037
0040

6011
5027
6016
7450
5027
7012
7010
3007
2036
5027

Load the HINDAC (Tape 1) program in the high-speed reader - begin anywhere
in the initial blank tape portion.

Load switch register with 0031.

Press ADDR LOAD.

Press CLEAR.

Press CONT.
NOTE
The tape should now read completely through the
reader and stop on the trailer portion (code 0200)
of the tape. The computer should also halt. At this
point, both RIM and the Binary Loader are in core.
If the RUN light does not go out or if the tape does
not read in properly, repeat this procedure.

4.3 LOADING THE MONITOR SYSTEM DUMP AND INDAC FILE TAPES
After the H INDAC program is successfully loaded, the Monitor Support program is loaded using the Binary
Loader:
Procedure

Step

Place the MSUP tape (Tape 2) in the high-speed reader. Set the leader portion
(code 200) of the tape under the read lamp.

Load switch register with 7777.

Press ADDR LOAD.

Set switch register to 3777.

Set the rotary console switch to AC.

Press CLEAR, then CONT.
The tape should now read until the trailer portion (code 200) is under the read
lamp. At this point, the computer will halt with the AC containing O; the link
may be on.

Load switch register with 200.
(continued on next page)

4-3

Step
8
9

Procedure
Press ADDR LOAD.
Press CLEAR, then CONT.
The MSUP program will begin a series of questions to determine the operation
required. Each question may be answered by a single letter followed by a
carriage return (designated) ). After it is loaded, the first question is:
LOAD, DUMP or VERIFY-

Type L)
The next question asked is:
ENTIRE DISK OR FILES?-

Type E)
The MSUP program will come to a halt after typing the "E)". The program is
now waiting for the user to load the Monitor System Dump (Tape 3) in the highspeed reader.

At this point, before loading the tape, set the switch register to 0001. The LSB
switch is used to control the loading of the Monitor System Dump tape. When
the switch is in the" 1" position, the program will idle after completion of loading
the current block (one "block" of information). When the switch is set to "O",
the program will resume loading.
NOTE
Make certain that all checksums are torn from the
end of the Monitor System Dump tape before loading.

Place the Monitor System Dump (Tape 3) in the high-speed reader. Start tape at
leader portion (code 200).

Press CONT. At this point, if you have correctly set the LS.B switch, the program
will be idling.

Set the LSB to "O" to start the loading process. At any time, setting the LSB to
"1" will stop the tape at the next leader/trailer. Loading will resume whenever
the LSB is reset to "O".
NOTE
During the loading process, if the reader malfunctions, MSUP will print "CHECKSUM OR VERIFY
ERROR" and halt. Back the tape up one block
(blocks are groups of punches separated by leader/
trailer, code 200). Place the leader of the block
that failed under the reader lamp. Set LSB for continued loading and press CONT. If the block will
not load correctly, the tape has been damaged and
must be replaced.

When MSUP completes the loading of the Monitor System Dump (Tape 3), there
will be a slight pause and the program will type out:
LOAD, DUMP OR VERIFY-

Place the I NDAC Support Programs (Tape 4) in the high-speed reader. Start the
tape at leader portion (code 200).
(continued on next page)

4-4

Step

Procedure

NOTE
Tear all checksums from the end of the tape (information following the last leader/trailer).
Type L)

The next question asked is:
ENTIRE DISK OR FILES?Type F)

When MSUP completes the loading of the INDAC Support Programs (Tape 4)
there will be a slight pause and the program will type out:
LOAD, DUMP OR VERIFY-

Place the I NDAC System Tables (Tape 5) in the high-speed reader. Start tape at
leader portion (code 200).

NOTE
Tear all checksums from the end of the tape (information following the last leader/trailer).
21

Type L)
The next question asked is:
ENTIRE DISK OR FILES?-

Type F)
When MSUP completes the loading of the INDAC System Tables (Tape 5) there
will be a slight pause and the program will type out:
LOAD, DUMP OR VERIFY-

When both file tapes have been loaded type CTRL/C (hold both CTRL and C
keys down), this will return control to the Disk Monitor System just loaded.
The Monitor will respond with a period. At this point, the disk system is
established for a single disk and the Disk Monitor is resident in core.

4.4 LOADING AND BUILDING THE INDAC COMPILER
Load the MAKE 8/2 (Tape 6) program using the Disk Monitor Loader. To load MAKE 8/2 from the high-speed
reader, use the following command string:
.LOAD)
*IN-R:)

*
ST=200)

tt (User types CTR LIP after each t ).
NOTE
Start the tape at leader portion (code 200).
MAKE 8/2 will execute and return to the Monitor. Now load the compiler tapes.
COMP1 (Tape 7)
COMP2 (Tape 8)
COMP3 (Tape 9)
COMP4 (Tape 10)

4-5

The loading sequence (using the high-speed reader) is the following for each tape:
.LOAD)
*IN-R:)

*
ST=2)
tt (User types CTRL/P after each t).
NOTE
Start tape at leader portion (code 200).
After the user's second CTR LIP control returns to the Monitor after about one minute. The same sequence is
repeated for each compiler tape.
4.5 RUNNING THE SGEN 8/2 PROGRAM
SG EN 8/2 takes a single-disk system and expands the file structure into as many disks as are connected to the
system (Limit 4), and protects the Executive and data file areas that will be used later in the system. Any disk
units on the system that are not to be used by the INDAC System should be switched to "off".
To run SGEN 8/2, type the following sequence:
.SGEN)
.SAVE (SD>! 1400;)
SG EN 8/2 expands the single-disk system to include as many as the user has DS32 expander units selected and
returns to the monitor after a short pause. The SAVE command initializes the System Devices Table in preparation for the system configurator program.
4.6 LOADING THE INDAC EXECUTIVE 8/2
Load the I NDAC Executive as follows:
Step

Procedure
Place the Executive 8/2 (Tape 11) in the high-speed reader. Start the tape at the
leader portion (code 200).

Set the switch register to 3600.

Type HELD)
The HELD program will load the Executive and halt after loading.
NOTE 1
Since paper-tape to disk operations are taking place,
the loading may appear "jerky" or "erratic"; this is
normal.
NOTE2
At this point the AC register should be "O" indic&ting that the checksum comparison is correct. If the
AC is not 0, repeat the Executive Loading procedure.

Set the switch register to 7600

Press EXTD ADDR LOAD.

Press ADDR LOAD.

Press CLEAR, then CONT.
4-6

NOTE
Control returns to the Monitor after pressing CONT.
This completes the procedures required to build the initial system. Continue with the following procedures to
build the specific sample system for checkout and experimental usage.
4.7

LOADING GENDAC TO CONFIGURE A SPECIFIC SYSTEM

The function of GENDAC is to configure the INDAC System for the specific requirements of each installation.
To configure the system proceed as follows:
Step

Procedure
Load GENDAC (Tape l-2) in the high-speed reader.

Set the switch register to 3600.

Type LELD)

The LELD program will load GENDAC and return to the Monitor after loading.
NOTE
Since paper-tape to disk operations are taking place,
the loading may appear "jerky" or "erratic"; this is
normal. About a third of the way through loading
the tape the Teletype will echo a carriage return/
line-feed.

When loading is completed, LELD will return to the Monitor.

Press HALT, then return switch to normal position.

Set switch register to 0200.

Press ADDR LOAD, CLEAR, then CONT.

GENDAC will begin execution (see Figure 4-1).

The initial dialogue from GENDAC types the version number and a reminder to
the operator that if he is unsure of the response to a question, type "?".

GENDAC will then ask if this is to be a 4K system. The response will be "N"
because you do not have a 4K system.

GENDAC will request the mode of operation to be used via the question "*OPT-";
the response will be "B" for binary mode.
NOTE
In the binary mode GENDAC expects to find a formatted binary image produced by the assembly of a
program such as the 1/0 Handler and Function
paper tapes. The contents of these tapes is summarized in Charts 4-1 and 4-2. In the system mode
GENDAC expects to find vector code produced by
the INDAC compiler.

The next request is for the input device via the question "*IN-". The response
will be "R:)" for the high-speed reader.

GENDAC then types "t", waiting for the binary tape to be loaded. Load the
1/0 Handler (tape 13) tape into the high-speed reader and reply by striking
CTRL/P (hold CTRL key and strike P key). GENDAC echos tP.
, ___ .... : - •• -...J - - - - · · .... ____ \

\\.#UllLlllUICU Ull lllCJl.l IJCIHICI

4-7

Step

Procedure

Answer all questions. Answer "Y" (yes) if the 1/0 Handler is to be inserted into
the system; answer "N" (no) if the 1/0 Handler is not to be inserted. In general,
if you do not understand what the routine in question is, type" I" (inspect).
Refer to Figure 4-1 for sample dialogue of current 1/0 Handler tape.
NOTE
GENDAC also asks whether CTRL/C is to be disabled when the console is in the Command mode. If
the user answers yes to this question return to the
Monitor is not possible via the Executive Command
Decoder.

When GENDAC reaches the end of the 1/0 Handler tape it will type:
END OF LIBRARY TAPE.
*OPT-

If you wish to add one or more of the following functions:
SIN

cos
EXP
LOG
ATAN
the response to GENDACs question "*OPT-" should again be "B" for binary
mode. If none of these functions are desired, respond with a carriage return to
the question "*OPT-" to terminate G ENDAC. G END AC will then release
control to the monitor system.
18

If GENDAC is terminated at this point in the system building phase, proceed to
Chapter 5 for information on program preparation or to Chapter 7 for information on modifying the basic system. If one or more of the specified functions
are to be added to the system, proceed as follows:

a. Respond with "B" to G ENDACs question "*OPT-"
b. GENDAC will then type "*IN-" requesting the input device specification.
Type "R:) " for the high-speed reader.

c. GENDAC then types "t" and waits for the binary tape to be loaded.
d. Place the Function tape (Tape 15) into the high-speed reader and strike
CTR LIP (hold CTR L key down and hit the P key). G END AC echos "tP."

e. Answer all questions. Refer to Figure 4-2 for sample dialogue of current
Function tape.
f.

When GENDAC reaches the end of the Function tape, it will type:
END OF LIBRARY TAPE.
*OPT-

Respond with carriage return to terminate GENDAC and refer to Chapters 5
and 7 for operating and modifying the system.

4-8

Chart 4-1
INDAC 1/0 Device Handler library Summary
SELECT ONE

60 Hz PDP-8iE LiNE FREQ CLOCK

CLOCK

50 Hz PDP-8/E LINE FREQ CLOCK

1/0 HANDLER

PDP-8/E PROGRAMMABLE CLOCK

(CLK8)
BY ANSWER ING

60 Hz PDP-8/1

YES(Y)

50 Hz PDP-8/1

TO ONE OF

60 Hz PDP-8

THESE QUESTIONS

50 Hz PDP~8

DIGITAL 1/0POINT-TABLE 1/0
HANDLER

SELECT ONE
ANALOG-TO-DIG ITAL
CONVERTER
(ADC) 1/0
HANDLER

UDCP
GENERIC 2 INTERRUPT MODULE

.-.

[j
D

B
B
E3

AF01
AF02
AF03
AFC8
AD01
AF04

DIGITAL 1/0UDCE

DIG IT AL-TO-ANALOG

AA01

CONVERTER (DAC)

AA50

PTP

TTY2

EXPLICIT 1/0
HANDLER
SELECT ONE

1/0 HANDLER
HIGH-SPEED PAPER
TAPE PUNCH

ADDITIONAL
CONSOLES

TTY3
TTY4

4-9

Chart 4-2
INDAC Function Library Summary
CR SIN, COS
SELECT

DR SIN, COS

FUNCTION

CR EXP

DR EXP

ANSWERING
YES (Y)

CR LOG

TO THESE

DR LOG

QUESTIONS

CR ATAN
DR ATAN

4-10

B
B

•
•
•

•

•
•
•
•
•
•
•
•
•
•
•
•
I•
1·
1·
I

l_~

·'••

.LELD
•
GENDAC
427
TYPE '?' IF CONFUSED.
4K?-?
4Y. INDAC EXECUTIVE?-N

•
•
•
•
•
•'
•
•
•
•

*OPT·B

* IH-R:
tfP

IIO LIBRARY INDAC 8/2 <U2IC>

SAVED <SD>!l400; FROM SGEN?-?
TYPE 'y' TO INCLUDE,
'N' TO OMIT,
'I' FOR MORE INFORMATION.-!

HAVE you SAVED <SD> FOR INITIAL RUH
IF NOT TYPE CTRL/C TO RETURN TO MONITOR7-Y
CHKS... StTORED AT 1021dU
INITIAL RUN?-I
IS THIS THE INITIAL LOADING OF ANY l/O HANDLER?-Y
CZCJ
STORED AT ~117'
END
ADDED TO IO CHAIN •
CISABLE CTRL/C IN COMMAND MODE?-I

D!) '!OU WISH TO DISADLE CTRL/C RETURNING TO THE
MONITOR FROM CONSOLE CO~MA~D M0DE?-Y
SCAN STORED AT 03235
G~

HZ PDP-8/E LINE FREQ CLUCK?-!

-~ HZ PDP-8/E WITH DK8-EA LINE FREQ CLOCK?·Y
CLK8 STORED AT 05102
CLK8 STORF.D AT ~5~45
CLK8 STORED AT 0531'
~~

HZ PDP-8/E LINE

HZ PDP-8/E WITH DK8-EA LINE FREQ CLuCK?-N

FRE~

•
•

CLUCK?-I

•
•
•
•
•

PDP-8/E PROGRAMMABLE CLOCK?-!
PVP-b/f WITH DK8-EP PROGRAriMABLE CLOCK?-N
iik1 HZ PDP-Bil OR PDP-8/L?-I

It: LINE F"REQ -.0 HZ F'UR Pl.JP-bl I OP PLP-8/L?-N
;~

HZ PDP-8/I OR

IS LINE

FH~

P~P-8/L?-I

Stl HZ FOR PVP-B/I OR Pl.)P-8/L?-N

Figure 4-1

GENDAC and 1/0 Handler Dialogue (Sheet 1 of 3)

4-11

I
'" HZ PDP-8 CLOCK?- I
1.

IS PROCEiiOR A PDP-B WITH A ' ' HZ CLOCK?-N

•

IS PROCESSOR A PDP-B WITH A 51 HZ CLOCK?-N

•
•
•
•
•
•

UDCP?-1

•
•
•

UDC POINT-TABLE AND INTERRUPT MODULE HANDLER?-Y
UDC3 ADDED TO <**>
UDCl ADDED TO <**>
UDCP ADDED TO <SD>
UDCP STORED AT 1~201
UDCI
ADDED TO IO CHAIN.
UDC3 STORED AT 12!ii

GENERIC 27-I

•

DO YOU HAVE UDC COMTACT-INTlRRUPT MODULES?-Y
UDC' ADDED TO <**>
UDC5 ADDED TO <**>
UDC5 STORED AT 012ae
UDC• STORED AT 07408
UDCO STORED AT 0224'

•

AF01 ?- I

•

AF02?-I

•

A-D CONVERTER MODEL AF02?-N

•

A-D CONVERTER MODEL AF~37-N

•

ANY Of THE ABOVE A/D?-1

•

AFC-87-1

•

DO YOU HAVE AN AFC-8?-Y
ADC
ADDED TO <SD>
CAFJ STORED AT 81584
AFC
ADDED TO IO CHAIN.

•

•
•

HZ PDP-8 CLOCX?-I

•

•
•
•
•
•

A-D CONVERTER MODEL AFll?-N

AF03?-I

•
•
•
•
•
•

DO YOU HAVE AN AF01, AF02, AF03?-N

AD01?-l

•

A-D CONVERTER MODEL ADml?-N

•

AF~47-I

DO YOU HAVE AN IDVM AF847-N

-----

Figure 4-1

GENDAC and 1/0 Handler Dialogue jSheet 2 of 3)

4-13

•
•
•
•

•
•
•
•
•
•
•
•
•
•
•
•
•
•

UDCE?· I
UDC EXPLICT CHANNEL I/ 0 HANDLER?-Y
UDCE ADDEO TO <SD>
CDOJ STORED AT 01715

•
•
•

AA., 1?- I
AA01 DIGITAL TO ANALOG CONVETER?-N
AA50?- I

•
•

AA5e DIGITAL TO ANALOG CONVETER?-N
PTP?-1
HIGH-SPEED PAPER TAPE PUNCH?-Y
ADDED TO c:;;O>
PIP
CPPJ STORED AT 01320
PIP
ADDED TO IO CHAIN.

•
•
•
•
•
•
•
•
•
•

TTY2 ?- I
ADDFD TELEPRINTER #2?-Y
CT2 l ADDED TO <**>
TTY2 ADDED T0 <SD>
CT2 J STORED AT 0134"1
ADDED TO IO CHAIN.
TTY2
£T2]
STORED AT 03•32
TTY3 ?- I
ADDED TELEPRINTER 13?-Y
(!3]
ADDED TO <**>
TTY3 ADDED TO <SD>
CT3 l STORED AT 01•4i
ADDED TO IO CHAIN.
TTY3
!T3l STORED AT 03•33
TTY4?-I
ADDED TELEPRINTER #4?-Y
[ T4]
ADDED TO <**>
[ T4]
STORED AT 0•153
TTY.ii
ADDED TO IO CHAIN.
TTY4 STORED AT 03•34
ADDED TO <SD>
TTY4
[T4)
STORED AT 01372
END OF LIBRARY TAPE •

•;

*DPT-

e:
Figure 4-1

GENDAC and 1/0 Handler Dialogue (Sheet 3 of 3)

4-15

•
•
•

•

•
•
•

•

-----------.i
*OPT-B
*I N-R:
INDAC 8/2 FUNCTION LIBRARY
[Cl J ADDED TV <**>
ANY OF: SIN,

COS,

EXP, LUG, ATAN?-I

TYPE 'y' IF ANY OF THE ABOVE DESIRED.?-Y
(80] ADDED TO <**>
CBS) STORED AT 175'1
***~Y,TEM BAiE ALTERED!***
CLll ADDED TO<**>
[Lll iTOR!D AT 12118
[LBJ ADDED TO <**>
[Lil iTORED AT 1•371
ANY DISX-RtS: SIN, COi, EXP, LOG, ATAN?-Y
CDl l ADDED TO <**>
CDlJ iTORED AT 18753
CD8J ADDED TO <**>
CD01 STORED AT 17585
CR SIN, COS?- N
DR SIN, COS7-Y

•

SIN
COS
CSD l

•

DR EXP?-Y
EXP
ADD!D TO <XS>
CED] STORED AT 17,7373

•

CR LOG?-N

•

•
•
•

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

ttP

ADDED TO <XS>
ADDED TO <XS>
STORED AT 17, 75 74

EXP?-N

DR LOG?-Y
CHJ l
ADDED TO <**>
CHl l STORED AT 111•5
CHBJ ADDED TO <**>
CH0l STORED AT &•574
LOG
ADDED TO <XS>
CGDJ STORED AT 17,,1,,
CR ATAN?-N
DR ATAN?-Y
ATAN ADDED TO <XS>
[ADJ iTORED AT 17,57•5
END OF LIBRARY TAPE •
*OPT-

•

GENDAC COMPLETED.

•

•
Figure 4-2 GENDAC and Function Dialogue

4-17

CHAPTER 5
PREPARING THE PROGRAM

5.1

INTRODUCTION

After configuring a specific INDAC software system in accordance with Chapter 4, the user can prepare his source
program on-line via the EDITOR and the Teletype. The EDITOR, COMPILER, and PIP are useful in preparing
the source program. These programs are resident on the disk and can be called into operation simply by typing
the assigned name (command string) in response to the Disk Monitor period(.). For example:
.EDIT
.PIP
.COMP
Other system programs that can be cal led into operation are:
.LOAD
.HELD
.LELD
.SGEN
.SPUT

5.2 MONITOR
INDAC uses the standard disk-oriented keyboard monitor, allowing the user to control the flow of INDAC 8/2
programs through the computer. The Disk Monitor also allows the user to LOAD, SA VE, or CALL any other
PDP-8 program he wishes to run. (Refer to Disk Monitor System Manual, DEC-08-SDAB-D for a complete explanation of the Disk Monitor facilities.)

5.2. 1 Monitor Residence
The Disk Monitor and I NDAC system programs reside on disk. Part of the Monitor, called Monitor Head, also
resides in core. It resides in the top page (locations 7600 through 7777) of field 0. The starting address of the
Monitor is 7600. Nonresident portions of the Monitor, such as the Command Decoder, are automatically called
into core as needed by the system programs.

5.2.2 Starting the Monitor
If the computer is turned off or stopped for any reason, the Monitor must be restarted at location 7600. This
can be done by setting the switch register to 7600, pressing ADDR LOAD (EXTD ADDR, if necessary), CLEAR
and CONT. If the Monitor is in the User mode, that is, a system program is running, CTR UC may be typed to
return to the Monitor mode. The rubout key may be used by the operator when his input command string is
erroneous.

5-1

5.2.3 Bootstrapping the Monitor
If the Monitor Head is destroyed for any reason (Monitor cannot be started), the Monitor must be bootstrapped
into core; turning the computer off and subsequently turning it on again does not destroy the contents of core.
Use the following procedure to bootstrap the Monitor into core:
Procedure

Step

Toggle the following routine into core.

Address

Contents

Symbol

0200
0201
0202
0203
0204
7750
7751

6603
6622
5201
5604
7600
7576
7576

OMAR
DFSC
JMP .-1
JMP I .+1
7600

After toggling this routine into core, set switch register to 0200 and press ADDR
LOAD, CLEAR and CONT.

5.2.4 Monitor Error Messages
If an illegal command string is entered, when the user types a carriage return, the Monitor responds with "?"to
indicate invalid input.
Error messages output by the Command Decoder are given in Table 5-1.
Table 5-1
System Error Messages
Meaning

Message

Illegal syntax or miscellaneous error condition

Directory on the systems device is full

Too many inputs or outputs were entered
No such inputs

System iiO faiiure

Monitor time read or write errors cause a halt to occur. Persistence of this condition indicates a hardware failure
or that write lock is enabled on the system device, as the system 1/0 routine attempts to read or write three times
before halting.
5.3 SYSTEM PROGRAMS
The following I NDAC System programs indicate other readiness to receive information by typing either an asterisk
or a query:

.EDIT
.PIP
.COMP
.LOAD
.SPUT

5-2

The most common queries are:
*OUT-

Requests that the user specify one output device name. In the case of disk or
DECtape, the filename to be assigned to the output data must also be specified.

*IN-

Requests that the user specify one or more (up to 5) input device names. For
disk and DECtape, filenames of input files must also be specified.

*OPT-

Requests that the user specify one option or switch, entered as a single alphanumeric character; options available in each system program are described in
system program descriptions (EDIT, PIP, COMP, etc.).

System program is ready to receive commands.

The following paragraphs describe the use and operation of EDIT, COMP, and PIP which are needed to prepare
the I NDAC source program. The system program, LOAD, is also discussed. This program is not needed in preparing the source program.

5.3. 1 Command String Format
Command strings are composed of a few basic elements and follow certain rules of punctuation. To call a system
program into operation and to give it necessary information, the following formats and conventions must be followed.

5.3.1.1 Device Names - Device names permitted in command strings are as follows.
S:

System device (disk or DECtape unit O)

High-speed paper tape equipment (reader or punch)

Low-speed paper tape equipment on the Teletype (reader or punch)

On:

DECtape unit, if both disk and DECtape are present in the system
( n = unit number, 0 through 7)

5.3.1.2 Filenames - Filenames are limited to four characters in length and can be composed of any combination
of alphanumeric characters or special characters with the following exceptions.

a. lmbedded spaces should not appear in a filename.*
b. A filename cannot be one of the following words or symbols.
CALL

SAVE

Extensions (n) to the filenames specified by the user are automatically appended by the system. They are used
internally by the system and cannot be referred to or modified by the user.
SYS (n)

System program file in core bank n

ASCII

Source language program file (input to PAL-I Assembler or INDAC Compiler)

BINARY

Binary program file (output from PAL-I Assembler)

Filenames (and extensions) are meaningful only for file structured devices (disk and DECtape). If they are specified for other devices, they are ignored. Both the filename and extension name appear on directory listings produced by the list feature in PIP.**

*Note that the INDAC Executive is given the filename I OS; one reason for this unconventional use of an imbedded blank is to protect Executive from
accidental destruction by the user (for examole. deletion via PIP)_
**AF in example means VERSION A, change F.

5-3

Example:
NAME TYPE
AF
PIP
.SYS
EDIT .SYS
LOAD.SYS
.CD. .SYS
HGB .ASCII

BLK
(O)

(0)
(O)
(0)

(0)

0025
0016
0011
0007
0001

5.3.1.3 Punctuation - Punctuation within command strings is as follows.
, (comma)

Used to separate device names, when more than one is given in a command
string.
Terminates each device name.

5.3.1.4 Special Characters - Special characters are used as follows.

If given while the system is in Monitor mode or while most system programs
are running, control is returned to Monitor start (location 7600). Monitor responds with a dot.
tC is typed by holding down the CTRL key and striking C. tC does not echo
(does not print).

Typed in response to at typeout. Instructs the system to proceed with the
next operation. tP can also be used to prematurely terminate certain operations while in progress (for example, the typing out of a file directory by the
list option in PIP). tP is typed by holding down the CTR L key and striking P.
tP does not echo (does not print).

)

Carriage return terminates current command string input. When typed alone,
in response to a system query, it indicates that the user does not desire to specify the item (for example, device name) requested.

RUBOUT

Causes the current command string to be ignored, and the system returns to the
beginning of the command string and is ready to receive a new command.
RUBOUT does not echo.

5.3.2 Examples of Command Strings
These foiiowing exarnpies illustrate the elements and rules previously explained. Samples of both Monitor mode
and System mode operation are given.

a. Monitor Mode Commands:
.EDIT)

Call system program file, EDIT, from disk into core for
execution

.PIP)

Call system program file, PIP, from disk into core for
execution

.COMP)

Call system program file, COMP, from disk into core for
execution

b. System Mode Dialogue:
* IN-S:PR02)

Use the file PR02 on the disk as the input file

*IN-S:TST1,R:)

Use the file TST 1 on the disk and one file from
the high-speed paper tape reader as the input files
(continued on next page)

5-4

b. (cont)
*OUT-S:SPEC)

Write the output file on the disk and assign it the
filename SPEC

*OUT-T:)

Punch the output on the Teletype paper-tape punch

*OPT-B

Select option B

5.3.3 Editor
The Editor (Disk System Editor) enables the user to generate and edit INDAC source programs on-line from the
teleprinter keyboard. The I NDAC program may be either printed on the teleprinter, punched on paper tape
using the high-speed punch, or saved on the disk as an ASCII file. To use the Editor, the user must call ED IT via
the Monitor; this can be done only in response to a period. If a period is not present as the last system response,
the user must type CTR L/C which should appear as follows:
.EDIT)
The Editor is transferred from disk into core and responds by typing:
*OUTThe user selects one of the following output devices: (T:) for low-speed reader/punch; ( R :) for high-speed reader/
punch; (S: name) for output to the systems device on a file called name and types his choice immediately after
OUT-. If the specified device is not valid, that is, not declared when building Monitor, Editor will respond with
an error message (refer to Paragraph 5.3.3.6) and return control to Monitor. Thus, the user must call EDIT and
respond to *OUT- with a valid device. When Editor recognizes a valid device, it responds with *) (asterisk,
carriage return/line feed) and *IN-, as shown below.

*
*INThe user now specifies the input device by typing T:) , R: ) , or S: name) or) in the same manner as when replying to *OUT-, above.
The Editor now responds with:
*OPTasking the user to specify one of the following options.

Preserve blanks. Editor normally replaces multiple blanks
(spaces) with tabs, resulting in considerable saving of space
on the system device.

Enter dynamic deletion mode if input is from the system
device. As the file is read, it is deleted from the system device, thus allowing space for output if desired. (Filename
remains on the directory but without any assigned blocks.)

Combine the functions of B and D options.

)

None of the above options; assume conversion of two or more
blanks to tabs, and not D.

After the user has specified one of the options listed above, Editor responds with a carriage return/line feed and
asterisk. The entire printout might appear as follows.

5-5

.EDIT)
*OUT-S:CLK 1 )

*
*IN-T:)

*
*OPT-8
*
*
The appearance of the last asterisk in the preceding example indicates that Editor is ready to accept and operate
on the user's symbolic program.
The user may now type the symbolic program directly into core by using the A command.
*A)
To list a file called CLK 1 with a new output file (CLK) declared, the command string and printout would be as
follows:
.EDIT)
*OUT-S:CLK)

*
*IN-S:CLK1)
*
*OPT-8
*R)
*L)
.STORAGE IA
#1

.PHASE

#10

DO SNAP #100 AT 15:00 PRIORITY 1

#11

DO SNAP #101 EVERY 1 SEC PRIORITY 2
.ACTION
DO SNAP #104 PRIORITY 5

#104

.SNAP
.PROCESS
TIMER (START,#10)
TIMER (START,#11)
EXIT

#100

.SNAP

#102

.FORMAT ("3:00:00 PM ALLS WELL")
.PROCESS
SEND (TTY,#102)
EXIT

#101

.SNAP

#103

.FORMAT (0,/)
.PROCESS
LET IA=IA + 1
SEND (TTY,#103) IA
EXIT
.END

5-6

5.3.3.1

Modes of Operation - To distinguish between editing commands and the actual text to be entered in the

buffer, the Editor operates either in Command mode or Text mode. In Command mode, all input typed on the
Teletype is interpreted as commands to the Editor to perform some operation, or to allow the operator to perform some operation on the text stored in the buffer. In Text mode, all typed input is interpreted as text to replace, be inserted into, or be appended to the contents of the text buffer.
After being loaded into core memory the Editor is in Command mode; that is, the program is waiting for a command. The user types the desired command code and terminates it by striking the carriage return (RETURN)
key. This nonprinting character (represented by) ) tells the Editor to carry out the command. The Editor then
enters Text mode and responds with a line feed character (represented by .J,) as soon as it has processed the command and begun the operation.
With the Editor in Text mode, the user types the desired corrections or insertions to his text. To terminate the
text he enters a form feed (CTR L/FOR M combination) to tell the Editor to return to Command mode. The
Editor answers by ringing a bell to indica\the transition back to Command mode.

CTP..l L
5.3.3.2

Input Commands Command

Action and Explanation

READ a page of text from the paper-tape reader or disk as specified. The
Editor will read information from the input tape or disk until a form feed
character (CTR L/FORM key combination) is detected. All incoming text
except the form feed is appended to the contents of the text buffer. I nformation already in the buffer remains there.
In the case of input via the photoelectric reader, the end of the tape will
be interpreted as a form feed and the Editor returned to Command mode,
if an actual form feed character does not appear on the tape. In the case
of input via the Teletype reader, a form feed must be entered via the keyboard to return the Editor to Command mode, if an actual form feed character does not appear on the tape. If this is not done, the READ command
is still in effect and all subsequent commands will be interpreted erroneously as text and appended to what was just read from tape.
Any rubout encountered during a READ command will be ignored. (See
RUBOUT.)

APPEND the incoming text from the teleprinter keyboard to the information already in the buffer (the buffer may be empty initially). The Editor
will enter the Text mode upon receiving this command and the user may
then type in any number of lines of text. The new text will be appended
to the information already in the buffer, if any, until the form feed (CTR LI
FORM) key combination is struck.
By giving the APPEND command with an empty buffer, a symbolic program
tape may effectively be generated on-line by entering the program via the
keyboard.
The APPEND command must not be used to read paper tapes from the
Teletype reader since every rubout on the tape will delete a character.
NOTE
In both of these commands, the Editor returns to
the Command mode only after the form feed charricter_

5-7

5.3.3.3 Output Commands - Output commands are subdivided into List and Punch commands. List commands
will cause the printout, on the Teletype, of all or any part of the contents of the text buffer to permit examination of the text. Punch commands provide for the output of leader/trailer, form feeds, corrected text, or for the
duplication of pages of an input tape. List or Punch commands do not affect the contents of the buffer.

a. List Commands - The following commands cause part or all of the contents of the text buffer
to be Iisted on the Teletype.
Command

Action and Explanation

LIST the entire page. This causes the Editor to list the entire
contents of the text buffer.

nl)

LIST linen. This line will be typed out, followed by a carriage
return and a line feed.

m,nl)

LIST lines m through n, inclusive. Lines m through n will be
printed on the Teletype.

The Editor remains in Command mode after a List command and the value of the current line
counter is updated to be equal to the number of the last line printed.
b. Punch Commands - The following commands control the punching onto paper tape of leader/
trailer, text and form feeds.

Command

Action and Explanation

Proceed and output entire contents of the buffer followed by a
form feed and return to Command mode.

nP)

Output linen, followed by a form feed, return to Command
mode.

m,nP)

Output lines m through n, followed by a form feed, return to
Command mode.
Process entire file (perform enough NEXT commands to transfer
the remaining input to the output file) and create an end-of-file
indicator (legal only for output to the system device).

TRAILER. This command causes about 4 in. of blank tape to
be punched. If using low-speed punch, turn punch off before
typing command then turn on immediately after typing carriage
return.

NEXT. This is a utility command that combines the functions of
four commands. It punches the contents of the buffer, punches
some blank tape, a form feed and more blank tape, kills the buffer,
and reads in the next page of text from the input device specified
(that is, it executes P), K), R) ).

nN)

Execute the above sequence n times. If n is greater than the
number of pages of input tape the command will proceed in
the specified sequence untii it reads the end of the input tape,
then it will return to Command mode.

5.3.3.4 Editing Commands - The following commands permit deletion, alteration, or expansion of text in the
buffer.

5-8

Command

Action and Explanation

Kl LL the entire page in the buffer. The values of special characters"/" and
"."are set to 0. The Editor remains in Command mode.

nD)

DELETE linen. Linen is removed from the text buffer. The numbers of all
succeeding lines are reduced by one, as is the line count.

m,nD)

DELETE lines m through n, inclusive. The line following n becomes the new
line m and the rest of the lines are renumbered accordingly. The value of the
current line counter,".", is equal to the number of the line preceding the
deleted line or lines. The Editor remains in Command mode after all
DELETE operations.

nl)

INSERT the typed text before linen, until a form feed (CTRL/FORM) is
encountered. The-Editor enters Text mode to accept input. The first line
typed becomes the new linen. Rubouts are recognized. Both the line count
and the numbers of all lines following the insertion are increased by the number of lines inserted. The value of"." is equal to the number of the last line
inserted. To re-enter the Command mode, the CTR L/FORM key combination must be entered to terminate Text mode. If this is not done, all subsequent commands will be interpreted erroneously as text and entered in the
program immediately after the insertion.

INSERT without an argument will insert text before line 1.

nC)

CHANGE linen. Linen is deleted, and the Editor enters Text mode to
accept input. The user may now type in as many lines of text as he desires
in place of the deleted line. Rubouts are recognized during any CHANGE
operation. If more than one line is inserted, all subsequent lines will be automatically renumbered and the line count will be updated appropriately.

m,nC)

CHANGE lines m through n inclusive (m must be numerically less than n).
Lines m through n are deleted and the Editor enters Text mode allowing the
user to type in any number of lines in their place. All subsequent lines will
be automatically renumbered to account for the change and the line count
will be updated.
After any CHANGE operation, return to Command mode is accomplished by
entering a form feed (CTR L/FORM key combination) to terminate input.
After a CHANGE the value of the current line counter,".", is equal to the
number of the last line of the change.
Lines which are changed or deleted do not physically disappear from the buffer area, thus the space they occupied is not recovered upon completion of
the command. This being true, it is possible to overflow the buffer by
changing, or deleting and inserting lines. This possibility may be effectively
eliminated by logically segmenting a program on paper tape into "pages" of
50 to 60 lines. This is done by punching groups of 50 lines followed by a
form feed character (see output commands). There is a way to retrieve lost
space in some cases by use of the SEARCH command (refer to SEARCH
command which follows).

m,n$kM)

MOVE lines m through n inclusive to before line k (m must be numerically
less than n, and k may not be in the range between m and n). Lines m through
n are moved from their current position and inserted before line k. The lines
are renumbered after the move is completed, although the value of the current
line pointer,".", is unchanged. Moving lines do not use any additional buffer
space.
(continued on next page)

5-9

Command

Action and Explanation

m,n$kM ~
(cont)

A line or group of lines may be moved to the end of the buffer by specifying
k as "/+1". Example: 1, 10$/+M). Since the MOVE command requires
three arguments, it must have three arguments to move even one line. This
is done by specifying the same line number twice. Example: 5,5$23M).
This will move line 5 to before line 23. The Editor remains in Command
mode after a MOVE command.

GET and list the next line that begins with a tag or label. The Editor begins
with the line following the current line (line .+1) and tests for a line which
does not begin with a tab, slash, or a space. This will most often be a line
beginning with a tag or label.

nG)

GET and list the first line after linen which begins with a tag. The Editor
begins with linen and tests it and each succeeding line as previously described.
Both G and nG update the current line counter after finding the specified line.
However, if either version of the GET command reaches the end of the buffer
before finding a line beginning with other than a tab, slash, or space, the current line counter retains the value it had before the GET was issued and a"?"
is typed to indicate that no tagged line was found. The Editor remains in
Command mode after a GET command.

nS)

SEARCH linen for the character specified after the carriage return. Allow
modification of line when character is found.
The SEARCH command is one of the most useful functions in the Editor. It
is also structured somewhat differently from the other Editor commands.
After terminating the command nS with a carriage return the user has told
the Editor to SEARCH linen, but he has not specified what to search for.
The Editor is, therefore, waiting for the user to type a character. The character he types is taken as the object of the search but is not echoed. The Editor
instead immediately begins typing out the specified line. After typing the
character for which it is searching the Editor stops. All of the editing features
are then available to the user. He may proceed using any of the following:

a. Delete the entire typed portion of the line by typing "*-" (back arrow).
b. Delete the entire untyped portion and terminate the line and the search
by typing~ (carriage return).

c. Delete from right to left one of the typed characters for each "\"
(rubout) typed.

d. Insert characters after the last one typed simply by typing them.

e. Insert a carriage return and line feed, thus dividing the line into two,
by typing -!- (line feed).

Continue searching to the next occurrence of the search character by
typing CTRL/FORM. When typing stops all options are again available.

g. Change the search character and continue searching by typing CTR LI

BELL followed by the new search character.
Each time the Editor types the character for which it is searching, typing stops
and all or any combinations of the above operation may be carried out.
(continued on next page)

b-10

Command

Action and Explanation

m,nS)

SEARCH lines m through n inclusive as previously described. The search
character is input after the carriage return and all of the options are available.
The only difference is in point b. Typing') (carriage return) deletl?S the entire untyped portion and terminates that line, but the search continues on
the next line.
By typing CTR LIB ELL to change search characters, all editing of a single line
may be done in one pass. Clearly, typing CTRL/BELL twice will cause the
search to proceed to termination, since the search character will now be BE LL,
which is not stored in the buffer.
An additional feature is availc;ible to the more sophisticated user: by typing
S with no arguments the entire buffer may be searched for occurrences of a
single character. It must be remembered, however, that as with every
CHANGE command, every SEARCH command uses additional buffer spac.e
for storage of the new line. This is obviously necessary, since the program
can have no prior knowledge of whether the size of the line will be less than,
greater than, or equal to that of the old line, and it must therefore assume
that it will be larger. As the entire buffer is searched, a new image of the
text is created in core that is guaranteed to occupy the same or less space
than previously, since all deleted spaces have been removed. The Editor recognizes this and immediately moves the text image back to the top of the
buffer space. Thus, the only prerequisite to condensing the text image is that
there be enough core space left to contain another image of the edited text.

5.3.3.5 Special Characters and Functions - A number of keys have special operating functions. These keys and
their associated functions are listed below. The nonprinting characters are noted; the symbols for these are shown
in parentheses. All others echo the character in parentheses.

a. tP (CTR LIP) - During output, processing stops and control is returned to the Command mode.
b. tC (CTR L/C) - Always returns to Monitor.

c. Carriage Return ( ') nonprinting) - In both Command and Text modes, striking the carriage return key (RETURN) signals the Editor to process the information just typed. In Command mode,
it allows the Editor to execute the command just typed. A command will not be executed until it
is terminated by striking the RETURN key (with the exception of= which needs no')). In Text
mode, it causes the line of text which it follows to be entered in the text buffer. A typed line is
not actually part of the buffer until terminated by a carriage return.

d. Back Arrow (+-) - The back arrow (+-) is used for error recoveries in both Command and Text
modes. When used in Text mode,+- cancels everything to the left of itself back to the beginning
of the line. The user then continues typing on the same line. When used in Command mode,
+-cancels the entire command and the Editor issues a"?" and a carriage return/line feed (CR/LF).
Back arrow cannot cancel past a CR/LF in either Command or Text mode.
A+-? (CR/LF)
THIS+- "HERE IS A TEXT MODE EXAMPLE" (CR/LF)
only the part in quotes is entered in the buffer

e. Rubout(\) is also used in error recovery in both Command and Text modes with one exception.
When executing a READ command from either the paper tape or Teletype reader, rubouts are
ignored completely and do not go into the buffer.
I - - ·- ~ ~ -- - - - -•
\l;UllLlllU~

5-11

- --

Ull

-- •
ll~X l

\
µcty~/

e. (cont)

It is necessary for the READ command to disable the rubout function since all tab characters on
paper tape are, for timing purposes, followed by rubouts which would destroy the tabs. Rubouts
are not stored in the text buffer but are inserted by the Editor following all tab characters on the
output tape.
At any other time in Text mode (specifically if Text mode was entered via the APPEND,CHANGE,
INSERT, or SEARCH command) typing rubout echoes a back slash(\) and deletes the last typed
character. Repeated rubouts delete from right to left up to, but not including, the CR/LF combination, separating the current line from the previous one. Example:
THE QUUICK \\\\ ICK BROWN FOX (CR/LF) will be entered in the buffer as
THE QUICK BROWN FOX
When used in Command mode, rubout is equivalent to back arrow and cancels the entire command.
The Editor then issues a"?" and a CR/LF combination.
f.

Form Feed (CTR L/FORM nonprinting) - Form feed signals the Editor to return to Command
mode. A form feed character is generated by depressing and holding the CTR L key and hitting
the FORM key. This combination is typed while in Text mode to indicate the desired text has
been entered and the Editor should now return to Command mode. The Editor rings a bell in response to a CTRL/FORM to indicate the transition back to Command mode. If Editor is already
in Command mode when CTR L/FORM is typed, no bell will sound. CTR L/BELL is equivalent
to CTRL/FORM except in the case of a SEARCH command (see editing commands).

g. Period (.) - The Editor keeps track of the implicit decimal number of the line on which it is cur-

rently operating. At any given time the symbol period (.) stands for this number and may be used
as an argument to a command. Example: .L) means list the current line .. -1,.+1 L) means list
the line preceding the current line, the current line, and the line following it.
After a READ or APPEND command, the current line counter(.) is the number of the last line in
the buffer. After an INSERT or CHANGE command,(.) is equal to the number of the last line
entered. After a LIST command,(.) is the number of the last listed line. After a DELETE command, (.) is the number of the line immediately before the deletion. After a Kl LL command, (.)
is 0. After a GET command,(.) is the number of the line typed by GET. After a MOVE or
SEARCH command, the current line counter(.) is not updated and remains at whatever it was
before the command.

h. Slash (/) - The symbol slash (/) has a value equal to the decimal number of the last line in the
buffer. It may also be used as an argument to a command. Example: 10,/L) means list from
line 10 to the end of the buffer.

Line Feed (t nonprinting) - Commands are terminated by a carriage return/line feed (CR/LF)
combination and the lines on each page of text are separated by a CR and LF. The user need
only strike the RETURN key, however, to terminate a command or input line, since the Editor
automatically generates a line feed to follow each carriage return.
On input from paper tape, line feed characters are completely ignored. On output the Editor
automatically punches a line feed following each carriage return.
Typing a line feed while in Command mode is equivalent to typing ".+1 L) "and will cause the
Editor to type out the line following the current one and increment the value of(.) by one.

ALT MODE (ALT nonprinting) - Hitting the ALT MODE key (ALT MODE) while in Command
mode will also cause the line following the current line to be typed out and (.) to be incremented
by one. If the current line is also the last line in the buffer, typing either ALT MODE or Line
Feed will be answered by a"?" from the Editor indicating there is no "next" line. Some Teletypes have an Escape (ESC) key in place of the ALT MODE. The function is identical for ESC
or ALT MODE.
(continued on next page)

5-12

k. Left Angle Bracket(<) - Typing Left Angle Bracket (<) while in Command mode is equivalent
to typing ".-1 L) "and will cause the Editor to echo< and then type out the line preceding the
current line. The value of(.) is decreased by one so that it still refers to the last line typed out.
I.

Equal Sign (=) - The equal sign (=) is used in conjunction with the pointers{.} and {/}. When
typed in Co"?mand mode, it causes the Editor to print out the decimal value of the argument
preceding it, followed by a CR/LF. In this way, the number of the current line may be found
(.=XXX), or the total number of lines in the buffer (/=XXX) or the number of some particular
line (/-8=XXX) may be determined without counting from the beginning.

m. Colon ( :) - Colon is a lower case character with exactly the same function as(=).
n. Blank Tape and Leader!Trailer - Both Blank Tape and Leader!Trailer (code 200) are completely
ignored on an input tape, as are line feed characters and rubouts. Line feeds and rubouts are
automatically replaced wherever necessary on output; blank tape and leader/trailer are not.

o. Tabulation(--* nonprinting) - The Editor is written in such a way as to simulate "tab stops" at
ten space intervals across the carriage. When the user holds the CTR L key and strikes the TAB
key, the Editor produces a tabulation. A tabulation consists of from one to ten spaces, depending on the number needed to bring the carriage to the next tab stop. Thus, the user may use the
Editor to produce neat columns on the hard copy.

5.3.3.6 Editor Error Messages - Editor will print an error message consisting of a question mark whenever the
user requests nonexistent information or uses an inconsistent or incorrect format in typing a command. The
question mark will be followed by a carriage return/line feed and the command will be ignored. If the computer
halts at location 2330, a system error has occurred while reading from the disk. The user should, therefore, run
the disk maintenance tests to determine the cause of the error.

5.3.4 INDAC Compiler
After the INDAC source program is prepared and saved on the disk or on paper tape using EDIT, the compiler is
used to translate the INDAC program from ASCII (source) to vector (object) code for disk storage. To use the
compiler, the user must call COMP via the Monitor. This can only be done in response to a period. If a period is
not present as the last system response, the user must type CTR L/C, which should cause the Monitor to type the
needed period. The printout should appear as follows:
.COMP)
The compiler is transferred from disk into core and responds by typing:
*OUTThe user must then respond with the output device name and the filename to be given the compiler program. The
device name will always be the disk (S:). The filename can be any four printing characters except the exclamation mark and the space. When the compiler recognizes the device and filename as valid, it responds with (*)
(asterisk, carriage return/line feed) and *IN-, as follows:
*
*IN-

5-13

The user must now respond with the input device name and the filename given to the I NDAC source program if
it was stored on the disk during the Editing operation. The compiler will accept an input from the disk (S:) or
the high-speed reader ( R :). Because the I NDAC source file is in ASCII and the object file will be stored on the
disk as a system file, no conflict is caused by using identical names if the input file is on the disk. No name need
be specified if the source file is received from the high-speed reader. Examples of the command strings for the
two input methods follow:
Reader

Disk
.COMP)

.COMP)

*0UT-S:TST1)

*OUT-S:TST1)

*IN-S:TST1)

*IN-R:)

*OPT-~

*OPT-)

t type CTR LIP in response
to up-arrow.
After the user responds to the input question correctly, the compiler types:
*OPTThis question is asked to inquire if a listing of the entire compiled program and accompanying data is desired. A
reply of L produces the list. No list is produced if the CR ()) key is pressed.
NOTE
Complete listings are lengthy and are normally used
for system diagnostic purposes only.
If the input file is on the disk, compiling starts when the CR ()) key is pressed. In the case of inputs from the
high-speed reader, the user must then type CTRL/P in response to the up-arrow (t).
If, during compilation, the system halts with a value of 0721 8 in the accumulator, there is no more room available
on the disk. Remove some files from the disk, via PIP, and recompile the program. Any extra user files should
be deleted.

5.3.4.1 Compiler Output - The compiler prints all system level variables, those items declared in the job-level
storage statement, regardless of the option used following termination of the *OPT- command. The format of
these variables is:
Variable name @ storage location
The storage location addresses are required later to modify or inspect parameters by the Executive Command Decoder. A possible system level storage table would resemble the following format:
IHR @ 1 IMIN@ 2 ISEC@ 3 IHR1 @ 4
ES @ 5 R @ 8 PRES @ 11 VPR @ 17
Immediately after the storage variables, the compiler prints pairs of identical 4-digit numbers and then a final 4digit number, in the format:
0162

0162

0221

0342

1236

5-14

These numbers are SNAP and external subroutine level requirements that refer to the highest location used in
each SNAP and SUBROUTINE segment. The row containing the three 4-digit numbers is always printed last by
the compiler. The third number in the last row is the core address for the start of the user's system. The sum of
the third number in the last rmN and the largest of the other numbers prirHed cannot exceed 6600 for an 8K sys8

tem consisting of one Teletype. For each additional Teletype subtract 2008 from 66008 and, if FI LE is used,
subtract an additional 2008 from it. Therefore, for a system composed of two Teletypes and FI LE, the sum
cannot exceed 62008 . If the sum of the two numbers is greater than the calculated limit, the program must be
rewritten to occupy less space in memory, by segmenting the largest SNAP into two SNAPS, for instance.
After compilation is complete, control is returned automatically to the Disk Monitor System, as indicated by its
reply of a period.

5.3.4.2 Errors During Compilation - If any errors occur in the user's source level program during compilation,
the compiler prints an error message that consists of a number or a letter and a number. The error messages are
printed out as they are encountered during compilation and are, therefore, interspersed with the system level
parameters. Each error message is always followed by an address indicating exactly where the error occurs in the
program. For errors found before the first program tag, the notation STAT is used; for errors found after the
first program tag, the tag number is printed. The exact address is determined by the +n value printed after the
tag, where +n means n statement down from the tag statement,
E7 STRT+3
TEMP@01ILOG@10
H2 #20o+4
17 #600+1
0614

0614

5376

In this example, the first error is an E7 encountered three statements from the start of the program; the second
error is an H2 located four statements from tag #200; and the last error is a 17 found one statement from program tag #600. Notice that the system level parameters TEMP and I LOG were printed as they were encountered
in the program. Refer to Appendix D for the meaning of each error message.

5.3.4.3 Correcting Compilation Errors- The Editor is used to correct any errors in the INDAC program. Use
the following procedure to correct the errors.
Step

Procedure
Call the Editor by typing ED IT after Monitor's period printed after compilation.

Answer Editor's *OUT- with a new n.ame to be given to the corrected program in
the form S: filename. A different name must be used from that of the error containing ASCII program already on the disk. As when the user's program was loaded
initially, the compiler will accept input from the disk or from the high-speed
reader. If the high-speed reader is to be used as the input for the Compiler, the
corrected I NDAC program must be on paper tape.

Respond to the Editor's* IN- with the present name of the program. Use the notation S: filename for the program stored on the disk with the errors.

Choose the appropriate Editor option for *OPT-; B is generally a good choice.

Type R) to cause the Editor to read in the first block (buffer) of the program.
(continued on next page)

5-15

Step

Procedure

If the exact location of the error is not known, type /L (slash and L) to list the
last line of the first block. If the error is in this block, use the Editor commands
to correct the error.

After the errors in the first block are all corrected, or if there are no errors in the
first block, advance to the next block containing an error. Type N) to write the
first block onto the device specified in Step 2 and to read in the next block. Type
N ) enough times to read in the program blocks up to the block containing the
next error. The same technique can be used to locate the block containing the
error, namely, type /L to list the last line of the block presently in the buffer. Use
the Editor commands to correct all errors when they are located.

When all errors have been corrected and if output is not to the system device, type
N until a ? is printed, indicating an end-of-file condition. If output is to the disk,
type E) to output the rest of the program to the disk.

When all the errors have been corrected, the program must be recompiled. Be sure to use the new file name given
to the Editor in response to the output question, Step 2, before correcting the errors. Most users may find it
advantageous to call PIP to delete the incorrect file from the disk to save space.
5.3.5 PIP
PIP (Peripheral Interchange Program) performs general utility operations, such as listing the contents of specified
directories, deleting unwanted files from the system device, and transferring files between devices, and copying
specified files. PIP enables the user to do any of the above operations merely by typing commands from the
teleprinter keyboard.
To use PIP, the user must call PIP via Monitor which can be done only in response to a period. If a period is not
present as the last system response, the user must type tC, which causes Monitor to type the needed period. The
printout appears as follows:
.PIP)
PIP is transferred from the disk into core and responds by typing:
*OPTThe user now selects one of the following options.
L

List entire directory of device to be specified

Delete a file to be specified

Aor)

Copy ASCII file (destination and origin(s) to be specified)

Copy binary file (destination and origin to be specified)

Copy FORTRAN binary file (destination and origin to be specified)

u
s

Copy user file (file structured origin and destination to be specified)*
Copy system file (file structured origin and destination to be specified)*

The user types only the option character, to which PIP immediately responds with a carriage return/line feed.
The user does not terminate the line with the RETURN key; it is a meaningful option.

*User and system files may not be copied onto paper tape because they are core images and have no defined paper tape format.

5-16

If the user selects an option using any character other than one of those listed above, the option is illegal, and PIP
ignores the request, types? (question mark), and asks for another option character. The output would appear as
follows:
*OPT-G

?
*OPTThe L option lists the entire directory of the system device (disk) or DECtape on which a directory exists. For
example,
.PIP)

User calls PIP

*OPT-L

list option of the

*IN-S:)

system device directory

FB=0241

PIP types number of free (unused)
blocks remaining on specified device

NAME

TYPE

BLK

followed by filename and descrip-

PIP

.SYS (O) 0025
EDIT .SYS (0) 0016

tion; for example, PIP is a system

LOAD.SYS (O) 0011

blocks of storage

.CD.

program in field 0 and occupies 25 8

.SYS (O) 0007

HELD.SYS ( 1) 0001
LE LO .SYS (0) 0001
SGEN .SYS (O) 0015
SPUT .SYS (O) 0012
When the user specifies the D (delete a file) option, PIP responds with
*FILE TYPE (A,B,F,U,S)where A, B, F, U, and Sare legal options from which the user may choose; indicating ASCII, binary, FORT RAN
binary (compiler output), user, and system program, respectively. Options F and U are not used in INDAC 8/2.
If the user's reply is S) , indicating a system file, PIP asks
REALLY?
PIP will not delete a system file unless the user answers by typing

(meaning yes)

to the question. Any reply other than Y) causes PIP to repeat the FI LE TYPE request. When the user types

Y) , PIP responds with
*INand waits for the user to specify the device and filename of the system file to be deleted. The printout would
appear as:

5-17

*OPT-D
*FILE TYPE (A,B,F,U,S)-S ~
REALLY?N)
*FILE TYPE (A,B,F,U,S)-S)

Delete option specifying
system file, user must reply
with Y),
Pl P repeats request,
user replied correctly,

REALLY?Y)
*IN-S:SGEN)
*OPT-

PIP needs device and filename,
file is deleted and PIP asks for
the next option.

When the file has been properly identified and deleted PIP returns to ask for another option. If filename SGEN,
in the example above, had not been on the specified device, PIP would have ignored the request and typed a "?"
before asking for another option. For example,
*IN-S:SGEN)

SG EN is not the name of

?
*OPT-

a file on the specified
device

The user should not try to delete system files .CD. or LOAD.
Options A, B, and Sare used to transfer files from one device to another in the INDAC 8/2 System. When the
user has requested any of these options PIP responds with
*OUTand waits for the user to specify the destination or output file and, if the destination is disk or DECtape, the name
of the file. For example,
*OPT-A

copy an ASCII file option

*OUT-S:TSTl ~

specifying the destination and filename

Only one destination is legal; if the user specifies more than one, Pl P will ignore the response, type the error message E, and return control to Monitor. For example,
*OPT-A

copy an ASCII file option

*OUT-S: TSTl, E

PIP recognizes the comma, which
is used to separate file names;
control returns to Monitor

NOTE
The Land D options return to PIP's option request
(*OPT-) when the user responds illegally, and all
other options return control to Monitor.
PIP indicates acceptance of the user's destination by responding with *,carriage return/line feed, and *IN-, and
waits for the user to specify the input, that is, to state from where the input is to originate. An attempt to
specify more than one input to any but the A option will cause PIP to ignore the response, type tne error message E, and return control to Monitor. For example,
*OPT-S
*OUT-S: TSTl ;
*
*IN-S: TST2, E

copy a System file option
specifying system device and filename
PIP accepts user's destination
input to system device, comma is
used to separate device names control returns
to Monitor

5-18

The A option will allow any combination of up to 11 ASCII input files to be merged into one output file in the
order specified by the input list. The user, therefore, may write generalized subroutines as separate files to do his
often repeated operations and then, by combining these with each specialized program before assembly, eliminate
the need to ieWiite such opeiations fo; each p;ogram. PIP acknowledges each legai inpui fiie by priniing an *.
If, however, the input file specified to any option is not found on the specified device, PIP prints I in place of the
* and returns to the Monitor. For example,
the file does exist; when the user types CTRL/P,
copying begins

*IN-S:TST1)

*t
*I N-S: TST2)
I

the file does not exist
control returns to Monitor

If the user requests the 8 option, indicating he wishes to copy a binary file but the filename he has specified appears as an ASCII file, it is not acceptable; therefore, PIP prints an I and control returns to Monitor. The user
can ascertain file types by using the L option and checking the file directory.
A summary of the copy features of PIP is presented in the following table.
Number of

High-Speed

Option

Input Files

Disk

DECtape

Reader/Punch

Teletype

ASCII

Yes

Binary

Yes

FORTRAN
Binary

Yes

User

Yes

System

Yes

5-19

Examples
.PIP )

User calls PIP

*OPT-L

and requests the list option

*IN-S:)

of the system device directory

FB=0111

PIP types number of free (unused)
blocks remaining on specified device

NAME TYPE

BLK

e.g., PIP is a system program in field 0

AF
PIP

followed by filename and description;

.SYS (0) 0025

and occupies 25 8 blocks of storage

EDIT .SYS (O) 0016
LOAD.SYS (O) 0011
.CD.

.SYS (0) 0007

HELD.SYS ( 1) 0001
LELD .SYS (O) 0001
SGEN .SYS (O) 0015
SPUT .SYS (0) 0012
<PZ>.SYS (0) 0001
<CM>.SYS (O) 0001
<IF> .SYS (0) 0001
<SD>.SYS (0) 0003
<XS>.SYS (O) 0001
COMP.SYS (O) 0022
OVAL.SYS (0) 0125
COPS .SYS (O) 0014
I OS .SYS (O) 0152
FILE .SYS (0) 0106
DOC .ASCII

0035

*OPT-D

User requests the delete option and

*FILE TYPE (A,B,F,U,S)-A)

specifies type of file, A(ASCI I) and

*IN-S:DCC)

device and filename; file is deleted

*OPT-D

User requests the delete option and

*FILE TYPE(A,B,F,U,S)-S)

specifies type of file, S (system)

REALLY?Y)

(PIP double checks); Y is only the
meaningful answer

*IN-S:SGEN)

User specifies file and filename;
file is deleted

*OPT-L

User requests list option

*IN-S: ~

and system device directory,

C'"O-r\11'.' ....

Note increase of 52 8 free blocks

10-u10..>

(see above)
(continued on next page)

5-20

NAME TYPE

BLK

AF
PIP

.SYS (O) 0025

Note removal of two files:

EDIT .SYS (0) 0016

ASCII file DDC

LOAD.SYS (O) 0011

and

.CD.

.SYS (0) 0007

System file SG EN

HELD.SYS ( 1) 0001
LELD .SYS (0) 0001
SPUT .SYS (0) 0012
<PZ>.SYS (0) 0001
<CM>. SYS (0) 0001
<IF> .SYS (O) 0001
<SD>.SYS (0) 0003
<XS>.SYS (O) 0001
COMP.SYS (O) 0022
OVAL.SYS (0) 0125
COPS .SYS (O) 0014
I OS .SYS (0) 0152
FILE .SYS (O) 0106
*OPT-D

User requests delete option

*FI LE TYPE (A,B,F ,U,S)-S)
REALL Y?N;

Y is only response for deletion of

*FILE TYPE (A,B,F,U,S)-S)

a system file; other responses

REALLY?W)

cause PIP to repeat the file type

*Fl LE TYPE (A,B,F,U,S)-S)

request

REALL Y?Y)
*IN-S:I OS.)

Even if user responds to REALLY?

?
*OPT-D
*FI LE TYPE (A,B,F ,U,S)-U

with Y, PIP will not delete the
Executive

*IN-S:NONE~

PIP knows NONE is not an existing

user filename on the system device

*OPT-D

and indicates by typing ?

*FILE TYPE (A,B,F,U,S)-A)

User requests ASCII file option

*IN-S:EDIT)

PIP also knows when the filename

and file type do not match; ED IT is

*OPT-D

a system program

*FILE TYPE (A,B,F,U,S)-8)
*IN-S:EDIT)
?
*OPTMerge, into an ASCII file on disk "TST1", one tape from the reader, one tape from the Teletype, one file from
disk called SRC, and one file from DECtape 7 called SRC1.

5-21

*OPT-A
*OUT-S:TSTl)

*
*I N-R :,T:,S:SRC,D7 :SRC1 )

*
*
*
*tttt

(type CTR LIP after each file)

*OPTCopy the system file Pl P from disk to DECtape 3 using filename Pl PX.
*OPT-S
*OUT-D3:PIPX)

*
*IN-S:PIP)

(type CTR LIP)

*OPTTry to merge two binary files onto disk called BIN from paper tape.
*OPT-B
*OUT-S:BIN ~

*
*IN-R:,E

(list exceeded)

Try to copy an ASCII paper tape from high-speed reader, a nonexistent file from DECtape 5, and a paper tape
from Teletype to high-speed punch.
*OPT-A
*OUT-R)

*
*IN-R:,D5:FOO,T:)

( R: accepted as legal)
(D5:FOO rejected, no such file
on D5:)

5.3.6 Loading Programs - Disk System Binary Loader
The Disk System Binary Loader takes as input the binary coding produced by the PAL-I Assembler or other DEC
assemblers and loads it into core in executable form. When loading is completed, Loader "disappears" after first
entering the loaded program at the starting address typed by the user just prior to loading. Loader accepts input
from the system device or paper tape.
5.3.6.1

Binary Loader Operating Procedures
.LOAD)

Direct Monitor to print Binary Loader from the
system device into core for execution.

(continued on next page)

5-22

*IN-

Loader requests source of input(s). Type one or
more device names, separated by commas. If an
input device is a file-structured device, include
filename(s).
Up to five files can be specified.

Examples
*IN-R:)

Input one tape from the paper tape reader.

* IN-R :,R:,R:)

Input three tapes from the paper tape reader.

*IN-S:INPT)

Input the file INPT from the system device.

*IN-S:BIN2,R:)

Input the file BI N2 from the system device and one
tape from the paper tape reader.

*IN-S:BIN1,S:BIN2 ~

Input the files BIN 1 and B IN2 from the system
device.

If device(s) are valid and filenames (if any) are
actually found on the system device, Loader responds with one asterisk for each correct input.

*ST=

Loader requests the starting address to which control is to be transferred when loading is completed.
The address is typed in the form
fnnnn
where f =field numberl (omitted if field 0), and
nnnn = location with field.

Examples

*ST=)

Load into field 0.

*ST=7600)

Return to Monitor after loading.

*ST=O;
*ST=30225)

Load into field 3.
Jump to location 255, field 3, after loading.

*ST=10000)

Load into field 1.
Return to Monitor after loading into field 1.
Loader now types a series of up-arrows, one at
a time, as explained below.
Following each up-arrow typeout, the user is
required to perform one or more actions.

First up-arrow: Loader is ready to load. If
paper tape input, put the tape in the reader.
Type tP.2
Second up-arrow: End of pass. Type tP to jump
to previously specified starting address.
(continued on next page)

1
2

The f-digit forces Loader to start loading into the specified field until a "field setting" is found in the input file or tape.
1f Teletype paper tape equipment is used, type

tP before turning on the reader.
5-23

tt (cont)

Multiple Input Files
An up-arrow is typed out as the processing of
each input file is completed. If paper tape input,
insert the next file in the reader and type tP.
Repeat the above step until all files given in response to the *IN- request have been processed.
After all files have been entered, type tP to jump
to the previously specified starting address.
NOTE
After each input paper tape is read, the high-speed
paper tape version of Loader loops until the user
types tP to continue.

At this point, Binary Loader disappears and control is transferred to the previously specified starting address.
5.3.6.2 Binary Loader Error Messages - An illegal checksum error condition causes Loader to type"?" and return to Monitor after the user types tP or tC. Error messages for illegal filenames or devices are as specified in
Paragraph 5.2.5.

5-24

CHAPTER 6
EXECUTING THE PROGRAM

After the I NDAC source prograJTI is compiled error free, the object program is stored on the disk as a system file.
SPLIT, the system-put-together program, converts the object program into absolute disk format so that the INDAC
Executive can locate and execute the source program at the most efficient speed. After converting the source
program, SPLIT bootstraps the Executive into core. The Executive performs scheduling, interrupt control, disk/
core management, input/output, handles operator terminal communication, and executes the compiler generated
code. Once bootstrapped into core, the Executive waits for the operator to start the program by giving it a
specific run command. These commands are interpreted by the Executive Command Decoder which can be
called by typing CTR LIA on any command console. After the run instruction (CTR LID R # [Phase tag]) is
given and the Command Decoder is released, the console reverts to the system mode. The only way the Command
Decoder can be called again is via the I NDAC program placing the console in the operator mode with the command mode permitted. Then, the operator may call the Command Decoder to inspect and/or modify system
level parameters without interfering with the execution of the INDAC program.
The INDAC object file is stored on the disk only after error-free compilation, as indicated by the compiler printout. Before running SPLIT, make sure that no SNAPS are too large to be run (refer to Chapter 5 for details). After
the compiler finishes with its printout, it will release control to the monitor which then types the monitor period.
The user may then call SPLIT by typing:
.SPLIT)
SPLIT is transferred from disk to core and responds by typing:
*INThe user must now respond with the input device name and the file name given to the I NDAC object program
during compilation. SPLIT will only accept an input from the disk (S:). When SPLIT recognizes the device and
file name as valid it responds by typing:
*Fl LE START XXXX
EXEC LOADED
The Executive is called directly by SPLIT and only the name given to the object program file during compilation
needs to be specified. SPLIT types a decimal block number after *FI LE START indicating where the file may
start. This information is useful when more than one job is to be run and data collected and stored in the Fl LE
by one job is to be used by another job. Different jobs usually vary in size and therefore the start of the FI LE
area will be different. Therefore, to run two related jobs where the FI LE is used, the start of the file must be
changed for the shorter job to eliminate the conflict. The start of the file can be changed under the control of
the Executive Command Decoder in the same way the system level variables printed by the compiler can be
changed.

6-1

After the Executive is called into operation, indicated by SPUT typing EXEC LOADED, system level parameters
may be altered and/or the job can be started via the Executive Command Decoder. To interact via the Command
Decoder, a Teletype must be in the command mode. Any Teletype may be a command Teletype. To attach any
Teletype to command mode, type CTR LIA. If an tA is echoed, the Teletype is now in command mode. A Teletype is released from command mode by typing CTRL/P and the RETURN key.
Commands to the Command Decoder are terminated by typing the RETURN key. A RUBOUT deletes the last
character in the current line and echoes that character; when no character is echoed, no characters remain in the
line. An entire line may be deleted by typing CTR LIU.
The Command Decoder has two basic operations: value modification and run commands. Value modification,
initiated by CTRL/V, executes interlaced; that is, phases are allowed to execute when the Command Decoder is
not processing commands. System level variables, including the Fl LE START parameters, may be displayed and
modified while the system continues to execute.
The run command processor, initiated by CTR LID, can flag phases for running. A run command, in the form
tD R#n, is used to flag the phase n, where n is the phase tag, for execution.
When the user types the RETURN key, the system begins to execute the flagged phase and will release the command Teletype to the system output mode (refer to Appendix H for a list of the available commands and their
meanings).
Such a series of commands appears as follows on the Teletype:
tA
*tDR#1)

(CTRL/A)

Typing CTRL/V indicates to the Command Decoder that some value changes are going to be made to the system
level variables. There are four number systems available for examining and modifying the contents of a particular
variable in memory.
Number System

Abbreviation

Variable Type

clock

System Clock

octal

Integer

decimal

n
LJ

Integer

exponential (floating-point)

Real

It may be helpful to remember that for access to memory you have a memory "CODE". The combination of
floating-point and integer arithmetic provides increased data acquisition capability. For an integer variable, one
whose name starts with an I, such as IMIN or ISEC, decimal or octal numbers are used for inspection. For a real
variable, one whose name does not start with I, such as VPR, exponential numbers must be used. To examine a
particular storage variable, type the abbreviation for the number system to be used and the storage location number of the variable (the number after the variable's name in the storage table printed by COMP). For example:
*tVD 3

+0003\+0000,)
*E17

+0017\+0.350000E-0002,
The Executive types the storage location number, a back slash, its present value, and a comma.

6-2

To change the present value in a location, simply type in the new value after the comma is printed. A carriage
return typed after the new value closes this location, leaving it with the new value. A line feed after the new
value closes the location with the new value and then types the contents of the next location. If the present
value in a particular location is not to be changed, but the user wants to check the next storage location, a line
feed will cause the next location and its contents to be printed out in the same format and same number system
as above. The value of the new location can then be changed after the comma is printed.
NOTE
The FILE START parameters are located at D-1,
D-2 and 0-3; they reflect the respective base of
FILE addresses for FILE #1, #2, and #3. These addresses refer to the disk page address beginning at 0
for disk 0. The locations may be called in their appropriate number system(s) in any order.
For exponential notation, a number may be entered with or without the power of 10 portion. For example:
*E21
(line feed)

+0021/+0.230000E-0002
+0024/+0.105000E-0001,

.01035

(line feed)

+0027 /+0. 780000E-0003,

73E-5

(carriage return)

The first location number, 21, was called by typing E21 and no changes were required, so a line feed was typed
to examine the next variable. Note that the next location number has been increased by a value of 3. This is
because each exponential (floating-point) number requires 3 memory locations for storage. Location 24 required
a change which was incorporated by typing in the new value in decimal point notation. The new number will
internally be changed to exponential notation. A line feed now causes another three locations to be advanced
(still using the exponential number system) and this time a variable was entered followed by a power of 10 exponent.
These variable changes are automatically performed while the system is running by starting the value change sequence with CTR LIV and then typing the appropriate number system abbreviation and storage location variable
after the response of a period.
Clock mode allows the user to inspect the value of the system clock and to preset to the time of day. Typing tC
will display the system clock in "HH: MM: SS,'' and open the clock for modification. The system clock is closed
by), or preset by XX:YY:ZZ).
*tVC
CLOCK 0: 1:35,12: 15:00
After the system is started, the keyboard is locked and will respond to the CTRL/A command only if the system
designer opens the console (command mode permitted). All of the possible CTR L commands are listed below:
CTRL/A

Attach this Teletype to command mode

CTRL/C

Return to Monitor (available only if not disabled by GENDAC)

CTRL/D

Initiate a "RUN" command

CTRL/V

Initiate a value change

CTRL/P

Return to operator mode

CTRL/U

Delete this line (start line over)

If the. same. program !s to be used aga!n, on!y SPUT need be ca!!ed because the program has already been compiled. Follow the same procedure for calling SPUT as already specified. Note that this may not apply if FI LE
is used.

6-3

CHAPTER 7
MODIFYING THE SYSTEM

7.1

INTRODUCTION

The INDAC Compiler {COMP) is unique among compilers in that the source-level language to be processed has
been defined in tables. While this feature alone is not exceptional, the fact that these compiler tables may be
modified by the user is unique.
The tables are called System Communication Tables, and they define, to the different programs in the INDAC
system, the location of code, the names of devices, the disk areas, and other system parameters.
It is the function of GENDAC to interact with these tables (under user direction) and to update certain disk
images to incorporate additions or modifications to the I NDAC system (see Figure 7-1 ). G END AC is a conversational program that accepts one or more formatted paper tapes called library tapes and/or accepts one or more
disk images developed by the user. The entire acceptance or rejection procedure is under operator control through
the Teletype keyboard.

7.2 SYSTEM COMMUNICATION TABLES
These tables provide the basic intercommunication link between the various component programs in the I NDAC
System. They are updated by GENDAC to contain all data relevant to the use of library routines within the system. These tables are described in the following paragraphs.

7.2.1

Intrinsic Functions (IF)

The (IF) table contains the vector locations of the intrinsic functions of the I NDAC language {the Executive routines that resolve the GOTO statement, the+ in arithmetic statements, the DO statement, etc). The (IF) table is
a by-product of the Executive Assembly and contains addresses generated by the assembler. In the I NDAC System, the user is permitted to update this table (using G END AC) to include certain core-resident functions in the
I NDAC Executive. Highly repetitive functions, such as bit shifting, thumbwheel input conversion, and output
digital formatting, can be made core-resident parts of the language to save the overhead time of continual disk
accessing.

7.2.2 External Subroutines (XS)
The (XS) table contains the alphabetic names and descriptions of all external (disk-resident) subroutines and
arithmetic functions. The INDAC Compiler uses this table to resolve such source-language statements as:
LET Y =SORT (X)
DO IRONC (INPUT, IOUT, REF)

7-1

SYSTEM PROGRAMS
LOADING

BASIC COMM
TABLES

MODES:

GENDAC

SYSTEM-ADD COMPILED SUBR (VECTOR CODE)
BINARY- RUN DEC SUPPLIED BIN IMAGE LIBRARY

USER ROUTINE

BIN CODE

USER ROUT I NE
L I BR A RY I B IN

COMPILE

DEC SUPPLIED

CREATED BY
ASSEMBLING
ROUTINES WRITTEN
IN PAL-I

INDAC SOURCE PROGRAMS
DUMP

COMPILED OBJECT

SYSTEM PUTTOGETHER

<PZ>
<CM>
<SD>
<XS>
I•OS
I/O
CONFIGURE
COMMTABLES
EXEC
DISK IMAGE

RUN TIME EXEC
SYST

FIELD 1

FIELD 0

HARDWARE
08-0657

Figure 7-1

System Communication Tables - Interaction

7.2.3 System Devices <SD>
The (SD) table contains the alphabetic names and definitions of all system hardware devices and pseudo-devices
(see Appendix I). The I NDAC Compiler uses this table to resolve all device requests such as the following:
GET (KEYS) IA
SEND (TTY, #101) ICLK
GET (CLOCK) ICLK
7 .2.4 Core Map (CM)
The (CM) table contains the core map of the Executive areas available for user insertions. This table is also generated through the assembly of the I NDAC Executive. This table contains, for each page of core, the first address
and the number of words available on that page.
7.2.5 Page Zero (PZ)
The <PZ> table contains pointers, parameters, and general data for use by the INDAC Executive, the INDAC Compiler, and the Utility Support Programs. This table is an external table and is mentioned only to aid the explanation of how the INDAC System functions. The (PZ) table is one page in length and becomes Page 0 of Field 0 in
the I NDAC Executive during runtime. This page is updated and supplied by the Compiler as part of each compiled INDAC program. During the loading of the program, the page 0 table overlays the INDAC Executive page
0 image. For this reason, no direct modification of the disk image of Executive page 0 is possible.

7-2

7.2.6 Special GENDAC Table(**)
G ENDAC creates a special table(**) for its processing that contains tags, fixups, and entry points to support the
"linking loader" function. This table is created for each separate run of G ENDAC and exists for that run alone.
7.3 UPDATING THE SOFTWARE
The I NDAC Compiler does not produce PDP-8 assembly-level code. The Compiler produces, instead, what we
shall call "vector" code.
While most compilers producing an interpretive code point to a location in a table that, in turn, gives the location
of the desired information in core; the INDAC Compiler's vector code points directly to the desired core location by using the (IF) table. Vector code is a highly compact form of information, allowing the I NDAC Executive
to receive its commands in an abbreviated form, and thereby reducing the time required to retrieve commands
stored on disk and the time required to "interpret" normal interpretive code.
The INDAC Executive contains routines to process arithmetic and logical operations, routing, 1/0, timing, and
other system functions. The Compiler supplies the vector code to call these routines into action, based on the
resolution of INDAC source-language statements. Two kinds of code appear in core while the Executive is operating:

a. Vector Code: The command-level code produced by the INDAC Compiler and used by the
Executive to run the system.

b. PAL-I Assembly Code: The operating-level code performing the functions and commands
specified by the vector code.
GENDAC can process both the vector code and the assembly-level code. Generally, vector code is a product of
the I NDAC Compiler and appears as a disk image; however, provision has been made to allow the user to produce
his own vector code and to direct GENDAC to include this code in the system. The provision also exists for a user
to create a new INDAC Executive function in the PAL-I code and to call this function into action through a new
(user-created) vector command.
To more fully understand the vector code that the Compiler generates for source-language statements, code a
simple INDAC program and run the Compiler under the List Option, as shown below:
.COMP
*OUT-S:TEST
*IN-S: TEST
*OPT-L
Paragraph 7.6, on Coding Details, contains the specifications and functional descriptions for each of the vector
commands.
7.3.1

System Mode of Updating

The I NDAC Compiler produces vector code as a normal part of its operation in compiling a program. Since
G ENDAC can process vector code, and since the Compiler can generate vector code, there ought to be some
method of combining both capabilities so that code generated by the Compiler could become defined to the
Compiler for future referencing. This would ensure that the next time the Compiler is run, the code previously
generated can be referenced in source langauge.
The provision to define previously compiled code to the Compiler is known as the System Mode of Updating and
can onl-y be done when addiny disk-resident subroutines. The procedures are summarized as follows.

7-3

Step

Procedure
Generate an INDAC program that consists of only one subroutine. This INDAC
program is then compiled, using the INDAC Compiler.

Run G END AC, using the "S" ("System") option.

Assign a particular name to the subroutine, give the current name of the compiled
subroutine, and other pertinent information as requested by G END AC via the
Teletype keyboard.

GENDAC updates the appropriate system tables with the new name and moves the
image of the subroutine to an area of the disk selected by a G ENDAC algorithm.

Refer to the processed subroutine code with the I NDAC statement:
DO aaa
where aaa is the name of the subroutine just processed.

7.3.2 Binary Mode of Updating
GENDAC operates in a second mode, termed the Binary Updating Mode. In this mode, GENDAC expects to
find a formatted binary image produced by the PAL-I Assembler either as binary paper tape(s) or as binary disk
image(s). In Binary Mode, the user can elect to manually simulate the INDAC Compiler output (vector code),
develop complete assembly language routines, or produce some hybrid code, making use of the conveniences of
both.
GENDAC requires information to update the System Communication Tables and disk images. One part of the
information is the actual code for the library routine. Additional information is needed to define:

a. The function of that code (subroutine, device handler, function)
b. The area within the Executive that the code may exist
c. The information that the code is to be considered a disk-resident function or subroutine external to the INDAC Executive image.
STRUCTURE OF A BINARY IMAGE
IDENTIFYING
iNFORiv1ATiON
USER
CODE
IDENTIFYING
INFORMATION
Because GENDAC sees only an assembled binary image, a technique for dividing and identifying code and information is required.
GENDAC requires a specific set of origin declarations (*76XX, *7777) and FI ELD declarations to separate and
identify the library routines to be processed. Because these declarations are assembled by PAL-I as distinctive
binary images, GENDAC is able to interpret them as declarations to separate and identify information and code.
These declarations, therefore, are not to be considered origins for code or data, they are merely identification
codes. (See Figure 7-2.)

7-4

PAPER TAPE
BINARY IMAGE
x x x
x x x

............................
x x x

xxx

xx x
xx x

x
x

----7--7'~
ORIGIN~

LEADER /TRAILER

FIELD

CHANGE
08-0671

Figure 7-2 Paper Tape Binary Image

7.4 LIBRARY STRUCTURE
G ENDAC Library Tapes may contain subroutines, functions, and 1/0 device handlers; the tapes may be in the
form of punched paper tape or disk binary files. Library tapes consist of blocks of code referred to as "modules."

7.4.1

Basic Module

A module is a disk-resident or core-resident code that is assembled and located as a single unit. To identify these
modules to G ENDAC, the "type" of module (core or disk, function, subroutine, or device handler, etc.) is specified and linkage information and other pertinent system data is provided. The following PAL-I Assembler pseudoops are used to provide information regarding modules to G ENDAC:

*76XX

FIELD X

*7777

Because this area in core is prohibited (Monitor Bootstrap), these origins are considered by G END AC to be descriptive information. The following origin numbers have English word equivalents (Equates) that are used in
coding the modules:

7600 =GROUP
7610 = OLDMOD
7620 = NEWMOD
7630 = CALMOD
7640 =ENTRY
7650 =CODE
7660 = FIXUP
7777 = ENDGRP
Disk-resident modules are used for subroutines and functions and must be one page of code or less to be loaded
by GENDAC. During Executive run-time, disk-resident modules are called into core when needed for execution
and are placed into one of two (foreground/background) system buffers. Each such module contains one or more
entry points defined in the (XS) table, for use by the compiler at compile time.
Core-resident modules are used for functions and device handlers. Each relocatable, core-resident module must
be one page of code or less if coded in PDP-8 machine code, or two pages or less if coded in I NDAC vector code
(all vector code must reside in Field 1).

7-5

7.4.2

Extension Modules

Each function or device handler can be constructed from several modules, if necessary. This set of functional
modules is considered a "functional group". In each case, one module contains the basic entry point(s) to the
routine as named in the (SD) or (XS) system table, while the remaining modules are "extension" types. The module containing the entry point must be loaded by GENDAC into the appropriate part of Field 0 according to its
purpose, that is, whether it is a function or device handler.
Any modules loaded, in addition to the initial module, as part of a core-resident function or device handler must
be loaded by GENDAC and may contain an "extension entry point," named in the G ENDAC (**)table, in order
that the modules may reference each other. These modules need not be loaded into prespecified core locations.
The user can specify the absolute location of any core-resident module or can allow G END AC to allocate any relocatable code according to its function and the current (CM) table.
The modules of an inter-related group are bracketed on a library tape between a *7600 and a *7777 code; the
entire group can be omitted (optionally) from any given GENDAC run. The *7600 is followed by a userformulated text comment, which GENDAC types as the tape is read. The operator determines whether the
group is to be loaded or bypassed. For example, consider the following code:
*7600
TEXT 'ADC'
When the above code is encountered on a library tape, G ENDAC recognizes that it has encountered another
library routine (*7600). GENDAC then types:
ADC?At operating time, the operator replies yes or no, depending on whether the routine is desired.
7.4.3 Call-Up Modules
If an extension module is common to each of several library routines, and the user desires to have GENDAC load
the module only if one or more of the routines requiring it has been selected for loading, then such an extension
module is defined as a "call-up" module. A definition of its allocation requirements, and a specially flagged entry
in(**) must then precede all modules that may reference it. The actual body of code (together with any necessary global definitions) must follow all referencing modules.
7.4.4 Additional Notes
To insert an entry into a system table (either (SO) or (IF)) pointing to code already present in the INDAC Executive, but omitted from the table, a pre-existing module may be declared with a *761x to avoid having G END AC
load the code again.
The special entry declaration * 7642 is used to extend the Executive 1/0 interrupt skip-chain. It may be used
only once in a module (any Field 0 module) to point to the first device skip IOT in the module. The last device
skip IOT, of one or more in the module, must be followed by a jump indirect through the first word of the moduie. G END AC inserts a connection address into the first word of an liO interrupt module, joining that module
into the interrupt skip-chain. The last module in the group points to a system error routine; thus, if no module
responds to the interrupt, the interrupt chain defaults to the system error routine.

7-6

(This page intentionally left blank.)

7-7

//EQUATES FOR GENDAC
GROUP=7600
OLDMOD=7610
l\EWtV;OD=7620
CALMOD=7630
ENTRY=7640
CODE=7650
FIXUP=7660
ENDGRP=7777
TEXT 'SPECIAL DEVICE HANDLER'
//FUNCTIONAL GROUP DECLARATION
*GROUP
TEXT 'MY DEVICE'
TEXT 'SPECIAL DEVICE 1~13

~OD

2•

//PHYSICAL DESCRIPTION OF MODULE
4
/REQUEST FOH I-0 MODULE ALLOCATION
*275
/ORIGIN WITHIN PAGE CDETERMINED FROM <CM> TABLE>
LAST-MYDFV+l
/SET UP LENGTH OF MODULE

*NEW~OD

//LOGICAL DESCRIPTION OF MODULE
1
/NU~BER OF ENTRY POINTS WITHIN MODULE
*ENTRY 3
/DEFINING AN I-0 HANDLER ENTRY POINT
TEXT 'MYDEVICE'
/NAME THAT CO~PILEH WILL RECOGNIZE
0
/RELATIVE ENTRY POINT WITH1N MODULE
2000
/8 IT PATTEkN FOR "GET .. DEV ICE
6361
/IOT CODE FOR DEVICE
0
/NO CONTROL WOkDS
//ACTUAL CODE OF MODULE
*CODE:

TEXT ' [ ~YJ'
*275

IY.YDEV,
DEV2,

CALL
SETUP
CALL
DATA
JMP

/INITIALIZE THE DATA kOUTINE
/GET THE ADDkESS OF THE DATA ITEM
NOf'v'.OR

I INTEGER RETUkN

SYS OUT
DCA
AD Dk

/kEQUEST FOR REAL VAkIABLE-EkkOk
/HOLD ADDkESS
/GET DATA FRO~ DEVICE
/STOkE DATA-~OTE DATA FIELD SET TO ONE ON ENTRY
/GET NEXT ADDkESS

6362

DCA

Jt'P

NO MOR,
LAST,

/END OF DATA LIST

SKP

CALL
DEVRET

ADDR
DEV2

/kETURN

FRO~

DEVICE HANDLER

//FUNCTIONAL GROUP TERMINATION
/END OF MODULE GROUP FOk
*ENDGRP

7-8

~YDEVICE

7.5 LIBRARY TAPE FORMAT
The GENDAC Library Tape Format is graphically represented by the following diagram. A typical library tape
comprises the elements indicated.
LIBRARY TAPE HEADER
MODULE HEADER
LIBRARY
TAPE

MODULE
MODULE HEADER
MODULE

7.5.1

Definitions

The library tape is coded with PAL-I origin definitions of 76xx for communication purposes. Because this area is
prohibited (Monitor Bootstrap), these origins are considered by G END AC to be descriptive information. The
equate statements listed in the Teletype example provide an English word equivalent that is more descriptive
than the origin number. Refer to the coding sample shown on the foldout in the subsequent discussion.

Gk C• l 1P

=7 6 :;:1 r'.

L!L D rv CD= 7 6 1 ~J
NE'!:'~/;OD=7 62Vl
CALFOD=763fi
FNTR Y =7 64(/1
CCD E 7 6 5Cil
FIXUP=7660
ENDGRP=7777

A PAL-I coded line such as:
*ENTRY 3
results in *7643 as an assembler output.
7.5.2 Library Tape Header
The first element of the GENDAC Library Tape Format is the title for the entire library tape. With the PAL-I
pseudo-op TEXT, the user creates a title for the tape as follows:
TEXT 'TITLE OF TAPE'

or 0
if no title is desired.

7-9

7.5.3 Module Header
The header data for modules is dealt with in subsequent paragraphs and consists of the following:
Functional Group Declaration .............. Defines a group of related modules (basic entry,
*GROUP
extension, and call-up, which can be selected or
bypassed).
Module Declaration
Physical Description ................. Declares the type of module and storage allocation requirements.
*OLDMOD Y
*NEWMOD Y
*CALMOD 2 or 3
Logical Description .................. Declares the purpose of the module, and conse* ENT RY
quently what system communication tables are
to be updated.
7.5.3.1

Functional Group Declaration - This data defines a group of modules, all of which may be either se-

lected for loading at GENDAC run-time or bypassed at the user's discretion. A functional group begins with a
*GROUP code and contains a description of the module to be printed at run-time. For example:

7-10

The *GROUP code indicates to G END AC that it has come to a new group. When the tape is read into the computer by GENDAC, the following is typed on the Teletype:
MY DEVICE?At this point, the user indicates whether or not the group is to be included in the system at this time. If the operator requests additional information about the group (by typing "I"), the Teletype output at this point appears
as follows:
MY DEVICE? - I
SPECIAL DEVICE 1013 MOD 2?
Now the user must answer either "Y" if the group is to be included or "N" if it is to be bypassed.
There may be more than one line of text to be printed out as a description of the group. This is done by supplying the material on subsequent lines with a TEXT pseudo-op for each line. All such descriptive material is printed
when and if "I" is typed by the operator.

7.5.3.2 Module Declaration - This data consists of two units of information:

a. A physical description of the module, defining the allocation requirements.
b. A logical description of entry point(s) within the module. There may be one or more such logical
descriptions.

7.5.3.2.1 Physical Descriptions of Module - The first information to be placed after the Functional Group Declaration is one of the following module definition codes.
Code

Definition

*OLDMOD Y

The module is already core-resident, and Y cannot be 0.

*NEWMOD Y

The module is being supplied for addition to the INDAC System. The module may be a basic entry module or an extension
module.

*CALMOD Y

A call-up module (where Y=2 or 3) is being defined at this
point. This call-up module is some form of common routine
and is to be loaded only if a request is subsequently made for
this module.

The "Y" in the module definition codes above defines the allocation requirements of the module and is described
inTable7-1.
The allocation code is followed by a set of parameters that depend on the chosen value of Y. Table 7-2 lists the
various parameter codes to be used. The user should insert the codes in the order listed, depending on whether
the type of module being handled belongs to the module type listed (the value in the allocation code). If the code
is required, the correct sequence must be observed.

7-11

Table 7-1
Allocation Codes

Y Code

Resident Area

Relocatability

Additional Information

Disk

External subroutine or function

Core

Nonrelocatable

User specifies absolute origin

Core

By page

Core

By word

Core

By page

1/0 handler only

Core

By page

Function

Core

By word

Function

User specifies page origin

Illegal

Table 7-2
Allocation Parameters
Module Types

Actual Code Used

1,2,3

FIELD X

Reason for Code
Where x may be 0 or 1. FIELD 1 is required for any module
written in vector code in an 8K system.
FIELD 0 is required for any module containing a *7642
interrupt-chain extension.

2,4,6

*XXXX

Where XXXX is the absolute core origin of the module. (XXXX
may not be in page 0 or of the form 76XX or 77XX.)

*OXXX

Where *OXXX is the origin within the page. All page-relocatable
code must be assembled in page 200 - 377.
All word-relocatable code must be assembled in pages 200 - 577,
and *200 should begin the body of code.

3,7
All cases
except 0

xx xx

Where XXXX is the size of the module in words. There is a maximum of one page for absolute, call-up, and page-relocatable
modules; a maximum of two pages for word-relocatable modules
(except ca!!-up); exactly one page is assumed for disk-resident
modules.

In the following example, the physical description of the module, the origin within the page is arbitrarily selected
as *275. In practice, the ESUP (Executive Support) program would be run against the current core map (CM)
and an address chosen based on the available core.

7-12

TEXT •MY OEVlCE•
•sPE,CIAL DEVICE 1013 MOO 2•

TE:XT

//PHYSICAL DESCRIPTION OF MODULE
/REQUFST FOR I -0 rv;QDULE ALLOCATION
•275
/ORIGIN WITHIN PAGE CDETER~INED FROM
LAST-MYDEV+l
/SET UP LENGTH OF MODULE

*NElvMOD Lt

<CM>

TARLE>

NOTE
The size of the module does not have to be counted.
Because this code will be processed by the PAL-I
Assembler, it is sufficient to generate an expression
for the length of code and let the assembler generate
the actual number.
7.5.3.2.2

Logical Description of Module - The next item in the module declaration is the collection of codes

that specify the purpose of the module, and correspondingly specify what system tables must be updated to refer
to points within this module.
Table 7-3 is a summary of the table update information, similar to Table 7-2. Each module to be considered uses
all or part of this information for each entry point. If used, it must be in the sequence indicated.
Table 7-3
Table Update Information
Code

Significance

xx xx

XXXX is the number of entry points to be defined within this module, or 0 if
none are defined. Each entry point definition is preceded by a *ENTRY X
code.

*ENTRY X

Where X defines the system table(s) to be updated, and the nature of the table
entry.
Xis one of the following codes:
0

Extension module or Extension Entry point names in(**) for use
in global definitions only.

Call-up module, used if this is a *CALMOD X call-up module. Limit
is one per module, also uses(**).

Device handler interrupt-chain extension. Uses no system table,
word 0 of module code is filled in by G END AC.

Device handler, called by source program, uses (SO).

Core-resident function, uses (IF).

Disk-resident function, uses <XS).

Disk-resident subroutine, uses <XS).

7-13

TEXT 'NAME OF ENTRY POINT'
For ENTRY 3, 4, 5 and 6, TEXT "name" specifies the name to be used in the
INDAC source program.

ox xx

Where OXXX is the location of the entry point relative to the base of the module.
(If the module is of the type *ENTRY 2, then OXXX points to the fast IOT of
the skip-chain segment in the module.)

When the module is of Type *ENTRY 3, a device handler called by the source program, the following additional
data is required:

xx xx

Where XXXX is the bit pattern defining the device usage to the INDAC Compiler,
bits may be ORed to give a combination of the following meanings:
4000

Allocated buffer required

2000

"GET" device

1000

"SEND" device

0400

Multiplexed device with channel assignments

0200

May reference a FORMAT statement

6XX1

The device code.

ooxx

Where XX is the number of device control words to follow, or 0 if there are no
control words:

TEXT 'control word'
Where the control word is the text to appear in a control line statement, for example: X1000, GAIN 2, etc.

xx xx

Where XXX is the octal value for the control word just defined. (The TEXT and
XXXX codes are repeated until the number of control words specified above are
completed.)

The logical description of the sample 1/0 handler follows:

DES CR IPT IC:''! OF MOOLILF~
/h·'l'f1 P i:h: f';F E :-.~Tk Y PC• I >ff S :,; i TH i ;--.; r·ICD ULF
ID E F I 1\1 I Nfi A f\' I - ('. HA 1\10 L Fr< F i'\ T k Y P 0 I NT
rr.xr 'i.'Yf)FVJcr.•
/f\!M•·F THAT cc~:PILF:f\ '!ILL j-,FCCG\,]J/F
0
/kFLATIVE ~NTkY POINT WITHIN MODULE
Y c-1 vH'l
I B I T P tn TE k !\: F 0 k " GET" DEV I C E
63~1

/IDT CODE FCR DEVICE

7.5.3.3 Module Body -When the header information required by GENDAC to define the module is complete,
the code of the module itself can be included.

7-14

7.5.3.3.1

Module Title -

*CODEX

Where x has the following meaning:

code of standard (*NEVv'MOD X) module follows
code of call-up (*CALMOD X) module follows

TEXT 'module name'

Gives the name of the module. If a call-up module has been
referenced by a *F IXUP X prior to this point, the code for the
module is loaded, if the module has not been referenced, GENDAC
skips to the next * ENDG RP separator and continues from there.

Details on preparing the code for a module are given in Chapter 3.
7.5.3.3.2 Fixup Declarations - GENDAC allows the use of relocatable code for modules. Such code cannot be
completely defined at assembly time, because the final location of the code is unknown. In this sense, GENDAC
operates as a linking loader, providing links within and between modules.
The basic link mechanism is the "FIXUP". A FIXUP is a declaration made before the actual body of code as
follows:
*FIXUP X

Where Xis the index to the following FIXUP definition TEXT 'AAAA'

TEXT 'AAAA'

Where AAAA is the name of an entry point of this or a previously
loaded module and may take the following forms:
AAAA

Absolute reference

AAAA-.

Displacement reference

-AAAA

Negative absolute reference

-AAAA-.

Negatived isplacement reference

0-.

Null displacement FIXUP definition

Repeat this declaration for each F IXUP needed in this module.
G ENDAC saves the text declaration stated and assigns a number to the saved text according to *FI XUP X.
This saved text, with its assigned number (X), is now a FIXUP. FIXUPs are called for within the actual code of
the module as follows:
FIELD X
word
At the Field call, GE NDAC uses the assigned FIE LO as an index to retrieve the saved FI XU P text. The saved
text is evaluated at the location in core, where the word following the FIELD call will be placed. The evaluated
text is arithmetically added to the word following the FIELD call (normally the following word is zero). In most
cases, the absolute reference format is used. The special cases are provided for generating relocatable, vector
displacement code.
FIXUP declarations are limited to 8 (0-7) for each single module. A new module can use the same FIXUP index
numbers again, because all saved text is deleted at the end of each module. The actual code of the sample 1/0
handler follows:

7-15

INO CONTROL WORJ.lS

//ACTUAL CODE: OF MODULE
*CODE

TEXT • [ rv; Y J •
*275

f'l.YDEV,

CALL

DEV2,

CALL
DATA

/INITIALIZE THE DATA kOlJTINE

SET UP

JfVP

/GET THE ADDRESS OF THE DATA
!\!Ot1~0R

/END OF DATA LIST

SKP

/INTEGER

SYS OUT

/kEGUE~T

DCA

ADDR

6362

DCA 1
Jt.".P

!'!Or" OR,

CALL

LAST,

DEVRET

kETUt~N

FOR REAL VARIABLE-EkkOR
/HOLD ADDRESS
/GET DATA FKOM DEVICE

ADDR

/STORE DATA-NOTE DATA

DEV2

/ONE ON ENTkY
/GET NEXT ADDRESS

//FUNCildNA[

7.5.3.4 Group and Module Termination -When the module is complete, the separating *ENDGRP code (*7777)
should be used in the following situations:

a. To terminate function groups of modules, and
b. To terminate a call-up module.

//FUNCTIONAL GROUP TERMINATIOf\I
*ENDGRP

/END OF MODULE GROUP FOk MYDEVICE

7.6 CODING DETAILS
To illustrate the coding, application, and differences between subroutines and functions, a common industrial
processing routine was selected and implemented both as a subroutine and as a function. The routine is a bitshifting operation fixed at 4 bits (BCD) to the right. This limitation is imposed because a function can only

7-16

accept a single argument; the subroutine can accept several arguments and could be made more general. The
structure of a library module is described in Paragraph 7.4.
7.6.1

Function (Core-Resident in Field 0)

The function is called by the language statement:
LET la= BCDSHF (e)
where BCDSHF may be isolated as a call or may be imbedded in an expression itself. The purpose of BCDSHF is
to take the last value on the arithmetic stack, perform the shifting, and replace the value on the stack. In this
sense it functions as a unary operator.
A core-resident function is entered in binary mode by the Executive. When the function completes its assignment,
the function must return to the Executive. The actual code to perform the shifting is as follows:
TAD I STACK

/get the value on the stack

CLL RTR
RTR

/perform the shifting

DCA I STACK

/replace the value on the stack

VECTOR

/return to the user program in Vector mode

The ESUP program provides a paper tape of PAL-I Equate Statements to resolve the STACK and VECTOR terms.
This tape must be loaded with the function module so that the terms can be resolved during assembly. The Executive enters the function routine with the data field set to Field 1. The arithmetic stack is located in Field 1;
therefore, the indirect instructions reference the correct data. If any data is to be referenced, indirectly, in F leld
0, then the field must be changed and later reset to 1 before exiting from the routine. The complete routine
embedded in GENDAC descriptive information follows:

TFXT
*GROUP

'SPECIAL FllNCTION'

TEXT 'RCDSHF'
TFXT 'BCD RIGHT-SHIFT FUNCTION CC-R>'
*NE'IJIVOD 7
/FUNCT ION-ll/OkD kELOC
5
/SIZE OF MODULE

4
TEXT 'BCDSHF'

*ENTRY
0

/ONE ENTRY POINT
/FUNCTION, C-R
/NAME COMPILEk WILL RECOGNIZE
/j;°'l\iTkY POl!'-1T ~0ITHIN IVODULE

*CODE
:+:200

/Y0RD-kFLOC OkIGIN

TAD I STACK

/GET VALllF FkN': STACK

CLL RTR
RTR
DCA

/PEkFOkM bCD SHIFT
I STACf<

VECTCF<
*n-JDGF;P

/kEPLACF VALUE ON STACK
/kET!IRN TO

VECTOk

IVODE

NOTE
This code is word-relocatable (unusual for PAL-level
code), because all instructions are either page 0 references or nonmemory instructions.

7-17

7.6.2 Subroutine
The subroutine is called by the language statement:
DO BCDSHF (IX, IY)
where IX is the input item, and IY is the output item.
7.6.2.1 Subroutine Requirements and Analysis - The following is a list of subroutine requirements:

a. The first two words of the subroutine must be Os. These words are used by the Executive to
control the dynamic allocation scheme.
b. A subroutine cannot exceed one page~

c. A subroutine is entered in VECTOR mode and, therefore, must be located in FIELD 1. The
subroutine is d isk·resident and is automatically called into FIELD 1 by the Executive.
d. A function deals with information supplied by the arithmetic stack. A subroutine deals with
data (generally located in FIELD 1); therefore, a subroutine must include some instructions
to retrieve and update this data.

For this reason, a subroutine usually contains some initial Vector code to retrieve the data specified as the argu·
ments of the subroutine call. There is also generally some final Vector code that updates the output data items
of the call. As described in Paragraph 7.6.1, the ESUP paper tape provides the equate statements required to re·
solve the Vector code and must be loaded with the subroutine module for assembly.
7.6.2.2 Sample of Subroutine Coding
The following is an example of subroutine coding.

7·18

/SAMPLE SUBROUTINE
/DO BCDSHF<JX,,

JY)

//EQUATES FOR GENDAC
GROUP =7 60(_,
OLD MOD =7 6 ll1
NEWMOD=7 620

ENTRY=? 64~'
CODE=7650

FIXUP=7 660
ENDGRP=7777

TEXT 'BCDSHF SUBROUTINE TAPE'

*GROUP

TEXT 'BCDSHF'
TEXT 'BCD SHIFT SUBROUTINE'
/NEW-MOlliJLE,DISK-RESIDENT

*NFWMOD

/ONE ENTRY POINT
/DISK-RESIDENT SUBROUTINE
TEXT 'BCDSHF'
2
/MODULE ENTRY POINT <MUST HAVE TWO WOKDS OF 2.EhOS >
TFXT 'BCDSHF'
1

*FNTRY 6
*CODE

•2f'J0
fl
0

ILE
1

/TWO LEADING WORDS OF ZEROS REQUIRED
/FOR ALL EXTEkNAL SlJbKClUTINFS Ok FUNCTIGNS
/GET THE VALUE FOR THE I" I kS T Al-\GU!VJENT OF THC:
/CALL ON THE AkITHMETIC ~TACK

IS
ALPHA-.
/STORE THE VALUE INTERNAL TO THE SUBKOUTINE
GBIN
/ENTER BINARY MODE
TAD ALPHA
CLL kTR
!HR
DCA ALPHA /PERFORM THE 8DC SHIFT

INT ERP
IL
ALPHA-.
ISE
2

RETE
2

ALPHA,

/KETURN TO THE VECTOR MODE
/LOAD THE ADJUSTED VALUE ON THE
/AkITHMETIC STACK
/STORE THE VALUE ON THE STACK INTO
/THF SECOND AkGUMENT OF THE CALL
/RETURN FROM THE SUBROUTINE, RYPASSING
/THE 2 ARGUMENTS OF THE CALL

*ENDGRP
$

7.6.3 1/0 Handler
7.6.3.1

1/0 Handler Operation - The compiler generates the following vector code for an 1/0 request:
where XXX is an index to the handler
5XXX
)control driver pointer
)format driver pointer
indicator
4000 - GET; 0000 - SEND
nr. data list words
data list

7-19

7.6.3.1.1 The Control Driver Table - The control driver table contains the information supplied through the
.EOU IPM ENT statement device-control line, and the starting channel of each item in the data list. The Executive
contains subroutines available to the user for extracting information from this table automatically.

7.6.3.1.2 The Format Driver Table - The format driver table contains information supplied through a .FORMAT
or .HEADER statement. The processing of this table is reserved for the Teletypes and high-speed punch. This
table is not available to the user.

7.6.3.1.3 Data Information and Calls - The remainder of the 1/0 call is processed by Executive subroutines
through the "CALL" statements as shown in Table 7-4.
NOTE
The AC must be cleared before issuing a CALL.
Table 7-4
CALL Statements
ESUP CALL

Function

CALL
SETUP

The first CALL in an 1/0 handler provides the initial setup and linkage to
the 1/0 tables. Returns with the LINK set to 1 for a GET request and 0
for a SE ND request. AC set to 0.

CALL
ALO CAT

Returns with the base address of a page of FIELD 1 core. Used to obtain a
page of buffer for dynamically allocated core. Not generally used.

CALL
DATA

Used to obtain addresses from the data list. Returns with the address of
the data item in the AC at:

a
b
c

a. no more data address
b. request for integer variable

c. request for real variable
If the data item specified is an array, then consecutive DATA requests return with the consecutive addresses of the array requested, until the array
is expired. The DATA routine then looks for additional data list items,
until the list is expired.

CALL
IOSUB

a
b

Request for the subroutine (if any) specified in the device control line to
operate on the data supplied. A return is made at:

a. subroutine did execute
b. no subroutine requested

CALL
CHANEL

Returns with the channel number in the AC. This is the channel associated
with the last address supplied by the DATA routine.

CALL
DEVRET

The 1/0 handler is finished. This is a return to the Executive in vector
mode to continue processing the compiled program.

7 .6.3.1.4 Executive Page Zero Parameters - Table 7-5 is a list of executive parameters.

7-20

Table 7-5
Executive Parameters

Field Location

ESUP Name

Use

ADDR

Loaded with the address supplied by DATA for use by the subroutine
called through IOSUB.

VALUE

Loaded by the 1/0 handler with information derived from, or being
sent to the 1/0 device. The processing subroutine (if any) loads
VALUE from the data list (after processing) or takes VALUE, processes it, and stores it into the data list. The method of operation depends on whether the device is an input or output peripheral.

IOCTRL

After SETUP, this parameter contains the bit configuration, specified
by the device control line, required to properly service the device
(gain control, resolution, initializing requests, etc.).

STATUS

This parameter is available for 1/0 devices to communicate with
language-level code. The parameter is retrievable by
GET (STATUS) IX
where IX contains whatever value STATUS was set to.

7.6.3.1.5 Interrupting Devices - A complete cross-reference listing of devices supported by DEC is included in
the G ENDAC library. For an analysis of interrupt processing, see the ADC handler included in the program listing (tag "ADCI" to "AFX = .").
7.6.3.2 A Typical 1/0 Handler - A typical 1/0 handler (D !GOUT) is presented below, showing the use of the

1/0 calls and parameters:
//DIGITAL

!ii3 5(?

C13 5f1
L!LJ65

(1)351

C,,0ill I

0352
0353
(t)3 54
\,1355
0356
~)3~7

Vl360
0361

0362
0363
036L1
('1365
0)366
C13 6 7
(13 70

L!-465

DIGO,

D IG02,

0002

537-4
7410

OUTPUT DEVICE HANDLER-SYS TH1 LEVEL

*35\Zi
CALL
SETUP
CALL
DATA
Jf'.~P

/LINK
/

/GET ADDRESS ANlJ TYPE OF
MOkE DATA ITEMS

ITEIVi

!JTG(iLJ

/[\Q

ADDR

/ERkOk-kFAL VAKIAHLE KE:uu~::sT
/SAVE AD Dk FOk SUBkOUTINE CALL

SKP

4510

S YSOUT

311 7
4465
(;)('.i04
5365

!JCA

CALL
IOSUEl
JMP

DIG03
ADDR
VALi iF

Vl\/WJ3

TAD I
DCA I
CALL
CHANEL

6366
73f,i0

6366
CLA CLL

1 51 7
3461

4Lt65

BIT INDICATES DikECTION

OF TRANSFER:"0"-0UTPUT;"l"-INPUT

DI (:i C13,

/CALL SUBkOUTINE-IF ANY RE8UESTED
/SUBf'<OLiT I NE CALLED-AND EXECUTED
/SUBROUTINE DID NC1T ~XECUTE,
I HAf\!DLE'K MUST STORE THE VALUE
/GET CHANNEL ASSIGNED TO ITEtv:
/SELECT CHANNEL

r13 7 1

1 Lj 6 1

0.3 72

6365

13 73

5352

TAD I
VALUE
6365
/5[t'-'D VALUE T (J DIGITAL OU1PUT
/t'.'CH.:: THIS DEVICE CL.EARS THE AC AF1Eh SENDING THF IT E.fv,;
I THE AC tl,LJ~T BE CLEAR BEFOkE ISSUING A "CALL"
J\w,p
/GET NEXT DATA ITEiii
DIG02

33 7 Lj

Lj Lj 6 ~

DJ(;(IL1,

03 7 5

(/) l~ Vi'::>

CALL
DEVkET

/DEVICE

7-21

KETUkN

EXIT.

7.7 VECTOR CODE
All vectors are processed by the vector interpreter of the Executive and result in the execution of some intrinsic
function or library routine of the Executive. The user, coding in assembly-level code, can make use of these routines by entering the vector mode and using these routines himself, perhaps in some combination not available
through the Compiler. The following paragraphs describe the characteristics of vector code.

7.7.1 Numeric Formats
The internal format of numeric data is shown in Figure 7-3.
0

Isl

SIGNED
BINARY NUMBER

INTEGER FORMAT
11

WORO I

WORD

EXPONENT

21 s

I
11

WORD •

I
FLOATING

REAL
VARIABLE
MANTISSA

POINT FORMAT
08-0670

Figure 7-3 Internal Format of Numeric Data
Integer numbers can be positive or negative (stored as 2's complement) according to the sign (bit 0). Floatingpoint numbers, called "real numbers", can be positive or negative according to their sign and are each assigned
three words of storage. Floating-point numbers are carried in normalized form. The exponent is a signed 2's
complement quantity in one 12-bit word. The signed mantissa is stored in two 12-bit words, maintaining 23 bits
of significance, making a total of three words of storage.
Arithmetic expressions are evaluated using a push-down arithmetic stack that grows in ascending locations (in
Fl ELD 1 of an BK system) with the current stack pointer (STK), pointing to word 3 of a real number or to the
integer number.

7.7.2 Vector Code Example
An example of the vector code produced by the compiler to resolve the INDAC statement follows:

7-22

LET IA = I B + IC -ID
IL
)IB

!L
)IC
ADD
IL
)ID
SUB
IS
)IA
The term ")IB" means the displacement address from the point at which )IB occurs to the actual location of 18.
In the PAL-I assembly, this term would be written as "18-.". The code as written above means the following:

1. Load the integer, the displacement address of which follows, [)IB], onto the arithmetic stack.

2. Load the integer, the displacement address of which follows, [)IC], onto the arithmetic stack.
3. Take the last two values on the arithmetic stack and perform an integer addition. Replace these
two values with the new sum (a single value).

4. Load the integer, the displacement address of which follows [)ID], onto the arithmetic stack.

5. Take the last two values on the arithmetic stack (the result of adding 18 and IC, and the value
of ID) and perform an integer subtraction. Replace these two values with the new value.

6. Take the integer value on the arithmetic stack and store it in the location the displacement
address of which follows [)IA]. This last action collapses the arithmetic stack.
GENDAC provides for displacement fixups to allow vector code modules to interact. Displacement fixups may
not be used in disk-resident modules, because a disk-resident module cannot reference another module. For the
same reason, fixups cannot refer to a disk-resident module.
7.7.3 Load and Store Instructions
Load and store instructions are divided into categories:

a. Single-word operations ( 1-word)
b. "Real" variable operations (3-word)
Single-word operations are considered integer types. Three-word operations are considered floating types. Operations for integer and floating-point instructions take the same argument list, and deal with the arithmetic stack in
the same fashion, with the single exception of the number of words handled.
7. 7 .3. 1 Load a Simple Variable FL
)ALPHA

Form:

IL
)18

Function:

Load the simple variable, whose displacement address follows,
on the arithmetic stack.

7-23

7.7.3.2 Store a Simple Variable Form:

IS
)IB

FS
)ALPHA

Function:

Store the last value on the arithmetic stack into the variable
whose displacement address follows.

7.7.3.3 Load An Array Element (Indexed Variable) Form:

Function:

I LI
)IX
)IB

FLI
)IX
)ALPHA

Load the arithmetic stack with the variable whose index value
(displacement address) and base address (displacement) follow.

7.7.3.4 Store An Array Element Form:

ISi
)IX
)IB

FSI
)IX
)ALPHA

Function:

Store the last value on the arithmetic stack into the variable
whose index value (displacement) and base address (displacement) follow.

7.7.3.5 Load An External ArgumentForm:

ILE
k

FLE
k

Function:

This vector is used within an external subroutine to load the
arithmetic stack with the kth argument supplied by the subroutine call. Arguments range from 1 to n.

7.7.3.6 Store Into An External Argument-

7.7.3.7

Form:

ISE
k

FSE
k

Function:

Store the last value on the arithmetic stack into the kth argument supplied by the subroutine call.

Load An External Argument, Indexed Form:

ILEI
)IX
k

FLEI
)IX
k

Function:

Load the arithmetic stack with the kth argument supplied by
the subroutine call, indexed by the internal subroutine variable
whose address (displacement) is supplied.

7-24

7.7.3.8 Store Into An External Argument, Indexed Form:

ISEI
)IX
k

FSEI
)IX
k

Function:

Store the last value on the arithmetic stack into the kth argument supplied by the subroutine call, indexed by the internal
subroutine variable whose address (displacement) is supplied.

7.7.4 GOTO Statements
7.7.4.1 Unconditional GOTO GOTO I
Compiles into:

~
~
7.7.4.2 Computed GOTO GOTO (1 1, ... , In), i
Compiles into:
CGO
---,

)POINTER
)11

...
)In

,....

)i
7. 7 .5

IF Statement
IF (e 1) o (e 2 ) then S
compiles into:
evaluate expression e 1, and

leave answer in stack
then evaluate expression e2,

leaving that answer on stack

FIF

,,•'••

FI F or IF depending on type

)POINTER

if e 1 greater than e2

)POINTER

if e1 equal to e2

)POINTER

if e1 less than e2

CGO

II
II
II

!!••

generates GOTO statement according to whether S is unconditior)al

GOTO

GOTO or computed GOTO

7-25

The pointers are set to point to the appended GOTO statement if the relationship is true or to the next source
statement if the relationship is false.

7.7.6 DO Statements
7.7.6.1

Internal Subroutine Call DO I
Compiles into:

fGOsBl

u=J

7.7.6.2 External Subroutine Call -

Compiles into:
4xxx

xxx - index into
call driver table

)V1

NOTE: The Call Driver table
is NOT available to G ENDAC
pointers to the base arguments

7.7.7 Return Statement
The RETURN statement compiles into one of two forms, depending on whether it is an internal or external subroutine return.
An internal subroutine has the form,
RETURN
and compiles into:
RET
An external subroutine return has the form,
Returns
and compiles into:
RETE
*count
where the count is the number of arguments over which to jump in the return.

7-26

7.7.8 Loop Statements
7.7.8.1

FOR Statement FOR i = m 1 TO m 2 STEP m 3
Compiles into:
IL

)m,
IS
)i

The parameters m 2 and m 3 are used in the NEXT statement. If Step m 3 does not appear in
the statement, m 3 is assigned the value 1.
7.7.8.2 NEXT Statement NEXT i
Compiles into:
NEXT
)i

)POINTER
where m 3 and m 2 were defined in the corresponding FOR statement. The pointer is a displacement address to the cell following the corresponding FOR statement.
7.7.9 Arithmetic and Logical Operators
This class of vectors has no argument following the vector. The routines supporting the vectors deal only with
the arithmetic stack. Within this class there are two types of operations.

a. Binary - operate on the last two values of the stack, replacing those values with the single result
of the operation.

b. Unary - operate on the single, last value of the stack, replacing that value with the result of the
operation.
7.7.9.1

Binary Vectors - Table 7-6 is a listing of the binary vectors.
Table 7-6
Binary Vectors
Integer

Floating

MPV

FMPY

Multiply

DIV

FDIV

Divide

EXP

FEXP

Exponentiation

Function

(continued on next page)

7-27

Table 7-6 (Cont)
Binary Vectors
Integer

Floating

Function

ADD

FADD

Addition

SUB

FSUB

Subtraction

12-bit Boolean OR

LAND

12-bit Boolean AND

7.7.9.2 Unary Vectors - Table 7-7 is a listing of the unary vectors.
Table 7-7
Unary Vectors
Integer

Floating

Function

NOT

12-bit logical negation

NEG

FNEG

Arithmetical negation

FLT

FIX

Integer to floatingpoint conversion
Floating-point to integer conversion (truncate)

7.7.10 Special Vectors
The vectors described to this point are all normal compiler generations produced in response to some languagelevel statement. In this paragraph, two vectors are introduced that are not generated by the Compiler; however,
they are available for PAL-I coding. Remember that all vectors must appear in FIE LO 1 and are processed with
the data field set to 1.
ESUP Name

Purpose

GBIN

Go-Binary: exit from interpretive mode, enter machine
mode at the next instruction in FIE LO 1 with data field set
to 1.

XTRP
xxxx

Exit from interpretive mode, enter machine mode in FIE LO
0 at the instruction whose address follows the XTRP command. Data field set to 1.

7.7.11 Special PAL-I Code
Occasionally, the vectors already described are not sufficient to service user needs, and routine must be created
to directly interact with the Executive. The following "macro-calls" are provided as equate statements through
ESUP to ease communication with the Executive. The commands in Tabie 7-8 are aii machine-mode instructions.
Another "macro-call" is available to allow communication for 1/0 handlers (FIELD 0). The call is "CALL" followed by an argument. Each argument references a subroutine within the Executive that supplies either information or a service. A complete discussion of "CALL" and the available arguments is presented in Paragraph 7.6.3
"1/0 Handler".

7-28

Table 7-8
ESUP Commands
ESUP Name

Purpose

INTERP
xxxx

Command given in FIELD 1 with data field set to 1, to enter vector mode
at the next word. xxxx is interpreted as a vector.

GOT ERP
yyyy

Command given in FIELD 0 with data field set to 1, to enter vector mode
in FIELD 1 at the address specified by yyyy.

VECTOR

Command given in FIELD 0 for Executive to continue processing the user
program. This is normally the last instruction of any library routine or
core-resident function created by a user.

ABSGO

Command given in FIELD 0 with data field set to 1 to continue operation
of the user program (vector mode) at the address specified in location
0100 8 of FIELD 0. Normally used to simulate the operation of a GOTO
... or a conditional test.

SY SO UT

Command given in FIELD 0, data field irrelevant, to abort the current
SNAP. An error has been detected and no reasonable method of continuing is available.

INTRET

Used in the interrupt 1/0 handler to return to the AC and LINK restore
routine, exiting from interrupt mode.

7.8 RUNNING GENDAC
GENDAC, in conjunction with the DEC-supplied or user-written library tape, is a powerful system configurator
(see Chapters 2 and 4). Its main features are as follows:

a. Conversational operation - G END AC requests the option required. If the user is uncertain as to
the correct response, he strikes the question mark (?) key. G END AC then prints an explanation
to the user.
b. GENDAC links DEC-provided 1/0 handlers and math functions (sine, log, etc.) into the Executive.

c. GENDAC links user-developed 1/0 handlers, functions, and subroutines into the Executive. Subroutines can be written in INDAC or PAL languages.
d. G ENDAC places subroutines in the top of the last supplied disk or on the first disk (disk 0) if
available.

7-29

APPENDIX A
SUMMARY OF INDAC STATEMENTS

LEGEND:

a = print string
c = i SEC i MIN i HR I HH:MM (AT statement)

e = expression (logical or arithmetic)
i = integer variable
k =constant

I = statement label

m =control parameter

s = subroutine name
t =tag

u = device name
v = variable name
SCN = control names for device

A.1

EXECUTABLE STATEMENTS

ASSIGNMENT STATEMENTS
LET v = e

II.

CONTROL STATEMENTS

GOTO Statements
GOTO I
GOTO (1 1 , 12

•••

In), i

IF Statement
IF (e 1 ) o (e2 ) THENS
o = Relational Operator
S= GOTO I
or
S =GOTO (1 1 , 12 , . . . , In), i

DO Statements
DO I
DO S ( V 1 , V 2 , • • • , V n )
DO PHASE #t PRIORITY k
DO SNAP #t PRIORITY INTERRUPT
DO SNAP #t PRIORITY k

A-1

RETURN Statement
RETURN
RETURN s

Program Control Statements
EXIT
EXIT THEN DO SNAP #t PRIORITY k
EXIT THEN DO PHASE #t PRIORITY k

Loop Statement
FOR i = m 1 TO m 2 or FOR i = m 1 TO m 2 STEP m 3
NEXT i

TIMER Statement
TIMER (STOP, #t)
TIMER (START, #t)

Ill.

INPUT/OUTPUT STATEMENTS

GET (u, #t) list or GET (u) list

SEND (u, #t) list or SEND (u) list

IV. ACTIVITY STATEMENTS

#t 1 DO SNAP #t2 EVERY c PRIORITY k

#t 1 DO SNAP #t 2 DELAY c PRIORITY k

#t 1 DO PHASE #t2 DELAY c PRIORITY k

#t 1 DO SNAP #t2 AT c PRIORITY k

#t 1 DO PHASE #t 2 AT c PRIORITY k

A.2 NON-EXECUTABLE STATEMENTS

SPECiFiCATION STATEMENTS
1.

.EQUIPMENT
t*u (device code)
tCHAN (m,n) ....... (multiplexer channels)
t#t SCN 1 , SCN 2 , . . . SCNn (control information)

11.

ALLOCATION STATEMENTS

.STORAGE v1 , v2 , ..• , vn

#t .FORMAT

#t .HEADER

A-2

111. PROGRAM UNIT SEGMENTATION STATEMENTS

#t .PHASE

#t .SNAP

.SUBROUTINE s

IV. COMPILER DIRECTIVE STATEMENTS

.ACTION

.PROCESS

.END

A-3

APPENDIX B
EXAMPLE PROGRAM

This program shows the technique of making an array available to a subroutine .

. ARRAY PASSING EXAMPLE
.STORAGE II
#1

.PHASE

#10

DO SNAP #20 EVERY 30 SEC PRIORITY 2
.ACTION
TIMER (START, #10)

#20

.SNAP
.STORAGE IK (1, 10)/1,2,3,4,5,6,7,8,9,0/,A(1,3)

#21

.FORMAT
XX.XX

XX.XX

.PROCESS
LET I= II
DO SUB (IK, I, A)
SEND (TTY, #21) A(1), A(2), A(3)
EXIT
.SUBROUTINE SUB (IDUM, IDM, OUM)
.PROCESS
LET IA= IDM
LETIB=IA+3
FOR I =IA TO IB
LET DUM(I) = IDUM(I)
NEXT I
RETURN SUB
.END

B-1

APPENDIX C
USING THE DISK MONITOR SYSTEM

C.1

EDITOR COMMAND SUMMARY

Command

Format

Meaning

READ

Read incoming text and append to buffer until a form feed is encountered.

APPEND

Append incoming text to any already in the buffer until a form
feed is encountered.

LIST

List the entire buffer.

nL)
m,nL)
PROCEED

P)
nP)
m,nP)

List the line n.
List I ines m through n.
Proceed and output the entire contents of the buffer and return to
command mode.
Output linen, followed by a form feed.
Output lines m through n, followed by a form feed.

TRAILER

Punch four inches of trailer.

Punch the entire buffer and a form feed; kill the buffer and read
next page.

nN)

Repeat the above sequence n times.

Kill the buffer.

KILL
DELETE

INSERT

CHANGE

nD)
m,nD)

GET

Delete lines m through n.

Insert before line one all text until a form feed is encountered.

nl)

Insert before linen until a form feed is encountered.

nC)

Delete linen and replace it with any number of lines from the keyboard until a form feed is encountered.

m,nC)
MOVE

Delete linen.

m,n$kM)

Delete lines m through n, replace from keyboard as above until
form feed is encountered.
Move and insert lines m through n before line k.

Get and list the next line beginning with a tag.

nG)

Get and list the next line after linen which begins with a tag.

S)
nS)
m nC: '
'''1''""'¥

Search the entire buffer for the character specified (but not echoed)
after the carriage return; allow modification when found.
Search linen, as above, allow modification.
<:>llrn111 mnrlifir<>+inn
''I-··-··

C:o<:>.-rh 1111-~
linoc m +h.-r'll 1nh n

....,_Ul-11

Ill

loolll--::::ttl

1 1 1 - - 1 1 1 -.... l o o l - l l •

(continued on next page)

C-1

Command

Format

END Fl LE

C.2

)

Meaning
Process the entire file (perform enough NEXT commands to pass
the remaining input to the output file) and create an end-of-file
indication; legal only for output to the system device. If the lowspeed paper tape reader is used for input while performing an E
command, the paper tape reader will eventually run out of tape,
and at this point typing a form feed will allow the command to be
completed.

EDITOR KEY FUNCTION SUMMARY
Text mode: Enter the line in the text buffer.

(carriage return)

Command mode: Execute the command.
Text mode: Cancel the entire line of text and continue typing on
same line.

(back arrow)

Command mode: Cancel command.
\

Text mode: Delete from right to left one character for each rubout typed (is not in effect during a READ command).

(rubout)

Command mode: Delete entire command.
Text mode: End of input, return to command mode.

FORM FEED

(period)

Command mode: Current line counter used as argument alone or
in combination with+ or - and a number.

(slash)

Command mode: Value equal to number of last line in buffer and
used as argument.

(line feed)

Text mode: Used in SEARCH command to insert a CR/LF into
the line being searched.
Command mode: List the next line.

>
<

(right angle bracket)

Command mode: List the next line:

(left angle bracket)

Command mode: List the previous line.

(equal sign)

Command mode: Used in conjunction with . and I to obtain their
value (.=0027).

(colon)

Command mode: Lower case character, same function as=.

(tabulation)

Text mode: On output is interpreted as a tab-rubout combination.

C.3 ERROR MESSAGES
As an input command string is being typed, Monitor recognizes any incorrect syntax and responds with one of
the error messages in the following table.
Message

Meaning

Illegal syntax or miscellaneous error condition

Directory on the systems device is full

Too many inputs or outputs were entered

No such inputs

System 1/0 failure

APPENDIX D
INDAC COMPILER ERROR MESSAGES

Error Message

Meaning
Incorrect representation for statement TAG.

Incorrect representation for statement LABEL.

This statement is not syntactically complete. The compiler needs more information.

Unrecognizable statement verb.

While scanning this IF statement, the Compiler could not find or recognize a relational
operator (EQ, NE, GR, GE, LS, LE) or the THEN conjunctive.

The assignment variable (to the left of the equal sign) is not recognizable as a variable
name; or, this variable is a dimensional array and must be subscripted.

The expressions in an IF statement cannot be real and integer (or logical). They must
be of the same mode. A mixed mode of integer/logical is permitted, however.

The GOTO in this IF statement is missing or unrecognizable.

There is a subscripted variable in this statement whose subscript is bad. It is either
( 1) unrecognizable (2) a real constant or (3) the integer constant used as a subscript exceeds the declared size of the array.

The equal sign is missing or misplaced in this LET statement.

This arithmetic expression is not syntactically complete; the Compiler expects more information. Generally this means that an operand is missing following some operator.
Occasionally, this condition is caused by some previously detected error in the same
expression.

More than one unary minus in a row encountered while scanning this expression (for
example, --A is not permitted, while - (-A) is O.K.).

There is a constant in this statement that is improperly written.

Only integer constants are permitted as exponents or within exponential expressions.

Mixed mode expressions are not permitted. Real and integer (logical) variables cannot
be intermixed in one expression.

Illegal operator in arithmetic expression.

Illegal operator in logical expression.

22
23

Right parenthesis encountered adjacent to left parenthesis.
Extra left parenthesis (or parentheses) encountered.

Statement labels can only be used within the PROCESS section of a SNAP unit.

A logical expression cannot be equated to a real assignment variable.

Mixed mode expressions are not permitted. Arithmetic and logical operators cannot be
intermixed in the same expression.

Unrecognizable label encountered during processing of a computed GOTO statement.
{continued on next page;

D-1

Error Message

Meaning

Missing or misplaced comma within a computed GOTO statement.

Missing or misplaced right parenthesis within a computed GOTO statement.

The transfer variable in a computed GOTO statement may only be ( 1) an unsubscripted
integer variable or (2) an integer variable subscripted by an integer constant.

Priority verb missing, mispelled or not found in expected sequence. This error may be
due to untagged activity statement not in activity section.

Array sizes must be declared by integer constants only.

Syntactial error in EXIT statement.

Time declaration in AT statement must be positive integer constants of form: HH:MM
with HH less than 24, and MM less than 60.

Missing or unrecognizable DE LAY or EVERY in an activity statement.

The time declaration in an activity statement must consist of ( 1) an integer constant or
an unsubscripted integer variable or an integer variable subscripted by an integer constant (2) which must be followed by the contractions SEC, MIN, or HR.

Priority level stated is invalid. Level must be integer constant, between limits 1 to 11.

Variable name has been used more than once in STORAGE statements in this or parent
program unit.

This tag has been previously defined, or this label has been previously defined within this
SNAP.

The first character of a variable name must be alphabetic.

Could not make sense out of what to DO. Probably unrecognizable characters. Subroutine names must begin with alphabetic character.

Illegal FORMAT statement.

This statement is not permitted at PHASE level. .FORMAT, .HEADER, and .STORAGE
are not permitted at PHASE level.

Error in declared FORMAT statement.

This statement must have a tag reference appended to it for other statements to reference.

Subroutine names must begin with an alphabetic character.

This symbolic device name does not exist in this INDAC 8 System.

The ACTION section can only appear in a PHASE unit.

Dot statements (for example, . FORMAT) are not permitted within executable code in
the PROCESS section of a SNAP unit.

The word STOP, or START is missing, misspelled, misplaced, or otherwise unrecognizable.

Same as 52, except first character of the description word is not even alphabetic,

Missing comma in this statement.

This timer statement refers to an activity statement that was never declared.

A timer statement can only reference an activity statement.

Missing right parenthesis in this statement.
(continued on next page)

D-2

Error Message

Meaning

PRIORITY INTERRUPT is misspelled, misplaced or occurs more than once in ACTION
section. Only one PRIORITY INTERRUPT per PHASE is permitted.

The .ACTION statement can only occur once within a PHASE program unit.

This statement can only appear in the ACTION section of a PHASE program unit.

DO PHASE may not appear as a freestanding statement. DO PHASE may be used only
in activity or EXIT statements.

DO SNAP out of order. "DO SNAP #t PRIORITY n" may only appear in the action or
process sections.

Activity statements may only appear immediately following the .PHASE statement and
before the .ACTtON statement.

This activity statement is incomplete.

The .PHASE statement is out of sequence. Generally this error occurs when a .PHASE
follows a .PHASE without an intervening SNAP program unit.

The .SNAP statement is out of sequence. Generally this error occurs when a .SNAP
erroneously follows a SUBROUTINE unit or appears in the system unit before a .PHASE
statement occurs.

Only one .INTERRUPT device list is permitted during compilation.

The .INTERRUPT statement is out of sequence. It may appear only at system level.

The .SUBROUTINE statement is out of sequence. This error generally is due to no
SNAP unit associated with the preceding PHASE unit.

Either ( 1) an executable statement cannot have a tag reference appended to it or (2) executable statements can only occur in the PROCESS section (with some minor exceptions in the ACTION section).

The first character of a device symbolic name must be alphabetic.

This subroutine name has been used before, either it appears in the subroutine library
or the user has declared it twice in two SUBROUTINE statements.

The tag reference for this PHASE or SNAP has been used before.

Missing left parenthesis in this statement.

The .PROCESS statement may only appear only in a SNAP or SUBROUTINE program
unit.

Subroutine size is exceeded. The compiler code for a subroutine cannot exceed 128
words.

While generating compiled code for the preceding program unit, an undefined label or
tag has been found. The user evidently referred to it but never defined it (for example,
GOTO 100, but statement label 100 never occurred within the unit).

The number of digits specified exceeds the format statement limit. Total width may not
be greater than 15 digits, only the first six digits output are significant.

The line containing .HEADER must be blank following .HEADER. The header line(s)
can only appear in the line(s) following.

There is an extraneous decimal point in this .FORMAT statement.

The system tables not on system - either (PZ), (IF>, (SD>, or (XS) do not appear on the
,..,,,..+n~

;,yaLVlll.

(continued on next page)

D-3

Error Message

Meaning

There is an illegal octal constant in this statement. Octal constants rr.ay not contain the
digits B or 9 and must be four or less characters in length.

There must be at least one executable statement (for example, DO SNAP) in the
ACTION section of a PHASE unit.

An array size declaration in a storage statement must appear as ( 1, N) where N is an
integer constant greater than a positive one.

The right parenthesis is missing in a "repeat" command in a .FORMAT statement.

The .EQUIPMENT statement. If present, must be the first statement in the system.

The compilation has been aborted. No object code has been generated.

The control variable in a FOR statement can only be an unsubscripted integer variable,
or an integer variable subscripted by an integer constant.

Missing equal sign in this FOR statement.

One of the loop control parameters within this FOR statement is a real constant or not
a val id constant.

Nested FOR statements cannot share the same control variable.

Internal DO statements (for example, DO 100) are only allowed in the PROCESS section of a SNAP or. SUBROUTINE.

This tag is missing from PHASE reference table.

All loop control variables in a FOR statement must be integer mode.

Activity statement tag is not defined in current PHASE.

"DO PHASE #t EVERY i ... "is not a legal activity statement.

EXIT statements are not allowed in the ACTION section of a PHASE.

An integer variable used as a time declaration in an activity statement must have been
declared in system storage.

Missing time dimension (HR, MIN, SEC).

Unrecognizable subscript in activity statement.

A time declaration in an activity statement can only be subscripted by an integer constant.

This statement has a variable name which is in system storage. Subroutines cannot reference variables in system storage.

Dummy variable cannot be subscripted by integer constant.

Table storage capacity has been exceeded in the Compiler. Try breaking the expression
into simpler expressions and recompile. This error can only occur only with unusually
large arithmetic (or logical) expressions with enormous quantities of parentheses.

This, too, is a rare error which can occur in an arithmetic (or logical) expression with
very large quantities of function calls and parentheses. As with error D4, try breaking
the expression into simpler expressions and recompile.

Too many right parentheses in this expression.

A dimensional variable which appears in the .EQUIPMENT statement may appear in a
.STORAGE statement only as the name alone, with no dimension appended. The
.STORAGE statement will allocate storage for the variable according to the implied
dimension appearing in the .EQUIPMENT statement.
(continued on next page)

D-4

Error Message

Meaning

A computed GOTO statement cannot have a variable which is subscripted by another
variable. The subscript, if present, must be an integer constant.

There is an unrecognizable line in the .EQUIPMENT statement.

The second channel must be greater than the first channel in a multiplexer declaration.

A variable cannot be declared twice in the .EQUIPMENT statement.

The .EQUIPMENT statement contains two control lines having the same tag reference.

This control action is not defined in this INDAC B/2 System.

Is this an attempt at a DO conversion subroutine?

The first character of a conversion subroutine must be alphabetic.

A comment line cannot have a tag reference or label appended to it.

Dummy variable is previously defined (either in system storage or as another dummy in
this statement).

When addressing a multiplexed device, al I variables must have appeared in the
.EQUIPMENT statement.

Missing left parenthesis in GET/SEND statement.

This 1/0 direction is not possible (permitted) with this device.

The tag reference in a GET/SEND statement must be to a .FORMAT or .HEADER
statement or to a control line in the .EQUIPMENT statement.

The tag reference in this GET /SEND statement is undefined.

The control tag referred to in this GET/SEND statement is not associated with this
device.

Undefined conversion subroutine called for by the control tag referred to in th is GET I
SEND statement.

This device is not formatable.

A data list is not permitted when outputting a .HEADER.

A subroutine dummy is not permitted as a subscript, or as variable in another subroutine
call.

The variable was not declared to be dimensioned, yet it is subscripted within this statement.

The syntax of this FOR statement is incorrect (for example, the TO is missing or misspelled, or the STEP is wrong, etc.).

Parameters in a FOR or DO subroutine statement must not be subscripted by other
variables; however, subscripting by an integer constant is permitted.

This NEXT statement refers to a loop control variable that either was never declared in
a FOR statement or that was satisfied by a previously occurring NEXT statement.

Same as H5, except that the difference appears in the subscript.

Subroutine dummies cannot be used in a FOR statement.

This RETURN statement references an unknown subroutine name.

The preceding PHASE unit contains a reference to an undefined SNAP program unit.
(continued on next pagej

D-5

Error Message

Meaning

EX IT statements cannot request execution of an internal subroutine. (For example,
EXIT THEN DO 100 is illegal.)

The first character of the subroutine name must be alphabetic.

Sometime during compilation there was a reference to a SUBROUTINE which never
got defined.

Sometime during compilation there was a reference to a PHASE unit that never got
defined.

The preceding SNAP or SUBROUTINE program unit contains a FOR statement that
was never terminated by a NEXT statement.

The SUBROUTINE name is illegal for one of the following reasons: (1) An external
subroutine cannot be core-resident. (2) A subroutine name is used as a function call.
(3) A function name is used as a subroutine call.

Multiple base entries have been detected in the external subroutine table. GENDAC has
set the "A" bit twice in the chain.

The END statement is not in order. It must follow either a SNAP unit or a subroutine
unit.

This statement is complete so far as the Compiler is concerned, but extraneous characters occur at the end.

Missing comma in GET /SEND statement.

A subroutine dummy cannot appear in a GET /SEND statement.

Error while unpacking data list in preset, unrecognizable constant.

Integer variable cannot be preset with real constant.

Array not large enough to preset with number of entries in data list.

Real variable cannot be preset with integer constant.

Real constant cannot specify the number of repeats, real constant encountered out of
order.

End of .STORAGE statement encountered within preset, terminate preset with a slash.

There is no END statement. The END statement was inadvertently omitted.
NOTE
The errors that follow are internal compiler errors.
They may occur when a user error has been encountered and recovery of compilation assumed a
condition that was later incorrect. These errors
may be ignored unless they are the only errors in a
compilation. If this occurs contact nearest Software
Support personnel.

Attempt to move tag from System tag table to Phase tag table but unable to find in
System tag table (tape 4).

Entry in Executable Verb table exceeds size of table (tape 4).

Label around ASCII string not present in .HEADER statement.

Label for base of executable code or base of timer code not present.

Error in Phase Tag table for this phase tag (tape 4).

D-6

APPENDIX E
SPUT ERROR CONDITIONS

SPUT has three classes of errors:
1.

Single letter - Monitor message same as listed in Appendix C.3.

Message which states the error condition. Usually a disk error. If CDT ERROR occurs list
source program and contact nearest software support personnel, because this is a system
diagnostic.

?! and Halt. A disk error has occurred while reading the Executive into core from disk unit
1. AC contains the disk error flags.

E-1

APPENDIX F
INDAC 8/2 EXECUTIVE ERROR MESSAGES

There are five general classes of Executive errors.
Error Message

Meaning

00-19

Interrupt level system error.

Interrupt level stack overflow.

20-39

Interrupt level user detected error.

Interrupt chain error. No device responded but a device interrupted.

Device not listed in compilation.

No devices specified for this interrupt. Interrupting device does not appear in
. I NT ERR U PT statement.

40-59

System level user detected error.

No file number specified in file command.

No data list specified in file command.

Real variable specified page in file command.

Real variable in GET (device) .... where device is "CLOCK", "KEYS", "STATUS", or
"PRIORITY".

Real variable output attempted via "O" format control.

Real variable in GET (TTY, ... statement.

Negative subscript encountered.

60-79

System level system error.

Invalid interpretive operation code encountered.

Subroutine nesting limit exceeded.

System level stack overflow.

90-99
99

Miscellaneous errors.
User or 1/0 Handler call to error routine PC and AC listed to locate caller and reason.

F-1

APPENDIX G
GENDAC ERROR MESSAGES

Some error messages are of the form:
ERROR NN (additional data)
Tables G-1 and G-2 are a complete list of GENDAC error messages; the following notes apply to this table:
NO NOTE

Parentheses contain "debug" number, it can be ignored; it distinguishes
possible sources of this error and can change with reassembly of GENDAC.

NOTE A

The above debug number is preceded by the content of the accumulator
when the error was detected; in many cases this bears some relationship to
the field or expression that is syntactically incorrect.

NOTE B

The above (both) are preceded by a code for the field or expression found
on the library tape when the error was detected. Meaning of code:

0 =end of tape
1 =code

2 =origin symbol other than *7777
3 = Field pseudo-op
4 = *7777
NOTE C

The above (except Note B) are preceded by a second "debug" number of
the same kind.
Table G-1
Numbered G ENDAC Error Messages

Meaning

Notes

DN full

Disk full

No free block above start of system table to be extended

Overlay not found in GOVL

Invalid (PZ)

06
07

(XS) pointed to nonsubroutine block (discovered at REPLACE? -Y)

HS R ti me-out

"Impossible" system errors

Library tape contains garbage
(continued on next page)

G-1

Table G-1 (Cont)
Numbered GENDAC Error Messages
NN

Notes

Meaning

EQT, *, FIE LO found when code expected - following cases distinguishable by
values dependent on assembly:
TEXT, length (*761N), length (*762N, *763N), #entry points, Rel.
addr. of e.p., device usage code, IQT, #ctl. wrds, Ctl. wrd. value.

FIE LO fol lows tape leader

long name of *7660 group missing

*761 N, *762N, *763N missing

*7610, *7630 illegal

modtyp = 5 illegal

FIELD after *761N, *762N, *763N missing

FIELD after *761N, *762N, *763N not 0 or 1

module origin requirement after *761 N, & 762N, *763N missing

length of call-up module> 2008

length 0 or too big

#entry points 0 or too big

missing control-origin after# entry points

missing *7650

# control words 0 or too big

more control words than declared

FIE LO in place of control word or control-origin at end of control words

*7647 illegal

*7641 not associated with *763N

entyp incompatible with modtyp or two *7641 or *7642 in one module

Core-res. function entry point invaiid

l/Q device handler entry point invalid

IQT invalid

*7651 name not in(**)

*7651 name refers to *7640 not *7641 or occurs twice on one tape

Missing *766N or actual origin of module code

Illegal syntax in *766N TEXT

Displacement fixups iiiegai in disk-resident moduie

46
47

Missing control-origin after *7777 separator (EQT is legal)
Manual repositioning of tape associated with "N" response did not leave tape on
a *7777 separator

G-2

Table G-2
Unnumbered G END AC Error Messages
Error

Description

Not an error message - indicates GENDAC awaiting keyboard
input (printing of previous message was suppressed) Type'?' to
discover.

Keyboard input not understood - (or illegal) - try again

User aborted run with CTR L/C from keyboard

NO SUCH FILE

No file by that name exists of type S if OPT=S or of type B if
OPT=B

SUBROUTINE FILE INVALID

File specified does not have proper format for INDAC compiler
output of a single subroutine

SUBROUTINE EXCEEDS ONE PAGE

Self-explanatory

AAAA ALREADY IN (BB)

Name specified (only four characters printed) exists in specified
system table - in general: logical error in constructing.or loading library tape(s) - may need to regenerate system from scratch.

FIXUP ILLEGAL

Reference to disk-resident module in a *766N

FIXUP UNDEFINED AAA BBBB

*766N did not appear for this "FIELD" .AAAA =index of fixup
(last digit of *766N)
BBBB =value of "ICI" for word to be fixed up

AAAA EXCEEDS ALLOCATED AREA

Module code lies outside declared module length

TURN ON LSR

LSR time-out has occurred - if caused by jam, turn off until jam
fixed, if caused by moving switch to stop, no data lost; if caused
by moving switch to free, or if tape sprocket holes ripped, must
restart system generation.

TURN OFF LSR

GENDAC is returning to disk monitor - follow instructions!

CHECKSUM ERROR AAAA

If AAAA changes or message does not readout on repeating entire system generation process, then hardware trouble with paper
tape reader or punch which produced library tape. If AAAA reoccurs, then bad copy of library tape (probably punch error).

DISK READ/WRITE ERROR

Hardware trouble with disk

CANNOT CONTINUE

Self-explanatory

***SYSTEM BASE ALTERED!***

Base of I NDAC program area in core has been altered by
GENDAC due to addition of core-resident code. All INDAC
main programs must be recompiled to incorporate this change.

USE BK (XXXX)

A library tape option valid only for BK INDAC system has been
used when K=4

(AA) MISSING OR INVALID

Self-explanatory

(IF) FULL

Too many core-resident functions have been defined (limit is
eleven) or (IF> is bad.

CANNOT ALLOCATE CORE

Either no room left in valid area for type of module now being
allocated or (CM) is bad.

G-3

APPENDIX H
EXECUTIVE COMMAND DECODER OPERATION

H.1

OPERATOR COMMANDS
CTRL/A

Attach to Command mode if echo successful. If no echo or if ? ,
Command mode is not permitted.

All other commands active only in Command mode.
CTRL/C

Return to Monitor mode

CTRL/D

Initiate "RUN" command

CTRL/U

Delete entire line (Start line over)

CTRL/P

Release Teletype from Command mode back to Operator mode

CTRL/V

Value changes to be made in following modes

Inspect and open clock for preset HH:MM:SS

O,D,E

Inspect and open location for modification in Octal, Decimal, Exponential
Notation
Delete and echo last character in line; if none, no echo

RUBOUT

H.2 COMMAND DECODER ERROR MESSAGES
BA

Bad Address:

The address requested is not within the defined system storage area

Error in Phase:

The phase requested cannot be found

Error in Syntax:

The input line cannot be resolved or the arithmetic information is
not correct for the mode of operation.

Cannot resolve command.

H-1

APPENDIX I
THE GENDAC LIBRARY

The GENDAC library supplied with the INDAC software kit contains a library of 1/0 handlers and mathematical
functions; they are:

1/0 Handlers
KEYS
TTY

PART OF
EXECUTIVE

STATUS
PRIORITY
CLOCK
UDCE (UDC)
UDCP (UDC)
ADC (AFOl, AF02, AF03, AFC-8, ADOl)
AF04
DAC (AAOlA, AA50)
TTY2
TTY3
TTY4
PTP

Mathematical Functions
SIN

cos
ATAN
LOG
EXP
PART OF {
EXECUTIVE

SORT

Sample Call for Functions
LET A =SORT (B)

1-1

APPENDIX J
SYSTEM COMMUNICATION TABLES

Table J-1 lists and describes the Systems Communication Tables.
Table J-1
System Communication Tables
Save

Used By

Content

Compiler

Page zero of eventual run-time system including various system communication parameters
used by SPUT, GENDAC and the Compiler.

400-777

Compiler

Device handler names and pointers; legal
.EQUIPMENT statement control words.

SGEN 1

1600

Compiler

Subroutine names and pointers; function
names and (IF) indexes.

Intrinsic
Functions
(IF)

Executive

7000

Compiler

Vectors to operators and functions.

Core Map
(CM)

Executive

200

GENDAC

Defines the areas of Executive core available
to core-resident modules.

(**)

GENDAC 2

GENDAC

Allocation data for call-up modules and entry
points used for fixup purposes alone.

Name

Located In

Page Zero
(PZ)

Executive

System
Devices
(SD)

SGEN 1

External
Subroutines
(XS)

Address

1 Created by SGEN and modified by GENDAC when the user configures a specific system.

(The SGEN program
examines the system configuration to determine the number of disks present and modifies internal Disk Monitor
System tables to reflect the disks present.)

2 (**) is cleared at entry to GENDAC.

J-1

APPENDIX K
ESUP OPERATION

K.1

INTRODUCTION

ESUP is a special support program for INDAC. ESUP produces an ASCII paper tape of equivalences (=) for use
with PAL-I. One set of such equates is the actual interpretive code produced by the INDAC compiler for the runtime Executive. Those users who wish to code their own interpretive routines will find these equivalences useful.
Other equates include addresses or instructions to link user code to the Executive. The file used to extract the
equivalence information is the (IF) file. This is a communication table for the Executive and the Compiler; it defines the locations within the Executive of various calls that the Compiler or user will require. ESUP is also capable of producing a listing of available core in the executive area. This will be useful to users who wish to write
1/0 handlers or core-resident functions.

K.2 LOADING PROCEDURE
The following procedure will enable the user to load ESUP.
Procedure

Step

Place the ESUP paper tape in the high-speed reader.

Turn on the high-speed punch.

Load the program using the disk monitor system as follows:
.LOAD)
*IN-R:)

*
ST=200)
tt (User types CTRL/P after each t)
K.3 PRODUCING THE EQUATE TAPE
The following procedures will produce the equate tape.

Step

Procedure
The program types:
*OUTThe user then responds to the output request with R:)

(continued on next page)

K-1

Step
2

Procedure
The program types:
*IN-

The user then responds to the input request with S:<IF>).

The program types:
*OPT-

The user then responds to the option request with I (no CR is required).

The program then punches the equate tape, using the high-speed punch. When
the output is finished, the program types:
FINI

and returns to request further outputs:
*OUT-

K.4 PRODUCING THE CORE MAP

The following procedure will produce a printout of the core map.
Procedure

Step
The program types:
*OUT-

The user then responds to the output request with T:).
2

The program types:
*IN-

The user then responds to the input request with S:(CM>).

The program types:
*OPT-

The user then responds to the option request with C (no CR is required).
4

The program then types a table of the first free location on each page of the executive image and the length of the free area on that page. When the output is
finished, the program types:
FINI

and returns to request further outputs:
*OUT-

K.5 RETURNING TO MONITOR

Typing CTRL/C to a program request returns the user to the Disk Monitor.
K.6 RESTRICTIONS

The system must run under Disk Monitor control. The system assumes a high-speed punch for the (IF) table output.

K-2

K.7

REFERENCES

See Disk Monitor System, DEC-D8-SDAA-D for an explanation of the loading procedures and for more insight
into the programmatic use of the Disk Monitor Command Decoder.

K.8 EXAMPLES
K.8.1

Loading and Usage
.LOAD
*IN-R:
*
ST=200

tt
*OUT-R:
*
*IN-S:(IF)
*
*OPT-I
FINI
*OUT-

K.8.2 Sample Equate Printout
A sample printout of a portion of the generated ASCII Equate tape.
NOTE
Equivalences may vary from one version of an Executive to another. The following is a sample of one
version.

K-3

IL=
0201
Ill=
0200
ILE=
0277
I LEI=
0276
FL=
0236
FLI =
0233
FLE =
0307
FLEI = 0304
IS=
0211
ISi =
0210
ISE =
0302
ISEI =
0301
FS =
0253
FSI =
0250
FSE =
0314
FSEI = 0311
MPY =
0512
FMPY = 1200
DIV=
0437
FDIV = 1077
GOTO= 0361
CGO =
0400
IF=
0366
FIF =
1317
OR=
0556
LAND= 0552
NOT=
0564
NEXT= 0327
FIX=
1015
FLT=
1000
NEG=
0563
FNEG = 1012
GOSB = 0057
STOP= 0000
PAUZ = 0000
EXIT= 0037
ABRT = 0000
DUM1 = 0001
DUM2 = 0002
DUM3 = 0003
DUM4 = 0004
DUM5 = 0005
THLD = 0052
TMER = 0011
DUM6 = 0000
EXP=
1376
FEXP = 0110
RET =
0072
RETE = 0013
ADD= 0425
FADD = 0603
SUB=
0423
FSUB = 0600

K-4

INDAC HANDBOOK INDEX
A (cont)

AA01, 3-33

Assignment, 3-3
Arithmetic, 3-3

AA50, 3-33
Activity, 3-42,'3-47

Boolean, 3-7
Function, 7-17

Action, 3-42, 3-49
ADC, 3-30, 3-34

Assembler, 5-3, 5-22, 7-3, 7-4

AD01, 3-34

AT activity, 3-49

Addition, 3-4

A tangent, 4-8

Address

Index, 3-12, 3-20, 7-26

B
Background tasks, 3-53
BIN loader, 2-2, 4-3
Binary files, 5-3, 5-22
Binary mode, GENDAC, 4-7, 7-4

Pointer, 7-25

Binary operator, 7-27

Base, 3-53
Channel, 3-28, 3-39
Displacement, 7-23

System parameter, 3-53, 5-14, 6-2

Bit pattern, 3-6

Start of user system, 6-1

Block (page)

AFC8, 3-34

Disk, 3-38

AF01, 3-35

Paper tape, 2-3, 4-4, 7-5

AF02, 3-35

Boolean, 3-6

AF03, 3-35

Bit pattern, 3-6

AF04, 3-36

Expression, 3-7

Algorithm, 3-2

Operators, 3-6, 7-27

Analog to digital converter, 3-34
AND operator, 3-7
Arguments

Vector code, 7-28
Bootstrap loader, Monitor, 5-2
Branching, 3-9

Arithmetic, 7-23

Conditionally, 3-9

MACRO calls, 7-20

Unconditionally, 3-9

Subroutines, 3-44, 7-18

Buffer

1/0, 3-54, 7-14

Arithmetic, 3-3

Text, 5-7

Expression, 3-5

Fixed point, 7-22
Floating point, 7-22

Calls

Operators, 3-4, 7-27

Function, 7-17

Vector code, 7-23

MACRO, 7-20
PHASE, 3-47

Array, 3-20, 3-24
Element, 3-25

SNAP, 3-47

I nteger, 3-25

SUBROUTINE, 3-44

Name, 3-24

CANCEL request, 3-61

Preset, 3-26

Channel declaration, 3-39

Real, 3-25

ADC, 3-34

Subscript, 3-25

AF04, 3-34

Window, 3-25

DAC, 3-34

ASCII Files, 5-3, 5-5, 5-13, 5-16
ASCII string, 3-22

UDC, 3-37, 3-55
Checksum error, 2-3, 4-4

Assembiy-ievei Code, 4-7, 7-H

lndex-1

INDEX (Cont)
C (Cont)
CLOCK, 4-9

C (Cont)
Control declaration

Display, 3-17

ADC, 3-34

Get, 3-17

AF04, 3-36

Set, 3-17, 6-3

Fl LE, 3-38

Code

TTY, 3-60
Interpretive, 7-3
INDAC statements, 3-2

UDC, 3-55
Control driver table, 7-20

MACRO calls, 7-20

Control word, 7-14

PAL-I statements, 7-3, 7-28, K-1

Conversion routine, 3-31

Vector, 7-3, K-1

Core

Cold start, 2-1, 4-2

Field, 5-23, 7-12

Command Decoder,

Functions, 4-8, 7-1, 7-12, 7-17

Disk Monitor, 5-1

1/0 handlers, 7-12

Executive, 3-51, 6-1

Job parameters, 3-53

COMMAND request, 3-60

Limitations, 3-54, 5-15, 6-1

Comment, 3-43

Map, 7-2, 7-12, J-1, K-1

Communication

Monitor head, 5-2

Compiler tables, 7-1

Scheduling parameters, 3-53

Executive to SNAP, 3-54

SNAP, 3-53

Operator guidance, 3-54

Subroutine, 3-53
Swapping, 3-53

COMP1-4, 2-5, 4-5
Compiler

System parameters, 5-14

Communication tables, 7-1

COS (change of state), 3-57

Dialogue, 5-14

Cosine, 4-8

Equates, 7-17, K-1

Error messages, 5-15, D-1

DAC, 3-28, 3-33

Loading, 2-4, 4-5

Data collection and control, 3-27

Printout, 5-14

Data I ist, 3-17

Configuration, GENDAC, 4-2

GET statement, 3-35

Initial run option, 4-2

SEND statement, 3-34

Re-run option, 4-2

Decimal notation, 3-4, 3-24

Constant, 3-3, 3-20

DELAY activity, 3-48

Integer, 3-3

Device, 3-33, 3-39, 7-14

Octal, 3-6

Channel, 3-39, 7-14

Real, 3-3

Control, 3-39, 7-14
Declaration, 3-39

Console, 3-59

Hardware, 3-33, 4-9

Operator mode, 3-59

1/0 handler, 3-33, 4-9

System mode, 3-59
Continuation line

Name, 3-33, 4-9, 7-14
Table, 2-5, 4-6

EQUIPMENT statement, 3-28, 3-39
FORMAT statement, 3-18

Dialogue

Program statement (see .STORAGE statement)

Compiler, 5-14

STORAGE statement, 3-20

Editor, 2-7, 5-5

lndex-2

INDEX (Cont)
D (Cont)

GENDAC, 4-8, 7-29

E (Cont)

EVERY activity, 3-47
Executive

PIP, 5-16
SPUT, 6-1

Commands, 6-3, H-1

Digital input, 3-37

Loading, 2-5, 4-6

Digital output, 3-37

Operation, 3-50, 6-1

Digital to analog converter, 3-33

Parameters, 7-20

Dimensioned array, 3-20

Explicit 1/0 handler, 3-37

Disk

Exponential notation, 3-19, 3-24
Functions, 3-53, 4-8, 7-12

Exponentiation, 3-4

Job, 3-53

Expressions, 3-5

Monitor, 4-2, 5-1

Arithmetic, 3-5

Object file, 5-13

Boolean, 3-7

PHASE, 3-53

SNAP, 3-53

Field interrupt, 3-51, 3-55

Source file, 5-5

Fields, 5-23, 7-12

Subroutine, 3-53, 7-12

File

System programs, 5-2

Copying, 5-18

Division, 3-4

Creating, 5-5

DO/RETURN statements

Deleting, 5-17
Listing, 5-17

Language statement, 3-15

Merging, 5-19, 5-21

Vector code, 7-26

File name, 5-3

Editor, 2-7, 5-5

ASCII, 5-5, 5-14

Commands, 5-7, C-1

Binary, 5-23

Dialogue, 5-5

Input, 5-3

Error messages, 5-13

Output, 5-3

END statement, 3-1, 3-31
Equates

System, 5-14, 6-1
FI LE, pseudo device, 3-31, 6-1, 6-3

Compiler, 7-17, K-1

Fixed point (integer), 7-22

GENDAC, 7-5

Floating point (real), 7-22

EQUIPMENT, 3-28, 3-39, 3-53

Foreground tasks, 3-52

Channel declaration, 3-28

Format driver table, 7-20

Control declaration, 3-30

Format, library tape, 7-9

Subroutine declaration, 3-31
Error messages
Compiler, 5-15, D-1

Format, numeric, 7-22
FORMAT statement, 3-17, 3-22, 3-53
FOR/NEXT statements

Editor, 5-13

Language statement, 3-12

Executive, F-1, H-1

Vector code, 7-27

GENDAC,G-1

Functions, 4-1, 7-1, 7-17

LOAD, 5-22

Atangent, 4-8

Monitor, 5-2, 6-2

Coding details, GENDAC library, 7-16

PIP, 5-18

Cosine, 4-8

SPUT,E-1

Logarithm, 4-8

ESUP, 7-12, 7-17, K-1

Sine, 4-8

lndex-3

INDEX (Cont)
I (Cont)

G
GBIN vector, 7-28

Input file, 5-3, 5-5, 5-13, 6-1

GENDAC

Integer constant, 3-3, 3-6

Binary mode, 2-7, 4-7, 7-4

Integer quantity, 3-3, 3-6

Dialogue, 2-7, 4-7, 7-29

Constant, 3-3, 3-6

Error messages, G-1

numeric format, 7-22
variable, 3-3, 3-6

Equates, 7-5, 7-9
Library tapes, 4-7, 7-5

Integer variable, 3-3

Library structure, 7-5

Address, 3-22, 3-39

Run options, 4-2

Array, 3-20, 3-24, 3-39

System mode, 7-3

Data, 3-20, 3-24

Generic code, 3-56

Integrating digital voltmeter, 3-36

GET statement

Iteration, 3-12

ADC, 3-31, 3-35

Interpretive Code, 7-3

AF04, 3-36

Interrupt devices, 3-51

CLOCK, 3-17, 3-20

INTERRUPT statement, 3-51, 3-53

FILE, 3-38

Interrupt task, 3-52

KEYS, 7-2

1/0 handlers, 3-39, 4-1, 7-19

PRIORITY, 3-54

ADC, 3-34

STATUS, 3-52, 3-54, 3-62

AF04, 3-36

TTY, 3-61

CLOCK, 4-9

UDCE, 3-38

Coding details, GENDAC library, 7-19

UDCP, 3-55

DAC, 3-33
Fl LE, 3-38, 4-1

Global Parameters, 3-20, 3-24, 3-41
GOTO Statement

KEYS, 3-39, 4-1
PRIORITY, 3-39, 4-1

Language, 3-9

PTP, 4-9

Vector Code, 7-25
H

STATUS, 3-39, 4-1

HEADER Statement, 3-17, 3-22, 3-53

TTY, 3-39, 4-1

Headings, 3-22

TTY 2, 4-9

Head of Monitor, 4-2, 5-1

TTY 3, 4-9

HELD, 2-5, 4-6, 5-

TTY 4, 4-9

HINDAC, 2-1, 4-2

UDCE, 3-37
UDCP, 3-55

IDACS, 3-1
Identification codes, library, 7-9

J
Job specification, 3-41

IDENTIFY request, 3-56
IDVM, 3-36

K
KEYS, 3-39, 7-2

I F statement

Language, 3-10

Label, 3-9

Vector code, 7-25

Language statements

INDAC language statements, 3-1, A-1

Compiler code (vector) equates, 7-17, K-1

INDAC support programs, 2-4, 4-4

INDAC, 3-2, A-1

INDAC system tables, 2-4, 4-5, 7-1

GENDAC equates, 7-5

INITIALIZE request, 3-56

MACRO calls, 7-28

lndex-4

INDEX (Cont)
L (Cont)
PAL-i, 7-28, K-1

Name

LELD, 2-6, 4-7

Array, 3-20, 3-24

LET statement

Device, 3-28, 5-3, 7-14

Arithmetic, 3-3

File, 3-38, 5-6, 5-16, 6-1

Boolean, 3-6

Function, 4-10, 7-17

Function call, 7-17

1/0 handler, 4-9, 7-14

Vector Code, 7-23

Integer variable, 3-3, 7-22

Linearization routine
Call, 3-31, 3-44, 7-1
Coding details, 7-16
Definition codes, 7-9
Library codes, 7-9
Modules, 7-5
Paper tape, 7-5
Lists, 3-17, 3-20, 3-24, 3-3, 3-55, 3-60
LOAD, 2-8, 4-2, 5-22
Loading, 2-1, 4-1

Programs, 5-1
Real variables, 3-3, 7-22
Subroutine, 3-44, 7-18
Nesting, 3-12
NOT operator, 3-8
Numeric formats, 3-3, 3-24, 7-22

0
Object file, 6-1
Octal constant, 3-6

Cold start, 4-2

Octal notation, 3-6, 3-24

Files, 5-5, 5-23

Operating mechanics, 2-1

Monitor head, 5-1

Operator guidance, 3-16, 3-59

Sample system, 2-1

Operator mode, console, 3-59

User system, 4-1

Operators

Local parameters, 3-43

Arithmetic, 3-4

Logarithm, 4-8

Binary, 7-27

Logs, 3-16

Boolean, 3-6

Looping, 3-12, 7-27

Relational, 3-10

LSV (last scanner value), 3-57
M

Unary, 7-27
OR operator, 3-7

MACRO calls, 7-28

Output format, 3-17, 3-22, 7-14

MAKE, 2-4, 4-5
Mask, 3-13

Output file, 5-3, 5-5, 5-13
p

Matrices, 3-21

Page zero, 3-53, 7-2, 7-20

Messages, 3-16

Page (Block)

Modules, library

Core, 3-54, 5-15, 7-5, 7-12

Basic, 7-5

Disk, 5-15, 6-1, 7-12

Call-up, 7-6

Paper Tape, 2-3, 4-4, 5-7, 7-5

Extension, 7-6

PAL-I, 7-3

Size, 7-5

Parameters, 3-3

Monitor head, 4-2, 5-1

Constant, 3-3

Monitor bootstrap, 5-2

Executive, 7-20

Monitor mode, 5-1

Global, 3-41

Monitor system dump, 2-3, 4-4

Local, 3-43

MSUP, 2-2, 4-2, 4-3

Scheduling, 3-53

Multiplex devices, 3-27, 7-14

System, 3-53, 5-14, 6-7

Mu!tip!ication, 3-4

\/,,r;,,hlo ~-~
• Y• •U-•-, """ ......

lndex-5

INDEX (Cont)

S (Cont)

P (Cont)
PDP-BE, 2-1, 4-2

Immediate, 3-49

PDP-81/L, 2-1, 4-2

Parameters, 3-53
Priority, 3-50, 3-53

PHASE, 3-1

Timed, 3-47

Parameters, 3-53
Scheduling, 3-46, 3-53

Segmentation, 3-40
GENDAC equates, 7-5

Structure, 3-42

Phases, 3-42

PIP, 5-16
Dialogue, 5-16

Snaps, 3-43

Error messages, 5-17, 5-18, 5-19

Subroutines, 3-44

Options, 5-16

SEND statement

Point table 1/0 handler, 3-55

DAC, 3-29, 3-34

Priority level, 3-50, 3-54

Fl LE, 3-38

Background, 3-53

PTP, 3-18

Foreground, 3-52

TTY 3-18 3-60

INTERRUPT, 3-51

UDCE, 3-30, 3-38

PRIORITY pseudo device, 3-54

UDCP, 3-55

Process interface devices, 3-27

SGEN, 2-5, 4-6, 5-1

ADC, 3-34

Signal conditioning, 3-31

DAC, 3-33

Sine, 4-10

Digital 1/0, 3-37

SNAP, 3-1

IDVM, 3-36

Background, 3-53

PROCESS statement, 3-1, 3-31, 3-43

Communication, 3-54

Pseudo devices, 3-39

Foreground, 3-52

FILE, 3-38, 5-15, 6-1, 6-3

INTERRUPT, 3-51

KEYS, 3-39, 7-2

Scheduling, 3-4~, 3-54

PRIORITY, 3-54

Size, 5-15, 6-1

STATUS, 3-54, 3-62

Structure, 3-43
Source file, 5-1, 5-5, 5-13

PTP, 3-17, 4-9
Q

Queue (Stack), 3-5i

SPUT, 2-9, 5-2, 6-1
Stack (Queue), 3-51

R
Real constant, 3-3, 3-6

START, timer action, 3-42, 3-49
Statements, 3-1, A-1

Real quantity, 3-3, 3-6

Data collection control, 3-27, 3-39

Constant, 3-3, 3-6

Formatting, 3-17, 3-22

Numeric format, 7-22

Interrupt, 3-51

variable, 3-3, 3-6

Messages, 3-16

Real variable, 3-3

Programming, 3-3

Array, 3-20

Segmentation, 3-40

Data, 3-4

Scheduling, 3-46

Relational operators, 3-10

Storage, 3-20, 3-24

Reports, 3-16

STATUS pseudo device, 3-54, 3-62

RIM loader, 2-2, 4-3

STOP, timer action, 3-42, 3-49

s
Scheduling, 3-1, 3-46, 3-51, 3-53

Storage requirements, 5-15, 7-5
STORAGE statement, 3-20, 3-24, 3-53

lndex-6

INDEX (Cont)

S (Cont)
Subroutines
Coding details, GENDAC library, 7-18
External, 3-44, 7-1

UDC, 3-30, 3-37
UDCE, 3-30, 3-37
UDCP, 3-37, 3-55

Implicit, 3-31, 3-45, 7-1

Unary operation, 7-27

Internal, 3-44

User mode, 5-1

Size, 5-15, 7-5
Subtraction, 3-4

Variable, 3-3

Subscript, 3-21, 3-25

Integer, 3-24

Swapping, 3-54

Real, 3-24

System files, 5-3, 5-14

Vector code, 7-22

System mode,

DO/RETURN, 7-26

GENDAC, 4-7, 7-3

Equates, 7-17

Console, 3-59

FOR/NEXT, 7-27

System parameters, 3-53, 5-14, 6-2

GBIN, 7-28

System programs, 5-2

GOTO, 7-25
IF, 7-25

COMP, 5-13
EDIT, 5-5
PIP, 5-16
SPUT, 6-1

T
Tables

1/0, request, 7-20
LET, 7-23
XTRP, 7-28
Vector mode, 7-22
Verify error, 2-3, 4-4

Communications, 7-1
control driver, 7-20

Word, computer, 3-3, 3-25, 7-22

Format driver, 7-20
STORAGE statement, 3-21

XTRP vector, 7-28

Tag, 3-17

Activity, 3-47

<CM> table, 4-2, 5-20, 7-2, J-1, K-1

Control option, 3-30, 3-40

<IF> table, 4-2, 5-20, 7-1, J-1, K-1

FORMAT, 3-17, 3-22

<PZ> table, 4-2, 5-20, 7-2, J-1

HEADER, 3-17, 3-22

<SD> table, 4-2, 4-6, 5-2, 7-2, J-1

PHASE, 3-42

<XS> table, 4-2, 5-20, 7-1, J-1

SNAP, 3-43

<**>table, 4-2, 7-3, J-1

Testing, 3-10
Time of day, 3-17
Timers, 3-49
Delay (DELAY), 3-48, 3-49
Interval (EVERY), 3-47, 3-49
Time of day (AT), 3-48, 3-49
Titles, 3-17, 3-22
Trailer, 2-2, 4-3, 7-5
TRANSFER request, 3-57
TTY, 3-17, 3-59
TTY 2, 3-59
TTY 3, 3-59

TTY 4, 3-59
lndex-7

INDAC 8/2 HANDBOOK

DEC-11\>J-GRZA-D

READER'S COMMENTS

Digital Equipment Corporation maintains a continuous effort to improve the quality and usefulness of its publications. To do this effectively we need user feedback - your critical evaluation of this manual.

Please comment on this manual's completeness, accuracy, organization, usability, and readability.

Did you find errors in this manual?

How can this manual be improved?

DEC also strives to keep its customers informed of current DEC software and publications_ Thus, the following
periodically distributed publications are available upon request. Please check the appropriate boxes for a current
issue of the publication(s) desired.
D Software Manual Update, a quarterly collection of revisions to current software manuals.
D User's Bookshelf, a bibliography of current software manuals.
D Program Library Price List, a list of currently available software programs and manuals.

Please describe your position.
Organization-------------------Street - - - - - - - - - - - - - - Department
City - - - - - - - - - - State - - - - - - - - - - - - - - Zip or Country - - - - - - - -

- - - - - - - - -

- - -

- - - - - - - - -

- - -

Fold Here - - - - - - - - - - - - - - - - - - - -

- - Do Not Tear - Fold Here and Staple -

- -

- - -

FIRST CLASS
PERMIT NO. 33
MAYNARD, MASS.
BUSINESS REPLY MAIL

NO POSTAGE STAMP NECESSARY IF MAILED IN THF UNITED ST!\1!-'S

Postage will be paid by

mamaama
Digital Equipment Corporation
Technical Documentation Department
146 Main Street
Maynard, Massachusetts 01754