Digital PDFs

DEC-08-OCOSA-F

July 1975

395 pages

Original

12MB

Document:	D COS 300 310 System Reference Manual
Order Number:	DEC-08-OCOSA-F
Revision:
Pages:	395
Original Filename:	http://bitsavers.org/pdf/dec/pdp8/cos-300/DEC-08-OCOSA-F_D_COS_300_310_System_Reference_Manual_Jul75.pdf

OCR Text

o system

COS 300/310
System Reference Manual
DEC~08-0COSA-F-D

digital equipment corporation • maynard. massachusetts

First Printing, July 1975

The information in this document is subject to change without notice
and should not be construed as a commitment by Digital Equipment
Corporation. Digital Equipment Corporation assumes no responsibility
for any errors that may appear in this document.
The software described in this document is furnished under a license
and may only be used or copied in accordance to the terms of such
license.
Digital Equipment Corporation assumes no responsibility for the use
or reliability of its software on equipment that is not supplied by
Digital.
Copyright ~ 1975

by Digital Equipment Corporation

The HOW TO OBTAIN SOFTWARE INFORMATION pages, located at the back of
this document,_ explain the various servi.ces available to Digital software users.
The postage prepaid READER'S COMMENTS form on the last page of this
document requests the user's critical evaluation to assist us in
preparing future documentation.
The following are trademarks of Digital Equipment Corporation:

DIGITAL
DEC
PDP
DECUS
UNIBUS
COMPUTER LABS
COMTEX
DDT
DECCOMM

DECsystem-lO
DECtape
DIBOL
EDUSYSTEM
FLIP CHIP
FOCAL
INDAC
LAB-8

MASSBUS
OMNIBUS
OS/8
PHA
RSTS
RSX
TYPESET-8
TYPESET-II

LIMITED RIGHTS LEGEND
Contract No.
Contractor or Subcontractor:

Digital Equipment Corporation

All the material contained herein is considered limited rights data
under such contract.

CONTENTS
Page
xii

PREFACE
INTRODUCTION
CHAPTER 1

INTRO-1
DIBOL LANGUAGE

1-1

1.1

STATEMENT NUMBERS

1-3

1.2

STATEMENT LABELS

1-3

1.3

COMMENTS

1-4

1.4
1.4.1
1.4.2
1.4.3
1.4.4
1.4.4.1
1.4.4.2
1.4.4.3
1.4.4.4
1.4.4.5
1.4.5
1.4.6
1.4.7
1.4.8
1.4.9
1.4.9.1
1.4.9.2
1.4.10
1.4.11
1.4.12
1.4.13
1.4.14
1.4.15
1.4.16
1.4.17
1.4.18
1.4.19
1.4.20
1.4.21
1.4.22
1.4.23

STATEMENTS
ACCEPT
CALL
CHAIN
Data Manipulation Statements
Clearing Fields
Moving Alphanumeric Data
Moving Decimal Data
Moving Records
Converting and Formatting Data
DISPLAY
END
FINI
FORMS
GO TO
Unconditional GO TO
Computed GO TO
IF
INCR
INIT
ON ERROR
PROC
READ
RECORD
RETURN
START
STOP
TRACE/NO TRACE
TRAP
WRITE
XMIT

1-5
1-6
1-8
1-9
1-10
1-13
1-14
1-14
1-15
1-15
1-18
1-21
1-22
1-23
1-25
1-25
1-25
1-27
1-29
1-30
1-33
1-35
1-37
1-38
1-48
1-49
1-50
1-51
1-53
1-56
1-58

MONITOR

2-1

2.1

OPERATING PROCEDURES

2-1

2.2
2.2.1
2.2.2

MONITOR COMMANDS
BATCH
DATE

2-5
2-6
2-7

CHAPTER 2

iii

Page
2.2.3
2.2.4
2.2.5
2.2.6
2.2.7

DELETE
DIRECTORY
PLEASE
RUN
SAVE

2-8
2-9
2-11
2-12
2-15

2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
2.3.6
2.3.7

EDITING COMMANDS
ERASE
FETCH
LIST
Line Number
Number Commands
RESEQUENCE
WRITE

2-16
2-17
2-18
2-19
2-20
2-22
2-23
2-24

2.4

EDITING EXAMPLE

2-25

2.5

BATCH EDITING

2-27

2.6

ERROR MESSAGES

2-30

SYSGEN

3-1

3.1

COS-300 OPERATING PROCEDURES

3-1

3.2

COS-300 SYSTEM SOFTWARE CONFIGURATION

3-1

3.3

COS-310 OPERATING PROCEDURES

3-4

3.4

COS-310 SYSTEM SOFTWARE CONFIGURATION

3-4

3.5
3.5.1
3.5.2
3.5.3
3.5.4
3.5.5

LOGICAL UNIT ASSIGNMENT FOR THE COS SYSTEM
How Logical units are Assigned by SYSGEN
COS-300 Logical Unit Assignment
COS-310 Logical Unit Assignment
DECtape Users
Disk Users

3-7
3-7
3-9
3-11
3-13
3-13

COMP

4-1

4.1

OPERATING PROCEDURES

4-1

4.2

CONDITIONAL COMPILATION PROCEDURE

4-8

4.3

ERROR MESSAGES

4-9

PIP

5-1

5.1
5.1.1
5.1.2
5.1.3
5.1.4
5.1.5
5.1.6
5.1.7

OPERATING PROCEDURES
Transfer Binary File
Copy Device
Transfer Data Files
Eliminate Spaces in Directory
Transfer Source Files
Transfer System Program
Return to Monitor

5-1
5-2
5-3
5-4
5-6
5-9
5-10
5-11

5.2

ERROR MESSAGES

5-12

CHAPTER 3

CHAPTER 4

CHAPTER 5

Page
CHAPTER 6

BUILD

6-1

6.1
6.1.1
6.1.2
6.1.3
6.1.4

BUILD CONTROL PROGRAM
Field Descriptor section
INPUT Section
OUTPUT Section
Storing the BUILD Control Program

6-1
6-3
6-4
6-8
6-10

6.2
6.2.1
6.2.2

OPERATING PROCEDURES
KBD - Data Input from Keyboard
SYS, CDR or RDR - Data Input from System
Device, Card, or Papertape

6-11
6-12
6-13

6.3

BUILD INPUT LINE

6-13

6.4

ERROR MESSAGES

6-16

SORT

7-1

7.1
7.1.1
7.1.2

SORT CONTROL PROGRAM
File Descriptor Section
INPUT/OUTPUT Section

7-1
7-1
7-2

7.2

SORT OPERATING PROCEDURES

7-4

7.3

RUNNING SORT AS PART OF AN UPDATE PROCEDURE

7-6

7.4

MERGE OPERATING PROCEDURE

7-6

7.5

RULES FOR USING DEFAULT UNITS

7-8

7.6

ERROR MESSAGES

7-9

UPDATE

8-1

8.1
8.1.1
8.1.2
8.1.3
8.1.4
8.1.5
8.1.6
8.1.7

UPDATE CONTROL PROGRAM
File Descriptor Section
UPDATE Statement
INPUT Statement
SORT Statement
KEY Statement
OUTPUT Statement
END Statement

8-1
8-1
8-2
8-2
8-3
8-3
8-4
8-4

8.2

UPDATE COMMANDS

8-5

8.3

UPDATE EXAMPLE

8-6

8.4

OPERATING PROCEDURES

8-7

8.5

ERROR MESSAGES

8-8

BOOT

9-1

OPERATING PROCEDURES

9-1

CHAPTER 7

CHAPTER 8

CHAPTER 9
9.1

Page
CHAPTER 10

PATCH

10-1

10.1

OPERATING PROCEDURES

10-1

10.2

ERROR MESSAGES

10-5

DAFT

11-1

11.1

OPERATING PROCEDURE

11-1

11.2

COMMANDS

11-1

11.3

DAFT ERROR MESSAGES

11-4

11.4

DAFT OUTPUT

11-6

COS-OSi8 CONVERTER

12-1

12.1

OPERATING PROCEDURES

12-1

12.2

ERROR MESSAGES

12-3

FORMAT PROGRAMS

13-1

13.1
13.1.1
13.1.2

RK8MRK
Operating Procedure
Error Messages

13-2
13-2
13-2

13.2
13.2.1
13.2.2

RKEMRK
Operating Procedure
Error Messages

13-5
13-5
13-5

13.3
13.3.1
13.3.2

TDMARK
Operating Procedures
Error Messages

13-7
13-7
13-8

13.4
13.4.1
13.4.2

DTMARK
Operating Procedures
Error Messages

13-10
13-10
13-11

CREF

14-1

14.1

OPERATING PROCEDURES

14-1

14.2

ERROR MESSAGES

14-9

15-1

15.1
15.1.1
15.1.2
15.1.3
15.1.4
15.1.5
15.1.6

DESCRIPTION OF THE PARAMETER FILE
IDENT
HEAD1 and HEAD2
INPUT
COMPUTE
PRINT

15-1
15-1
15-2
15-2
15-3
15-4
15-5

15.2
15.2.1
15.2.2

OPERATING INSTRUCTIONS
Parameter File
PRINT

15-6
15-6
15-6

CHAPTER 11

CHAPTER 12

CHAPTER 13

CHAPTER 14

CHAPTER 15

Page
lS.2.2
lS.2.3
lS.2.4
lS.2.S

PRINT
CONVEX
Compilation
Program Execution

lS-6
lS-6
lS-7
lS-7

lS.3
lS.3.1
lS.3.2
lS.3.3

TEST CASES
Example 1
Example 2
Example 3

lS-7
lS-9
lS-12
lS-16

lS.4

ERROR MESSAGES

lS-17

lS.S

COMPILING PRINT

15-19

CHAPTER 16

LINCHG

16-1

APPENDIX A

COS CODES

A-I

APPENDIX B

LOADING COS

B-1

B.l

AUTOMATIC LOAD

B-1

B.2
B.2.1

TC08 DECTAPE AS THE SYSTEM DEVICE
System Restart on DECtape

B-1
B-2

B.3
B.3.1

DISK AS THE SYSTEM DEVICE
System Restart on Disk

B-3
B-4

B.4

TD8E BOOTSTRAP

B-4

B.S

PDP-12 USERS

B-6

B.6

RX8E USERS

B-7

APPENDIX C

SIZE OF CODE PRODUCED BY DIBOL COMPILER

C-l

APPENDIX D

DESIGNING A RECORD

D-l

APPENDIX E

BUILD CHECKDIGIT FORMULA

E-l

APPENDIX F

HARDWARE DESCRIPTIONS

F-l

F.l

HARDWARE COMPONENTS FOR COS-300 ARE:

F-l

F.2

HARDWARE COMPONENTS FOR COS-310 ARE:

F-3

F.3
F.3.1
F.3.2

VTOS TERMINAL
VTOS Start Up Procedures
Additional Features:

F-S
F-8
F-9

F.4
F.4.1

VT SaTE RMINAL
Additional Features

F-IO
F-13

F.S

CARD READER

F-14

F.6

LP08 LINE PRINTER

F-16

F.7

LS8-E LINE PRINTER

F-18

vii

Page
F.B
F.B.l

LPOS LINE PRINTER
Operator Controls

F-20
F-2l

F.9
F.9.l

LA30 DECWRITER DATA TERMINAL
Operating Procedures

F-23
F-24

F.lO

LA36 DECWRITER II DATA TERMINAL

F-27

F.ll
F.ll.l

HIGH-SPEED PAPER TAPE READER AND PUNCH UNIT
Papertape Positioning

F-30
F-3l

F.l2
F.l2.l

DECTAPE TRANSPORT UNIT

F-33
F-34

F.l3
F.l3.l

RKOB/RKOl DISK CARTRIDGE SYSTEM
Mounting and Dismounting the RKOB Disk
cartridge
Mounting
Dismounting

F-36

F.l3.l.l
F.l3.l.2
F.l4
F.l4.l

F-3B
F-3B
F-39
F-40

F.l4.l.l
F.l4.l.2

RKBE/RKOS DISK CARTRIDGE SYSTEM
Mounting and Dismounting the RKBE Disk
Cartridge
Mounting
Dismounting

F.l5
F.lS.l
F.lS.l.l
F.lS.l.2

RXBE/RXOl DISKETTE
Mounting and Dismounting the RXOl Diskette
Mounting RXOl
Dismounting RXOl

F-42
F-43
F-43
F-43

COS FILES

G-l

G.l

COS SOURCE FILES

G-l

G.2

COS DATA FILES

G-l

G.3

COS DIBOL COMPILER BINARY

G-2

G.4

COS SYSTEM PROGRAMS

G-2

G.S

SYSTEM AND DATA TAPE FORMATS

G-2

DIBOL DEBUGGING TECHNIQUE (DDT)

H-1

H.l

ENTERING DDT MODE

H-l

H.2

DDT COMMANDS

H-l

APPENDIX I

DIBOL STATEMENT SUMMARY

H-4

APPENDIX J

ERROR MESSAGE SUMMARY

J-l

APPENDIX K

COMMAND SUMMARY

K-l

K.l

COMMANDS FOR MONITOR FUNCTIONS

K-l

K.2

COMMANDS FOR EDITING FUNCTIONS

K-3

APPENDIX G

APPENDIX H

viii

F-41
F-41
F-41

Page
APPENDIX L

ADVANCED PROGRAMMING TECHNIQUES

L-l

L.l
L.l.l
L.l.2
L.l.3
L.l.4
L.l.4.1
L.l.4.2
L.l.4.3
L.l.4.4
L.l.4.S

ACCEPT AND DISPLAY
L-l
Background
L-l
Interaction of ACCEPT and DISPLAY Statements L-l
Simple Example Using ACCEPT and DISPLAY
L-2
Generalized ACCEPT Subroutines
L-2
RUBOUT and CTRLju
L-2
Sample CLEAR Program
L-4
Programmable Numeric Keypad
L-S
Special Function Keys
L-B
Escape Codes
L-B

L.2
L.2.1
L.2.2
L.2.3
L.2.4
L.2.S
L.2.6
L.2.7
L.2.B

DIRECT ACCESS TECHNIQUES
Background
The Reason for Direct Access
How It Works In DIBOL
Unsorted File
Sorted File
Rough Table, No Index File
Rough Table Plus Index File
Summary

L-B
L-B
L-9
L-9
L-9
L-ll
L-ll
L-13
L-13

L.3
L.3.1
L.3.1.1
L.3.1.2
L.3.1.3

DIRECT ACCESS NOTES
XMIT Statements Used With Direct Access
Truncating a File
Appending to a File
Rewriting a File

L-14
L-14
L-14
L-lS
L-1S

L.4

NUMERIC FIELD VERIFICATION

L-1S

L.S
L.S.l
L.S.2
L.S.2.1
L.S.2.2
L.S.2.3
L.S.3
L.S.4

CHAIN STATEMENT NOTES
Interaction of CHAIN and IN IT (n,SYS)
Communication Between CHAINs
File Status
Clearing CHAINed Records
Transferring Variable Values
Multiple CHAIN Entry Points
Miscellaneous CHAIN Facts

L-16
L-16
L-17
L-17
L-17
L-17
L-1B
L-1B

L.6
L.6.1
L.6.2
L.6.3

DIBOL PROGRAMMING OF SYS FILES
Operating Procedures
Data Section
PROCedure Section

L-19
L-19
L-19
L-19

L.7

VTSO ESCAPE SEQUENCES

L-20

DISKETTE UTILITY

M-l

APPENDIX M

GLOSSARY-l

GLOSSARY
INDEX

INDEX-l

FIGURES
Number
INTRO-1
INTRO-2
INTRO-3
1-1
3-1
3-2
4-1
4-2
6-1
11-1
14-1
14-2
14-3
F-l
F-2
F-3
F-4
F-5
F-6
F-7
F-S
F-9
F-IO
F-ll
F-12
F-13
F-14
F-15
G-l

Page
Business Application System
Typical Disk Operation Direct
Processing
Programmer's System
Sample Program
SYSGEN Flow Chart
Question and Answer Sequence
Compiler Listing
Storage Map Listing
Sample BUILD Control Program
DAFT Sample
Square Root Subroutine
CREF of Square Root Subroutine
Output of Square Root Subroutine
DS 300 Hardware Configuration
DS 310 Hardware Configuration
VT05 Terminal
VT50 Terminal
Card Reader
LPOS SO-Column Line Printer
LSSE Printer
LP05
LA30 DECwriter
LA36 DECwriter
High-Speed Paper Tape Unit
DECtape Transport Unit
RKOS Disk
RKSE/RK05
RXSE/RXOI
Monitor Organization

INTRO-3
INTRO-4
INTRO-5
1-3
3-2
3-5
4-4
4-6
6-2
11-6
14-3
14-6
14-7
F-2
F-4
F-6
F-l1
F-14
F-16
F-lS
F-20
F-23
F-27
F-30
F-33
F-36
F-40
F-42
G-l

TABLES
Number
1-1
1-2
1-3
2-1
11-1
A-I
F-l
F-2
F-3
F-4
F-5
F-6
F-7
F-S
F-9
F-IO
F-ll
F-12
F-13
F-14

Page
Source Program Limitations
Terminator Codes
Special Characters
Monitor Key Commands
DAFT Command Summary
COS Codes
VT05 Controls
VT50 Controls
Card Reader Switches and Indicators
LPOS Switches and Indicator Lights
LSS-E Controls
LP05 Controls
LA30 Controls and Indicators
LA30 DECwriter Specifications
LA36 Controls and Indicators
Specifications
High-Speed Reader Controls
High-Speed Punch Controls
DEC tape Controls
RKOS Controls and Indicators

1-5
1-7
1-17
2-3
11-1
A-I
F-7
F-12
F-15
F-17
F-19
F-21
F-24
F-26
F-2S
F-2S
F-31
F...;.32
F-34
F-37

PREFACE

This document is a source reference manual.
It is intended for
those
interested
in applying
the DIBOL language in a COS-300/310 system
environment.
Because of the COS - 300 and 310
systems
family
resemblance, most of the information contained herein applies to both
systems. Therefore, unless stated otherwise within this document, all
statements made in the COS-300/310 SYSTEM REFERENCE MANUAL apply to
both the COS-300 system and the COS-310 system.
Specific reference is
made to the particular system under discussion, whenever a difference
occurs.
The manual is constructed as follows:
DIBOL language program preparation is discussed
in Chapter
1.
Chapters 2,
through 5 cover the main systems material:
loading and
editing programs;
software system configuration and logical
unit
assignment, and movement of files and programs between devices.
Utility programs used
for building, sorting and maintaining data
files,
are found in Chapters 6 through 8, while Chapters 9 through 16
describe non-related utility programs.
The manual provides additional reference material and summaries in
Appendices A through M and
the glossary at the end of this book
contains a list of all possibly unfamiliar
terms.
Readers of this
manual are assumed to possess a basic knowledge of DIBOL.
Additional
background may be obtained
however,
by
consulting
COS - 300
SELF-TEACHING MANUAL (DEC-08-0CSTA-A-D).

INTRODUCTION

THE LANGUAGE
DIBOL (Digital Equipment Corporation Business Oriented Language) is a
COBOL-like language used to write business application programs. The
DIBOL language consists of data definition and procedures specified in
easy to learn statements. Figures INTRO-l and INTRO-2 illustrat~ a
possible application work flow using DIBOL and system programs.

THE COS SYSTEM
The COS-300 and COS-310 systems are of the same family and are
similar. The newer sophisticated COS-310 system however, is initially
designed for the small system user. Notwithstanding this, the COS-310
has expansion flexibility built into the system to accommodate the
degree of complexity normally associated with the COS-300 system.
The COS-310 system is streamlined for efficiency within its intended
environment.
Among other
things, this means that some of the
peripherals have been removed.
Paper
tape and card readers for
example, are not available on the 310. But, new devices have been
introduced to the system.
In particular,
the new flexible RX8E
Diskette comes with the 310, as does the VT50H terminal.
Both systems are discussed in detail throughout this manual.
minimum COS system hardware requirements are shown below:
COS-300 SYSTEM

PDP-8/E Computer
Mass
with
User
Line

storage device
16KB of core
terminal
printer

COS-310 SYSTEM

PDP-8/A
16KB of core
2 RX8E Dual Diskettes
User terminal
Line printer

INTRO-l

The

For a detailed description of the basic hardware and
supported equipment, refer to Appendix F of this manual.

additional

Also provided as part of COS are the following system programs:
MONITOR/EDITOR

controls the calling and executing of all
other programs in the COS system;
provides
the I/O control for
the peripheral devices
and an editing capability for correcting user
programs.
This program is
referred
to
throughout the manual as Monitor.

SYSGEN

specifies the
system
assigns logical device
startup.

COMP

translates DIBOL language source statements
into a binary object program which can be run
on the COS 300 computer.

PIP

transfers data, source
between two devices.

BUILD

creates data files based on the
data fields typed by the user.

SORT

sequences records according to key characters
or fields specified by the user.
Records may
be sorted into ascending
(0-9 or A-Z)
or
descending (Z-A or 9-0) sequence.

UPDATE

performs maintenance on data files.
UPDATE
makes the specified changes, creates a new
file and prints a report showing all changes,
deletions and insertions.

Figure INTRO-3 suggests possible usage of
DIBOL program development.

these

configuration
and
numbers for system

binary

system

files

contents

programs

COS FILE STRUCTURE
There are four types of files used in the COS system:
compiler binary, and system program.
Data files are completely devoted
to
processed by DIBOL or system programs.

the

storage

source,
of

data

data,
to

Source files contain control programs (for BUILD, SORT, etc.) or
DIBOL programs.

user

Binary files are the output of the compiler and contain user
programs translated into a code which can be executed by COS.

DIBOL

System program files include those programs
(MONITOR,
SYSGEN,
PIP,
BUILD, COMP, SORT and UPDATE) supplied as part of the COS package.
DIBOL source programs, control programs and compiled binary programs
can also be saved on system devices.
(Data files cannot be saved as
system files.)

INTRO-2

This figure illustrates
the possible flow
of
work from a
source
document (such as an
invoice
or
mailing
list)
through
the
creation of a data file
to the eventual output
of a report or listing.

Figure INTRO-l

INTRO-3

INPUT
PROGRAM

DIBOL

PROCESS MASTER
FILE TO PRODUCE
REPORTS
DIBOL

I
I

e
This figure illustrates
the input
of
data
(transaction file) from
terminal,
card reader
of DECtapei
the
processing against the
master file by a user
created DIBOL
program
and
the
output
(depending on the DIBOL
program being run)
of
reports and files.

Figure INTRO-2

INTRO-4

Transfers files,
Enlarges binary
scratch area

DIBOL
program
creation (also
SYSGEN tables
and control
programs) and
editing

Compiles
source
programs
into object
(binary) program

Dump and
Fix data
files

This figure illustrates the COS
system programs of particular
interest to the programmer.

Figure INTRO-3

INTRO-5

Cross
reference
of DIBOL
program

MANUAL AND SYSTEM CONVENTIONS
The symbols and terms used throughout this manual are described below.
SYMBOLS USED

EXAMPLE

Lower-case
characters

PROC n

Represent information that must
be supplied by the user, such as
values,
names
and
other
parameters.

Upper-case
characters

NO TRACE

Words or characters that must be
used exactly as shown.

... (Ellipsis)

RU prog(, ... ,progn)

Indicates optional
of arguments.

Underscored
char ac ter s

COS MONITOR 2.1108
Ind icates
system.
RU,-,PIP

START{~~ }

{}
[

START~~}]

]

..J( car r iage
return key)

EXPLANATION

.R. COMP, PROG~

output

continuation

from

the

Indicates a space or blank.

Braces indicate a choice of
of the items enclosed.

Brackets
feature .

indicate

one

optional

Line and command terminator.

TERMS
Expressions
Are variables, constants or arithmetic expressions (made
of variables, constants and the operators #,+,-,*,/).

Filnam,pronam,label and keyword
Are used to
identify names assigned
to files,
programs
statements, and
input lines,
these names may be of any
length but only the first 6 characters are recognized.
Pronam can contain any characters
comma, plus, period and minus.

INTRO-6

except

space,

slash,

PART I

DIBOL LANGUAGE

CHAPTER 1
DIBOL LANGUAGE

This chapter provides a reference source on the DIBOL language.
If
more basic information is desired, refer to the COS 300 Self-Teaching
Manual (DEC-08-0CSTA-A-D).
A DIBOL program consists of a series of
sections, Data and PROCedure.

statements

arranged

two

The Data section contains the RECORD statements which specify the type
and size of the information to be used in the program and where it is
to be stored.
The PROCedure section consists of:
ACCEPT
CALL
CHAIN
DISPLAY
END
FINI
FORMS
GO TO
IF
INCR
INIT
ON ERROR
PROC
READ
RETURN
STOP
TRACE/NO TRACE
TRAP
WRITE
XMIT
and data manipulation statements. These statements are arranged in a
logical order
for program execution.
Refer to the sample program
shown in Figure 1-1.
Use the editing commands (refer to Chapter 2) to
enter newly created DIBOL programs into the COS system.
Spaces and
tabs may be inserted in the program for clarity and are ignored by the
Compiler except when enclosed in single quotation marks (I).
Statement length is limited to 120 characters and spaces.
When the
121st character is typed, an error message LINE TOO LONG is displayed

1-1

and that line is lost.
CTRL/U deletes the entire line.
If the
terminal line is filled before the 120 characters are entered, the
Editor executes a carriage return/line feed to the terminal and the
line can be continued.

1-2

START
RECORD

RECORD

RECORD
PROC 2
LOOP,

EOF,

INBUF
STOCKN,
DESC,
UCOST,
QORDER,

D4
A25
05
04
09

OUTBUF,X

ECOST,
LABEL,

D4
A25
05
D4
09
A6,P

INIT(l,IN,LABEL)
INIT(2,OUT,'OUTPUT')
XMIT(l,INBUF,EOF)
IF(STOCKN.LT.lOOO)GO TO LOOP
ECOST=UCOST*QORDER
XMIT(2,OUTBUF)
GO TO LOOP
FINI(2)
FINI(l)
STOP

END

Figure 1-1
Sample Program

1.1

STATEMENT NUMBERS

Statement numbers are assigned manually by the user when typing in the
program or automatically by the Monitor when the program is entered
for editing (Refer to Chapter 2). These numbers control the order of
the statements in the file.
The OIBOL language makes no use of these
numbers, but error messages and the OIBOL debugging refer to them.

1.2

STATEMENT LABELS

Labels are the symbolic names assigned to identify the location
statement in the procedure section of a program.
TEST, A=B

t
(

label)

1-3

Labels consist of a sequence of letters or digits the first of which
must be a letter and only the first six characters are significant.
Labels are separated from the statement by a comma.
Statements with
labels are the transfer point for other statements such as GO TO, IF
and CALL.
Examples:
GO TO BEGIN

(

label)

BEGIN, A=B+4
t

(

label)

(

labels)

/,-,

GO TO (PRT,SEC,TO),K

- PRT, XMIT(lS,RPT,EOF)

(

label

(

yC,

READ(lO,CUST,88)

label

(

yT,

SUB=COST+TAX

label

1.3

COMMENTS

Comments and notes to explain the source program may be added
following a semicolon (i) on any statement line. A program line that
begins with a semicolon contains only comments.
Examples:
RECORD CUST i THIS IS THE CUSTOMER RECORD
i THIS PROGRAM PRINTS THE ACCOUNTS
iPAST DUE REPORT
Table 1-1 summarizes DIBOL source program limitations.

1-4

Table 1-1
Source Program Limitations

1.4

Maximum characters

about 8,000 per file

maximum number of
source files per
program

Maximum number of
symbols in 8K
systems

365

in 12K and
larger :'

511

Line numbers
available

0-4095

Maximum characters
per line

120

STATEMENTS

There are six types of statements in DrBOL:
1.

Compiler statements
(START,
END and PROC)
which label
the
beginning, end and procedure section of the program.
These
statements are non-executable and START and END are optional.

Data specification statements (RECORD) which describe the type and
size of data to be stored and the location in which it is to be
stored.

Data manipulation statements and INCR, which control the
of data within memory.

Control statements, (GO TO, IF,
CALL,
ON ERROR,
RETURN,
TRAP,
CHAIN,
STOP) which govern the sequence of execution of statements
within a program. Without these statements, the program would be
executed in the order written.

Input/Output statements (ACCEPT, DISPLAY, INIT, XMIT, READ, WRITE
and FINI)
which control data movement within memory or between
memory and peripheral devices, and to open and close the files
used by the program.
Line spacing to the line printer is
controlled by the FORMS statement.

Debugging statements,
program execution.

(TRACE, NO TRACE) which

1-5

permit

movement

tracing

ACCEPT
1.4.1

The ACCEPT statement takes input from the keyboard, stores it in the
specified alphanumeric field and causes the decimal equivalent of the
last key typed to be stored in the specified decimal field.
ACCEPT is
primarily used with the DISPLAY statement (section 1.4.5) and has the
form:
ACCEPT (dfield, afield)
where
dfield

is a dec imal
equivalent of
stored.

variable where the decimal
the last key typed is to be

afield

is an alpha variable
indicating
keyboard input is to be stored.

where

the

The code stored in the decimal field depends on the last key typed
(special terminator character). The keys and equivalent codes for the
allowable terminator characters are shown in Table 1-2.
For example,
when all characters are entered, filling up afield, the code 00 is
stored in the specified decimal field, which should be defined as two
or more characters.
This code can be tested later in the DIBOL program.
When an ACCEPT statement is encountered during program execution,
the
system waits for
keyboard
input.
Execution continues when the
characters typed fill the afield or when a special character such as
carriage return, line feed, altmode, rubout, CTRL/U, etc.
is typed.
Any non-COS character terminates the ACCEPT and sets dfield
to the
decimal equivalent of the last character typed. When the ACCEPT is
terminated before afield is full, the remaining character positions in
the afield are unchanged.
Examples:
Legal

Illegal

ACCEPT(A,B)
ACCEPT(A(3) ,B(4,5))

ACCEPT(5,CHAR)
ACCEPT (TCHAR,CHAR+3)

See Appendix L for more examples.

1-6

Table 1-2
Terminator Codes
Decimal
Char ac ter
00
01
02
04

05
06
07
08
09
10
11
12
13
14
16
17
18
19
20
21
22
23
24
25
26
27
29
30
32

VT50

Null or end of field
CTRL/A
CTRL/B
CTRL/D
CTRL/E
CTRL/F
CTRL/G
_+ OR CTRL/H
TAB
LINE FEED or LF
~ or CTRL/K
CTRL/L
RETURN or CR

Null or end of field
CTRL/A
CTRL/B
CTRL/D
CTRL/E
CTRL/F
CTRL/G
BACKSPACE or CTRL/H
TAB
LINE FEED
CTRL/K
CTRL/L
RETURN or ENTER
CTRL/N-(COS-3l0 only)
CTRL/P

CTRL/P
CTRL/Q-(COS-300 only)
CTRL/R
CTRL/S-(COS-300 only)
CTRL/T
CTRL/U
CTRL/V
CTRL/W
-+ or CTRL/X
CTRL/Y
" or CTRL/Z
ALT
HOME
EOL
RUB OUT

Code

Character

VT05

CTRL/R

CTRL/T
CTRL/U
CTRL/V
CTRL/W
CTRL/X
CTRL/T
CTRL/Z
ESC
not used
not used
DELETE

*Certain codes should not be used because
software functions.
These are either
marked with an asterisk.

they conflict with other
omitted from the table, or

A command protocol is built around the Escape code (033) (27 dec imal)
to
implement those commands needed by the VT50 but not found in 7-bit
ASCII.
Upon receiving the Escape code 033, the terminal
is set to
Escape mode and treats the next character received as a command.
Commands created in this manner are called Escape Sequences.

1-7

1.4.2

CALL

The CALL statement causes control to branch to a
the form:
CALL

subroutine

and

has

label

where label is the label of the first statement of a subroutine in the
PROCedure section of the program.
The CALL statement saves the return location in a
(pushdown)
list.
Additional
subroutine CALL and RETURN statements may be nested within
a subroutine to a depth of 50.
Examples:
Illegal
CALL SET

CALL
CALL (SUBR+3)

SET, PROFIT=PRICE-COST

RETURN

1-8

1.4.3

CHAIN

The CHAIN statement allows a DIBOL program which will not
available memory to be split up into two or more programs.

fit

The form of the CHAIN statement is:
CHAIN decimal expression
where decimal expression refers to the position of the
in the Monitor RUN command and is in the range 0-7.

CHAIN

program

When the CHAIN statement is encountered in a DIBOL program, execution
of the current program is halted and the CHAIN program referred to by
the decimal expression in the CHAIN statement is loaded.
Execution of
the CHAINed program begins with the statement immediately following
the PROC statement. All CHAIN programs must be properly declared in a
.RUN command. The program initially run is chain 0, the next program
chain 1, etc.
For example, a CHAIN 2 statement refers to program TRY2
in the .RUN command •
. RUN START+HELPl+TRY2+STOP3
Refer to section 2.2.6 for a description of the RUN command, section
1.4.16 for
the RECORD ,C statement and to Appendix L for more
information about the CHAIN statement.

1-9

DATA MANIPULATION STATEMENTS
1.4.4

Data Manipulation Statements

Data manipulation statements clear data fields, move data between
fields, calculate decimal expressions, convert data from decimal to
alphanumeric and vice versa and format data.
The form of this
statement is
destination field = source field
where source field is a field name, variable, constant or expression,
and destination field is the field where the data is to be stored.
The contents of the source field are moved to the destination field.
The destination field must have been defined in the data definition
section as either an alphanumeric or decimal field.
If
the
destination or source field is decimal, the data from the source field
is right-justified; if both fields are alphanumeric, left-justified.
The source data is always converted to the type defined for the
destination field. Record names can be used in data manipulation
statements.
Arithmetic Expressions:
Arithmetic expressions are allowed on the right side of the equals
sign.
The expression can contain decimal elements, subscripted data
elements, constants, variables and arithmetic operators.
There are
five binary arithmetic operators:
# (rounding)
/ (division)
* (multiplication)
+ (add i tion)
- (subtraction)

Order of priority:
1. round ing
2. multiplication and division
3. addition and subtraction

These operators require decimal operands.
Operators with the same priority are executed left to right.
The result of a division operation
example, 5/3 is 1 and -14/5 is -2.

truncated

toward

zero.

For

An error occurs when a divide by zero is attempted or when the result
of an addition, subtraction or multiplication is too large or too
small.
There are three unary operators (-,+,#). The (+) has no effect;
the
unary - negates the decimal value to which it is specified. More than
one unary minus can be specified for a number, e.g., A= - -3 is
equivalent to A = +3 and A = --+-4 is equivalent to A
-(-(+(-4))) or
A = -4.

The number sign (#) operator is also used to convert an alphanumeric
character to its equivalent internal code and make that decimal number
(0-63) available to the program.
The # operator takes that value from the alphanumeric or decimal field
on the right side of the expression, checks the first character

1-10

(left-most), determines the equivalent internal code and returns the
code as a decimal number. The result is a decimal value. In this
application, the number sign (#) appears before the character.
Examples:
A

A is a decimal field and B is an
alphanumeric field.
If B contains the
characters XYZ, X is converted to its
internal
code, 71, and the decimal
equivalent, 57, is stored in A.

3 + #B

A is a decimal field and B is an
alphanumeric field.
After conversion,
the decimal equivalent of the first
character of B, 57, is added to 3 and
stored in A.

A = 10
Fl = 100*A/2+3-1
The order of priority can be altered by the use of parentheses, since
expressions in parentheses are executed first. The above expression
equals 502.
Fl=100*(A/2+3-l)
The result of this example is 700.
Illegal
A+(B*C
C*3)-F
INTRST*((4*AMT+2)-PAYMNT/2
LENGTH+WIDTH+
6/3005434+'ABCDE'
Rounding:
Variables may be truncated and rounded.
The resulting variable is
rounded upward if the digit which followed the remaining variable
before truncation is 5 or more. The variable is rounded in terms of
magnitude; sign is unchanged.
The format for rounding is:
A#B

where A is truncated by B places and rounded in magnitude.
B cannot
be greater than 7 and should be a positive integer. If B is negative,
it is treated as positive. When used for rounding, the number sign
(#) appears after a character. For example:

1-11

MONEY,

TEMP=MONEY#2
If MONEY equals 123456, TEMP equals 1235.
If MONEY equals 123446, TEMP equals 1234.
If MONEY equals -1473, TEMP equals -15.
Typically, this feature would be used for rounding to the dollar.
variables:
Variables may take on one of the following three formats:
a.
b.
c.

name
name (subscript)
name (position 1, position 2)

name - consists of a sequence of letters and digits beginning with
a letter, only the first 6 characters are significant. Used in
record name, field name, etc.

name(subscript) - name should be a subscripted array and the value
of the subscript must be between 1 and the dimension specified in
the array.
The subscript must be decimal and can be
an
expression.
If either name is not an array or the value of the
subscript exceeds the dimension other locations in core outside
the limits of name are referenced. No error message is generated
unless the user tries to reference data so far away from his data
area that he is referencing his program or the system.

name (positionl, position2) -positionl must be less than or equal
to position2;
position2 should be less than or equal to the
maximum size of the array associated with name.
positions 1 and 2
must be both positive and decimal. This form of subscripting
references those characters from positionl to position2 inclusive.
If the variable name is subscripted, you may consider
the
successive array elements strung out left to right.
For example,
in referencing A(3,9) in the array A declared as 4D4, the 7 digits
are referenced as follows:
the last 2 digits of A(l), the entire
A(2), and the first digit of A(3) •

Decimal Literals:
These consist of a sequence of from 1 to 15 decimal digits.
Alpha literals:
These consist of a sequence of any legal COS characters (except single
quotes) enclosed in single quotes.

1-12

Example:
'ABCD'
RECORD Literals:
A RECORD literal is a constant that can be used anywhere in the
PROCedure section of a DISOL program where a record is required. It
is similar to an alpha literal except it begins with a double quote
(II) and ends with a single quote ('). For example:
REC=IITHERE'
INIT (8,TTY)
XMIT (8,IIHELLO')
XMIT (8,REC)

1.4.4.1

Clearing Fields

Data manipulation statements clear fields when used in the form:
field name =
The destination can be a single field or an array.
If alphanumeric,
it is cleared to all spaces; if decimal to all zeros. Fields, parts
of fields or fields in an array, can also be cleared by the use of
subscripts (expressions in parentheses). Any part of an array can be
accessed in a program statement by listing the position of the field
of character(s} in parentheses. Specifying an array name without any
subscripts, clears the first array element only.
Examples:
Meaning

Field Name
Fl(S,7}

clears character positions 5,
field Fl

and

Fl(S}

clears the fifth element in an array

Fl(A}

clears the Ath element in array Fl

Fl(l,l}

clears the first character.

To clear a RECORD area to all spaces use the form:
RECNAM=
Record names can be subscripted to allow reference to RECORD areas as
though they were in an array. All records to be so referenced must
follow one another and be of the same length. For example:

1-13

RECORD Z
,A3
RECORD Z2
,A3

Z(2)=
Clears the next record defined after record Z in the data
the program.

1.4.4.2

section

Moving Alphanumeric Data

To move the contents of one alphanumeric field to another alphanumeric
field, use the following form:
Alpha destination = Alpha source
If the source is shorter
than the destination,
the
data
is
left-justified with the right-most characters of the destination field
undisturbed.
If the source is longer
than the destination,
the
right-most characters of the source field
are not moved into the
destination field.
The source field retains the original data.
For example:
RECORD
A,A5,'ABCDE'
B,A3,'FGH'
PROC
A=B
A now contains

FGHDE

RECORD A
NAME,A4,'FRED'
NAME1,A7,'JOHNSON'
PROC
NAME=NAMEl
The name in RECORD A is now JOHNJOHNSON.

1.4.4.3

Moving Decimal Data

Use a statement of the following form to
decimal field to another decimal field.

move

the

contents

one

decimal destination = decimal source
If the value in the source field is shorter than the length of the
destination field,
zeros are inserted on the left.
If longer, the
most significant digits are not moved.

1-14

1.4.4.4

Moving Records

An assignment statement may be of the form
RECl=REC2
where RECI and REC2 are record names or subscripted record names or
REC2 may be a record literal.
Its effect is to move the contents of
REC2 into the space reserved by RECI.
If REC2 is smaller than RECl,
the right-most characters of RECI are undisturbed.
If REC2 is larger,
the right-most characters of REC2 are not moved.
Extreme caution must
be used if the destination record is subscripted.

1.4.4.5

Converting and Formatting Data

Decimal values can be
purposes of I/O and
to contain spaces and
records on DECtape or

converted to alphanumeric and vice versa for the
calculations. Any data field may be reformatted
punctuation marks which are not stored with the
disk and cannot be present during calculations.

Alphanumeric to Decimal Conversion - The form used for this conversion
is:
Decimal variable = Alpha expression
Alphanumeric values to be converted to decimal must be 16 or less
characters in length.
The source is an alphanumeric expression which
contains an optionally signed representation of a number.
The
destination is a simple or subscripted decimal variable.
Example:
DEC=ALPHA
The numeric alphanumeric field (ALPHA) is converted
to decimal
and
stored in the decimal field (DEC).
If the decimal field is larger,
the data is right-justified and zero-filled.
If the decimal field
is
smaller,
the excess characters to the left are not moved.
If the
alpha field contains characters other than digits, spaces and signs (+
or-), the message BAD DIGIT results at run time.
The statement should
be preceded by an ON ERROR statement if the contents of ALPHA may
contain BAD DIGIT's
(refer
to the ON ERROR statement for more
information).
Spaces and signs are not counted as characters that are
moved.
Spaces in the alpha field are ignored.
Signs may be imbedded
anywhere in the alpha field.
Two minus signs are equivalent to a
plus;
three minus signs are equivalent to a minus; etc.
Decimal to Alphanumeric Conversion - The form of this conversion is
Alpha variable = decimal expression
The source is a decimal expression and the
alphanumeric variable.
Example:
ALPHA = DEC

1-15

destination

simple

The contents of the decimal field (DEC) are placed in an alphanumeric
field
(ALPHA).
If the alphanumeric field
is larger, the data is
right-justified and space-filled on the left.
If the alphanumeric
field
is smaller, the excess characters on the left are not moved.
Leading zeroes in the decimal field are converted to spaces. However,
if the decimal field equals zero, the alpha field is set to spaces
except for the right-most character, which is a zero.
Data Formatting - Decimal fields can be
alphanumeric fields using the form:

reformatted

and

stored

Alpha variable = decimal expression, format
The source is a decimal
alphanumeric variable.
inserted with numbers.

expression and
the destination a simple
Format specifies special characters to be

Example:
A = D,

'-XXX, XXX. ZZ '

The eight-digit decimal number at D is converted to alphanumeric code,
reformatted with specified punctuation and stored in alphanumeric
field A.
The format string must be an alpha expression.
The formatted
value
is
placed
in
the
destination
field,
right-justified and padded with spaces if necessary.
If the formatted
value is larger than the destination field, the left-most characters
are not moved.
Most printing characters on the line printer or terminal can be used
in a
format string but the following characters must be used with
care: X,Z,*,-,.,', and comma.
Table 1-3 shows the special characters
to be used in format strings.
Examples:
TOTL=TEMP,'XXX,XXX.XX-'
If TEMP contains 12345678, TOTL will contain 123,456.78
AMT=PAY,'*XXX,XXX.XX'
CDATE=DATE,'XX/XX/XX'
GTOT=TOTl+TOT2,'-XXX,XXX.ZZ'
WAGES=RATE*HOURS,'XXXX.XX'
RATIO=(TOTI/TOT2)#2,'XXX.XX'
TOTI is divided by TOT2;
the result is
result is formatted and stored in RATIO.
Al=NUM, 'XXXO '
If NUM contains 987, Al will contain 9870.
A2=DATE,' . XX/XX/XX ,
A=Q, 'Q=XX'

1-16

rounded

two

places;

that

Also:
TOTFMT, All, '-XXX,XXX.ZZ'
GTOT=TOTl+TOT2, TOTFMT
TOTFMT is specified in data section and is used to specify
the PROC section.

format

When using a comma, period, slash, minus sign or any other notation,
each must be counted as a character position. In the first example
above, for instance, TOTL must be defined in a RECORD statement as an
eleven-character alphanumeric field.
In the example A=B,C, B contains a decimal value being converted to
alpha and stored ln A (an alpha field). The C represents an alpha
field which contains the format strings to be used in the command.

Table 1-3
Special Characters
Character

Explanation

Used in a format to arrange a decimal field for
printout.
Each X represents a digit and leading zeros
are automatically suppressed.

Special character
formatting output.

Used in a format string to replace leading zeros and
eliminate trailing spaces on printout. If the * is not
the first character in the string. digits may also be
replaced.

Inserts an arithmetic sign in a number to be printed.
The sign may be placed before or after the number.
If
the number is positive, a space is inserted where the
is placed. If the hyphen is placed in a position
following the first significant digit, but previous to
the last position of a format string, it is printed
like any other insertion character.

used

suppress

digit

Inserts a decimal point in a format string and
zeros
to the right of the decimal point
significant.

when

forces
to be

Used to insert a comma in a format string if there
significant digits to the left.
All other characters
characters.

1-17

are

treated

are

insertion

IDISPLAY I
1.4.5

DISPLAY

The DISPLAY statement is used with the VT05 terminal under both the
COS 310 system and
the COS 300 system but is used with the VT50
terminal only under the COS 310 system.
DISPLAY is used to show
messages on the scope and to move the scope cursor to the specified
line and character position.
The DISPLAY statement is a means of
displaying questions on the scope and storing the replies.
The form of the DISPLAY statement is:
DISPLAY (y ,x,

{!~~:f~l'}
dfield

where
y

is a decimal expression representing
the scope
line number (1-20).
If the line number specified
is not within 1-20, the results are unpredictable
when used with VT05.
If the line number specified
is greater than the number of lines on a VT50, the
cursor
is positioned
to the last line of the
screen.
NOTE
If Y is zero, no positioning is done and x
is ignored. The statement with y equal to
o is used to output a message without a
carriage return/line feed
i.e.
DISPLAY
(0 , 0, 'NAME' ) .

is a decimal expression representing the character
position
(1-72).
If the character is not within
1-72 the results are unpredictable.
If the line
number specified is greater
than the number of
lines on a VT50, the cursor is positioned to
the
last line of the screen.

'literal'

is an alphanumeric or decimal character string.
An alphanumeric string must be enclosed in single
quotes (')
and
is displayed at the character
position
specified.
There
is
no carriage
return/line feed after the message and the cursor
remains in the character position at the end of
the message.
A decimal literal must be
codes:

one

O=position cursor.
l=clear to end of scope.
2=clear to end of line.
25=emit a bell or beep sound
7=emit a bell or beep sound

1-18

the

following

(COS 300 only).
(COS 310 only) .

The above codes are the only special
codes
recognized by the COS - 300 and COS-3l0 systems.
The use of other codes is possible.
However,
equivalent results will not be guaranteed on
different or future versions of COS-300
and
COS-3l0.
afield

is a field defined as alphanumeric which
a message to be displayed.

dfield

is a decimal defined field containing a code that
either causes a particular operation to occur or
causes a particular character to be displayed.

contains

The special codes are as follows:
O=position cursor.
l=clear to end of scope.
2=clear to end of line.
25=emit a bell or beep sound.

(COS-300 only)

NOTE
The above codes are the only special codes
recognized by the COS - 300 and COS-3l0
systems. Any other codes found in the 300
field produce unpredictable results. When
using COS-3l0 system, codes are converted
into octal and transmitted to the terminal
as ASCII characters. The 310 system uses
code 007 to ring the bell.
The 300 system
uses code 25 to perform the same function.

Examples:
DISPLAY(O,0,25)

iring terminal bell.

(COS-300 only)

DISPLAY(0,0,7)

iring terminal bell.

(COS-3l0 only)

DISPLAY(l,l,l)

iclear entire scope.

DISPLAY(lO,l,l)

iclear from character position 1
iline 10 to end of scope.

DISPLAY(2,20,DAY)

idisplay the contents of
ialpha field DAY on line
i2 starting at character
iposition 20.

DISPLAY(l,lO,O)

iposition cursor at first
iline, 10th character position.

DISPLAY(ll,1,2)

iclear line 11 starting at
icharacter position 1.

DISPLAY(11,37,2)

iclear second half of line
ill.

1-19

DISPLAY(OO,'DATE')

idisplay the word DATE starting
;at the current cursor position.

DISPLAY(11,12,25)
(COS-300 only)

;position cursor at line 11,
;character position 12 and
;emit beep.

DISPLAY(11,12,7)
(COS 310 only)

;Position cursor at line 11,
icharacter position 12 and
;emit beep.

DISPLAY(Y,X,BEEP)

iposition cursor at coordinates
idefined in X and Y and
;execute code defined in
iBEEP. If the programs
;data division contains
RECORD
Y,D2,20
X,D2,36
BEEP,D2,25 (COS 300 only)
(For COS 310 replace above line with
BEEP, Dl,7)
;the cursor is set at line 20,
;character position 36 and an
;audible beep is sounded.

DISPLAY(20,36,FIRST)
;if the data division contains
RECORD
FIRST, A4,'NAME'
;NAME is displayed at line
;20 character
;position 36. The cursor is
;left in the 40th position of
i 1 ine 20.
DISPLAY(l,l,'HELLO')
;display HELLO in the
;upper left hand corner
;of the screen.
DISPLAY(lO,lO,BEEP)

;If the data division contains
RECORD DISP
POS'l', D2, 00
CSCOPE,D2,01
CLINE,D2,02
BEEP,D2,25(COS 300 only)
BEEP,D2,07(COS 310 only)
;the cursor is placed at
;character position 10 on line
;10 and an audible beep
;is emitted.

See Appendix L for more examples.

1-20

1.4.6

END

This is an optional statement and is used to indicate the
of a program. The form of the statement is

termination

END [Ix]
where

is an optional switch to the compiler

no compiler listing of source program
map

List source
printer

List source program storage map on terminal.

program

storage

map

storage
on

line

If no options are specified, a compiler listing is
produced as usual. The END option switch or lack
of it can be overridden with an option switch in
the compiler RUN command. If no symbol table is
printed, the symbol count and number of free
locations are not printed.

1-21

1.4.7

FINI

The FINI statement writes an end-of-file mark, closes the file,
rewinds the tape (if DECtape) and writes the file length in the file
label (wr i te only).
(In mul ti-reel files the file leng th is wr i tten
only on the last reel.) The form is:
FINI (n)
where n is a decimal expression whose value is 1-15 specifying a
channel number which was previously specified in an INIT statement.
FINI disassociates the channel number from the device specified in the
INIT statement.
(See Section 1.4.12.) Assignments of channel numbers
and devices can then be changed with another INIT statement. FINI is
necessary for mass storage output files only but it is considered good
programming practice to FINI each data file INITed. If specified for
an input file (other than from SYS), reading of that file stops. The
file may again be read (from the beginning), if another INIT statement
is executed.
Examples:
Legal
FINI(l)
FINI (005)
FINI (A+B)

Illegal
FINI
FINI (
FINI(4
FINI,3
FINI(O)
FINI(16)

If an output file is not INITed, the end-of-file mark is not written
and records may be lost. If an update file is not FINIed, the last
few records written may be lost.

1-22

1.4.8

FORMS

The FORMS statement is used to format line printer
has the form:
FORMS

(only)

output

and

(channel, skip-code)

where
channel

is a decimal expression
(value 1-15)
associated
with the line printer in a previous INIT statement
(refer to Section 1.4.12).
If
the
channel
specified
is not associated with a line printer,
the statement is ignored.

skip-code

is a decimal expression with a value of 0 to 4095
which causes the line printer to go to the top of
the page or skip the number of lines specified.
The codes are:

o
1-4095
-1

-2

-85

go to top of next page
skip this number of lines
skip to channel 2
(LS8E printer only)
and
move paper forward n lines as
specified on vertical
forms
control
tape.
print enlarged characters (LS8E printer
only)
for
next XMIT statement. Allows
only half as many characters per line as
normal size printing.
sound alarm bell (LS8E printer only).

Other
negative
numbers
cause
unpredictable
results.
The skip-code is compiled modulo 4096.
For example, FORMS(6,4096) causes a top of form to
be executed.
Example:
INIT(l,L)

FORMS(1,3)
1 is the channel number specified in a previous INIT statement
is the number of lines to be left blank.

1-23

and

INIT(5,L)

FORMS(5,O)
5 is the channel number and

° specifies top of line printer page.

INIT(5,L)

FORMS(5,-1)
5 is the channel number and -1 causes the page to move up the number
of lines specified on the vertical forms control tape of the LS8E
printer.

1-24

1.4.9

GO TO

The GO TO statement has two forms:
(1) unconditional and
(2)
computed.
It transfers control to the statement whose label is
specified in the GO TO statement.

1.4.9.1

Unconditional GO TO

The form of the unconditional GO TO is:
GO TO label
where
label

is a statement label assigned to the statement in
the PROCedure section where control
is to be
transferred.

Example:
Legal

Illegal
GO TO
GO TO,TARGET

GO TO SET

SET, TOT=TOT+SUB

1.4.9.2

Computed GO TO

The computed GO TO has the form:
GO TO (labell ,labe12, •.. ,labeln) ,var iable
where
labell,labe12, ... ,labeln are statement labels.
number of labels.)
variable

(There can be any

is a decimal variable
representing a value.

expression

Control
is
transferred
to
the
statement
labeled
labell,
labe12, ... labeln if variable has the value 1,2, ••. n respectively.
If
variable is negative, zero or greater
than the number of labels,
control passes to the next statement in sequence.
Example:
GO TO (LOOP,LIST,TOT) ,KEY

1-25

KEY acts as a switch transferring control to LOOP if
KEY=2, or TOT if KEY=3.

KEY=l,

Other examples:
Legal
GO TO(LABl,LAB2,LAB3) ,VAR
GO TO(TRY,LOOP2,TRY,TRY,TOP) ,Q+l
GO TO (TEST) , IN
GO TO(Ll,L2,L3,Ll,L2,L3,OVER,L2,L3,LA83,W,Ll) ,NOW
GO TO(EI,EI,C) , (TEST+2)*1-FLAG
ILLEGAL
GO TO(
GO TO(LABl,LAB2
GO TO(Ll,L2)
GO TO(Dl,D3),
GO TO(Dl,D4)KEY
GO TO() ,COUNT
GO TO(SO,LO,) ,PEAR
GO TO(Ll"L3) ,KIT

1-26

LIST

1.4.10

An IF statement conditionally executes certain statements on the basis
of the result of a comparison of expressions.
The form of the
statement is:
IF (expressionl.rel.expression2)stmnt
where
expressionl,expression2

are
a
combination
of
constants,
variables or arithmetic expressions of
the same type.

• reI.

is one of the following
relational
operators
.EQ.
Equal
.NE. Not equal
. LT.
Less than
.LE.
Less than or equal
.GT. Greater than
.GE. Greater than or equal

stmnt

is a control statement which is executed
if the relation is TRUE.
"stmnt" must
be one of the following
GO TO label
STOP
CALL label
TRACE
RETURN
NO TRACE
ON ERROR label

The two variables must be of the same type:
both decimal or both
alphanumeric.
Numeric data may be in decimal or alphanumeric form.
In a decimal comparison the shorter field is internally filled to the
length of the longer field, then the comparison is made between the
longer field and the zero-filled field.
In an alphanumeric comparison
the compare is made on the number of characters in the shorter field,
i.e.
in comparing a 3 character field and a 5 character field, only
the 3 left-most characters of each field are used.
If the result of comparison is false, the next statement
is executed.
Examples:
Legal
IF (A .EQ.
B) GO TO LABEL3
IF (SLOT.NE.
2) CALL BAD
IF (SALES. LT. PROFIT+TAX-RENT) NO TRACE
IF (A+B+C
.GT. D+E*F) RETURN
IF(NUMBER.LT.D+3) CALL FIXUP

1-27

sequence

Illegal
IF
IF
IF
IF
IF

(C.LE.76)
(D2.EQ. 'ABC') THEN GO TO RAL
(3.NE.5) D='FALSE'
(Q.LE.R), GO TO TRADE
(DIST.GT.RATE*TIME) GO TO (Ll,L2) ,KEY

IF
IF
IF
IF
IF
IF

)E.GT.8) GO TO LAND
(ABC ,GT.
DEF) GO TO CHI
(TIME.GT:lO) GO TO BED
(P-l .LT.
0, GO TO UPDATE
(D.LT.l .AND.
D.GT.-3) GO TO NOW

IF (COST .GR.
IF (TAX .LTE.

EXPNSE) GO TO WR
100) GO TO BUY

1-28

1.4.11

INCR

The INCR (increment)
has the form:
INCR

statement adds 1 to the

specified

variable

and

variable

where
variable

is a decimal variable to be incremented by 1.

INCR can be used with positive numbers only and cannot be the object
of an IF statement.
INCR is typically used to add one to a counter
and is faster than a data manipulation statement.
For example:
A2=A2+1
or
INCR A2
Illegal

Legal
INCR A2 (3)
INCR B (5)
INCR C (H, 6)

IF (A.LT.16)INCR B
INCR A+B
INCR 3

1-29

1.4.12

IN IT

The INIT statement associates a channel number with a mass storage or
character oriented device for
input or output and initializes the
device.
The form of the INIT statement is:
INIT (channel, dev [ ,da ta file name] [, un i t] )
where
channel

is a decimal expression which evaluates to a number
1-15 specifying a channel
to be associated with a
logical or physical device.
(Refer to section 3.3 for
logical unit assignments.) This number is then used in
other statements such as XMIT to refer
to the same
device.
If the number specified exceeds 15, it is taken
16.

modulo

If not specified, the following channels are
initially
associated with devices at program startup.
(The user
may change these assignments with an INIT statement.)
5
6

7
8

Paper Tape Reader (COS 300 only)
Line Printer
Keyboard
Terminal scope or printer

A channel that is associated with a logical
be FINIed prior to another INIT.

unit

must

dev

is the designation for the COS device to be associated
with the channel number.
These designations can be
abbreviated since only the first character is checked.
The designations are:

Designation

Abbreviation

INI

Mass storage device to be used for input.

OUT

Mass storage device to be used for output

UPDATE

Mass storage device to be used for both input
(See READ and
and output
(direct access)
WRITE)

KBD

Input from terminal keyboard

TTY

Output to terminal printer or display

LPT

Line Printer

CDR

Card reader

PTP

Paper tape punch (COS 300 only)

Meaning

(COS 300 only)

1-30

RDR

Paper tape reader

SYS

Input from a file located on the system
device.
This file name must have been
(Refer
to
specified in a RUN statement.
Appendix L, Section 6.0.)

(COS-300 only)

data file name is an alpha constant or variable that identifies
the data file on the logical COS unit. A data
file name is necessary with the I.
O.
and U
devices and
illegal with other devices.
If the
name is more than six characters only the first
six are used.
If less than six, the name is
left-justified and padded with spaces on the
right.
It can also be the name of a field defined
in the data section with the P option which allows
flexibility in the files specified to the program.
unit

is an optional decimal expression used with 1,0
and U device codes which specifies the logical
unit (1-15) where the data file name is stored or
to be stored.
(If the number specified exceeds
15, it is taken modulo 16). On input this unit is
checked for
the data file name if it is not
present, a MOUNT message is displayed;
on output
this unit is checked for the data file name or a
temporary (scratch file) name, if another file
is
present, a REPLACE? message is displayed.

The Monitor uses the name specified in the INIT statement in the run
time MOUNT message
(refer
to Chapter 2).
The reply to the MOUNT
message is the logical COS unit where the file is stored (INPUT) or to
be stored
(OUTPUT). The logical unit is then checked by the MONITOR
to be sure the named file is indeed stored there.
In the case where a
logical unit is specified in the INIT statement, the Monitor checks
the logical unit for the named file and if found bypasses the MOUNT
message.
If not found on that logical unit the Monitor displays the
usual MOUNT message.
Examples:
INIT (15,I,'RENEW')
initializes channel number 15 for. input,
the input file.

1-31

(I).

"Renew" is the name

RECORD D
TAPEID,A6,P

PROC

INIT (3,IN,TAPEID)
The previous code causes the message ENTER TAPEID to be displayed.
The user responds with the name of the file;
a MOUNT message is then
displayed and must be answered with the proper logical
unit number.
This allows different file names to be specified each time the program
is run.
INIT(l+l,IN,FILEl)
initializes channel 2 for input (I).
Legal

Other examples:
Illegal

INIT(l,IN,'FILEOl')
INIT(2,INPUT,'FILE02')
INIT(3,I,'FILE05')
INIT(4,IXYDREQW, 'LABEL')
INIT(5,OUT,'TAPEl')
INIT(6,OUTCH, 'TAPE2')
INIT (7 ,O,RCRD)
INIT(l,UPDATE,'LABEL')
INIT (3 ,UPD,' L')
INIT(6,U,'L3')
INIT(6,UP,'T453')
INIT(5,KBD)
INIT(5,TTY)
INIT(4,K)
INIT(l,TELETYPE,NAME)
INIT(9,LPT)
INIT(lO,CDR)
INIT(ll,RDR)
INIT(12,PTR)
INIT(N,SYS)
INIT(15,IN,'MASTER' ,A+B)

INIT
INIT (
INIT(4
INI'r<4 >
INIT,3

1-32

ION ERRORI
1.4.13

ON ERROR

This statement can be inserted in the source program to cause a branch
to the statement with the specified label when a non-fatal executed
error occurs in the next program statement.
The form of this
statement is:
ON ERROR
where

label

label is a statement label assigned to the statement in
PROC section where control is to be transferred.

the

The ON ERROR statement can be the object of an IF statement.
Examples:

ON ERROR ERRTRP
TEMP=lOO*ORIG/NEW

ERRTRP,

ON ERROR TRAP
DEC = ALPHA

TRAP,
The ON ERROR statement eliminates a return
following run time errors.

the

Monitor

for

the

zero divisor
bad digit in an alpha to decimal conversion
more than 15 digits in decimal field used in calculation
an end of file label was not specified in an XMIT statement
input line overflowed block it was read into

1-33

no file specified in
statement

RUN

command

satisfy

INIT

(SYS)

in a random-access operation (see READ and WRITE)
the file
contains variable length records, record number is too large
or 0, or length specified in the record header word does not
match the length of the XMIT block.

1-34

1.4.14

PRoe

The PROe statement separates the two sections of a DIBOL program,
the
data definition section and the procedure section.
It is of the form:
PRoe [n] [Ix] [icomment]
where
n

is a
single digit,
0-7,
(not an expression)
indicating the maXimum number of mass storage
logical units which the program will have open
simultaneously.
If no number is specified, the
compiler assumes 7.
The available core is divided
into buffers to handle the number of logical units
specified.
The more buffers
necessary,
the
smaller
they must be and the slower
the 1/0
process.

is an optional listing switch to the compiler
IN

temporarily halts compiler listing
program

source

resume listing source program
line printer

and

errors

resume listing source program
terminal

and

errors

The PRoe option switch is active until disabled by
a START or END statement with an option switch.
There can be only one PRoe statement.
If no
option switch is specified, a compiler listing is
produced as usual.
The PRoe option switch can be
overridden with an option switch at compile time.
(Refer to Chapter 4.)
comment

is an optional string of text preceded by a
semicolon which is stored for output as a heading
for the procedure section of the compiler listing.
When the compiler encounters the PROe statement
the line printer moves to the
top of the next
sheet of paper and outputs the comment, if any, as
a header line.

If IN is specified in the RUN eOMP command, it overrides any IL's or
in the source program.
However, IL and IT determine on what
device errors are listed.
If the program simultaneously opens more
mass storage logical units than were specified in the PRoe statement,
a run time error occurs.

IT's

1-35

Examples:
Legal

Illegal

PRoe
PRoe 3;BILLING V3
PRoe 3/N
PRoe;
PRoe 4/L; TEST PROG.

PRoe 32
PROe, 4
PROe 5+2
PROe BILLING V3
PROe A

1-36

1.4.15

READ

The direct access READ statement allows a specified data record to be
moved from a named file to a specified area in core and has the form:
READ (channel, record,rec#)
where
channel

is a decimal expression with a value of 1-15
specifying a channel number which links the READ
statement to the related
INIT statement.
(The
INIT statement must specify INPUT or UPDATE as the
COS dev ice.)

record

is the record into which data is to be read.

rec#

is a decimal arithmetic expression specifying
sequence number of the record to be read.

the

If the program READS past the end-of-file mark,
the results
undefined.
See Section 1.4.22 for restrictions on usage.

are

Examples:
READ (5,REX,88)
reads the 88th record of the device linked to the channel which was
INITed with the INIT (5, .•. ) statement and places it in the core area
labeled REX.
READ (6,BLT,EXPR)
reads the record specified by the expression EXPR and stores it in the
core area labeled BLT.
Refer to Appendix L for more
access techniques.

examples

1-37

and

discussion

direct

IRECORD I
1.4.16

RECORD

The RECORD statement reserves areas of core where records are stored
during processing.
Block can be used interchangeably with RECORD.
The RECORD statement is of the form:
RECORD [name]
and is followed by statements of the form
fldnam ,xn
where name labels the record area;
fldnam names a
field within the
record and xn specifies whether the field contains alphanumeric or
decimal data, and the number of characters in the field.
Storage
is
allocated
in the order that the data appears.
The RECORD name can be
any number of alphanumeric characters in length
(only the first
6
characters
are
significant)
and the first character must be
alphabetic.
The name is optional unless the program is to transfer
(XMIT,READ or WRITE) the data.
If no name is specified, the data is
assigned to contiguous locations but is not available for
input or
output.
Therefore, RECORD statements with no names are used for
temporary storage.
Examples:
Legal

Illegal

RECORD
RECORD ALPHA
RECORD T1376
RECORD R

RECORD $
RECORD, ABC
RECORD 564
RECORD,

Each RECORD statement should be followed by one or more data field
definitions.
The data field name (fldnam) is optional and may be any
number of alphanumeric characters in length
(only the' first 6
characters are significant).
The first character must be alphabetic.
A comma may be used without a field
name if the program does not
reference the
individual
field.
This specification is used when
formatting an output line for
the printer,
for example,
so that
intercolumn spaces do not require a data name, only a comma followed
by the type and size.
Data field
type
(x)
is alphanumeric
(A)
or decimal
(D).
All
input/output data to the line printer, terminal or high-speed paper
tape reader should be in alphanumeric form
(A).
Decimal
(D)
is
specified for data to be used internally by the program.
However, the
alpha representation of a positive decimal number is the same as its
decimal
representation.
This is not true of negative numbers (refer
to Appendix A).
Data stored
in
alpha
fields
is
generally
left-justified and if necessary padded with blanks on the right.
Data
stored in decimal fields is always right-justified and padded with
zeros on the left if necessary.
The data field size (n) is a decimal
number 1-510 specifying the number of characters in the field.
Decimal
fields to be used
in arithmetic operations have a maximum
length of 15 digits.
If data size is omitted,
1 is assumed.
The

1-38

total size of all data fields in a record must not exceed 510, if the
record has a name (i.e. can be input or output), otherwise the total
size must not exceed 4094.
Examples:
Legal
RECORD MASTER
A, D3
Ml, D4
M2, A6
, A6
RECORD TRANS
CUST, A16
ADD, A30
KEY,
D
D03
Illegal
RECORD NUMBER
SETNO Dll
ORDERS, D632
RECORD NUMBER
B,
A-6

1-39

Inserting Initial Values
Initial values can be assigned to data fields as part of the field
statements following a RECORD statement.
The initial value may be set
as part of the data field definition or marked for insertion via the
terminal when the program is ready for execution.
The maximum size
(limited by the Monitor) of an
initial value
is approximately 110
characters.
If no
initial value
is assigned to a field it is
initially set to all blanks (if alphanumeric) or zeroes (if decimal) .
If entered as part of the program,
the alphanumeric or decimal
constant is placed after
the type and size specification. A comma
must separate the constant from the type and size specification.
The
initial value
to be entered must agree with the size and type
specified for the field.
The alphanumeric constant must be enclosed
in single quotation marks
(' ')
and may contain any printable
characters except \ , ' ,or
Tabs inside the single quotation marks
are treated as spaces,
1 space per tab. A decimal constant is a
string of decimal digits optionally preceded by a sign (+ or -)
and
optionally followed by a minus sign.
The signs (+ or -) and quotes do
not count as part of the size of the initial value.
The compiler does
not insert leading zeros or trailing spaces in initial values.
Examples:
Illeg al
,A4, 'TEXT'
M,D6,000400
PA,A2,'ll'
AA,D3,123
AC,A3,'l23'
A,D3,-146
B,D3,146C,D3,+123

,A4:
'NOTE'
, A4,
, 3Q'
,A3,345
,D4,
'WEED'
,D4,
12
,D4,12345
A3 '2'
A3,
'HELLO'
AIT
AD,A3,

If the initial value is to be entered when the program is executed,
put a comma and the character P after the type and size specification.

1-40

Example:
fldnam,A8,P
The character P is used in the data field specification to allow
user to enter data at run time.
Before execution, the message

the

ENTER fldnam
is displayed on the terminal.

The specification

QZBI35,A6,P
causes the message
ENTER QZBl35
to be printed at the terminal at run time.
characters

The

may

use

six

999999
The terminal entry is stored in the field named.
If the value entered
at run time in response to the ENTER message is shorter than the
field, the value is padded with blanks,
if longer, an error will
occur.
The P should only be used with named fields and is executed
once each time the program is run.
NOTE
When using the P option all
information entered
from the
terminals is treated as alphanumeric
e.g.,
LABEL,A6,P
in the data section causes the message
ENTER LABEL
If the user replies
ABCD..}
The sequence

LABEL will equal 'ABCD
LABEL,D7,P
ENTER LABEL
AB123..J

causes LABEL to equal AB12300.
A variable to be initialized with the P option can also
initial default value.

1-41

given

Example:
NAME ,A3, 'DOE' ,P
The default value must agree in type and size with that specified
for
the field.
The program uses the default value if CTRL/Z is typed in
answer to the run time ENTER message.
Examples:
Legal
ENT,A6,P
DUP,D8,P
ZIP,A5,' OOOOO',p

Illegal
,D2P

1-42

The character D is used in the data field specification to insert the
current date (as specified to the Monitor at start up) in a field to
be used in a report heading. The date is stored in the form mmddyy,
e.g., June 25, 1972 would be stored as 062572. The field must be
exactly six characters long or the date stored in it will be
incorrect.
This eliminates the necessity of specifying the P option
and entering the current date each time the program is run. The P and
D options cannot be used on the same field.
Example:
RECORD A
DATE,D6,D
Arrays
A number of values can be represented with a single field name by
placing a repetition count before the type and size specification.
All elements of the array must be the same size and type.
For
example:
TAG, 4D2
defines TAG to be 4 decimal fields, each 2 digits long. Data can be
initially entered in an array as a continuous string separated by
commas. For example,
TAG, 4D2,88,44,22,55
It is not necessary to initialize all fields of an array.
However,
those fields to be initialized must be the first field and any
continuous number of fields after the first field, for example:
TAG, 4D2,88,44
initializes the first two fields.
The
initialize the first array element only.

option

Examples:
Legal
C,4D3
D,5A2
, 2A4 , , BLUE'
,lD7
,16Al
GOAL,3D2,22,23,33
UP, 4A2 , , AB ' , 'CD' , , EF ' , 'GH '
INPUT,3D5,P

Illegal
,2D3, 34
,2D3,4545
, 2A2 , , AMERICA'
LABEL,103,Al
,3E4
,OD6
,-3A4
,6D-3
GOAL, 3D2,222333
UP,4A2,'ABCDEFGH'

1-43

can

used

The character S is used in the data field to assign the value of
variable (A2) equal to the value of the options used at run time.
Example:
VAR,A2,S
will set VAR to 'XY' if .RU pronam/XY is run.

1-44

RECORD Overlay Option
The X option in a RECORD statement allows multiple definition
fields within a RECORD.
Specify the overlay option as follows:

RECORD,X
or
RECORD name,X
The X option must be preceded by a comma.
There can be one or more
overlays defining the same record but there must be a preceding RECORD
statement without an X.
The overlay (X) record must be equal to or less than the size
record being overlayed.

the

Examples:
Legal

Illegal

RECORD ONE
RECORD X,X
RECORD, X

RECORD TIME,X,
RECORD X (X IS A LEGAL RECORD
NAME BUT IN THIS FORM DOES NOT
INDICATE AN OVERLAY)

The X overlay option is useful in reformatting
record area.
For example,

previously

defined

RECORD COLOR
F 1, A4, 'BLUE'
,AI
F2,A3,'RED'
has three alphanumeric fields with a
total of
Fields Fl and F2 can be referenced individually.

eight

characters.

RECORD RAINBO,X
ALL,A8
overlays the previously described record and permits treatment as one
field.
The variable ALL contains BLUE RED and setting ALL to blanks
in the PROC section sets Fl and F2 to blanks.
The X option also
allows access to a particular portion of the record by specifying
RECORD B,X
,A4
Q6,A4
which skips the first four characters and uses the last four.
contain the quantity' RED'.

Q6 will

The most common use for record overlaying is with print records.
Each
unique format for a report is kept in a separate record overlay.
For
example,

1-45

RECORD HEADI
HI, AID, 'HEADING #1'
RECORD HEAD2
H2, A23, 'THE SECOND HEADING LINE'
RECORD PRINT
PR132, A132
RECORD ,X
CUSTNO, AIO
iCUSTOMER NUMBER
,
A20
CUSTNM, A30
iCUSTOMER NAME
RECORD, X
,
A3S
CUSADD, A20
iCUSTOMER ADDRESS
,
AS
CUSZIP, DS

PR132=Hl
XMIT(6,PRINT)
PR132=H2
XMIT(6,PRINT)
PRINT=
CUSTNO=NO
CUSTNM=NAME
XMIT(6,PRINT)
PRINT=
CUSADD=ADDRI
XMIT(6,PRINT)
CUSADD=ADDR2
CUSZIP=ZIP
XMIT(6,PRINT)

1-46

Records in programs which are loaded directly by the .RUN command are
automatically cleared.
That is, all decimal fields are set to zero
and all alpha fields are filled with blanks, unless an initial value
is specified.
However, if the record is loaded as the result of a CHAIN statement,
the record will retain whatever contents it may have had on the
previous CHAIN unless the clear option (,C)
is specified in the
record.
See Appendix L and Section 1.4.3 for more detail on CHAINing records.

1-47

RETURN

1.4.17

RETURN

The RETURN statement is placed at the logical end ofa subroutine
has the form:

and

RETURN
The RETURN statement causes control to return to the next
after the last CALL (or call implied by a TRAP statement) .

statement

I t i s illegal to attempt a RETURN if no CALL has been
previously and the message RETURN WITHOUT CALL results.

executed

Example:

CCALL

LIST

LIST, i=B+C
CALL

SECOND,

1-48

I START I
1.4.18

START

This is an optional statement and can
program.
It is of the form:

inserted

anywhere

the

START [Ix] [i commen t]
where
is an optional listing switch to the compiler

temporarily suspends listing of source program

resumes listing of the source
errors on the line printer

program

and

resumes listing of the
errors on the terminal

program

and

source

The START option switch is active until disabled
by another option switch.
The STARTlx statement
may be inserted in a source program as often as
necessary.
If no option switch is specified a
compiler listing is produced on the line printer.
The switches can be overridden with an option
switch at compile time.
(Refer to Chapter 4.)
comment

is an optional method to store a line of text to
be output as a heading on the compiler listing.

Each time a START statement is encountered during compilation, a
eject occurs and a new page heading is printed.

page

Examples:

START
STARTiKR3766-CUST UPDATE
START/L
START/TiTEST PROG.
LISTING
START;
Illegal
ST
STARTK176-REPORT
STARTI
If IN is specified in the RUN COMP command, it overrides any IL's or
IT's in the source program. However, IL and IT will then determine on
what device errors are listed.

1-49

1.4.19

STOP

STOP terminates program execution and returns control to the
The form of the statement is:

Monitor.

STOP
and it may be inserted anywhere in the PROCedure section.
There may
be more than one STOP statement in a program. STOP does not close
files;
a FINI statement must be inserted before the STOP to close
files previously INITed.
The END statement has the same effect as STOP but an END can
used as the last statement in a program.

1-50

only

ITRACE NO TRACE I
1.4.20

TRACE/NO TRACE

These statements are helpful
anywhere in the PROCedure
statement is:

debugging tools and can be inserted
section of a program. The form of the

TRACE

NO TRACE
Execution of the TRACE statement enables program tracing.
Execution of the NO TRACE ~tatement discontinues
and is optional.
The initial mode is NO TRACE.

the

TRACE

function

The appearance of TRACE statements in a program does not cause any
TRACE output to be generated unless the /T option is specified when
running the program.
When program tracing is enabled and the /T switch is
statement executed causes the line

on,

each

DIBOL

AT LINE xxxx
where xxxx is the source line number of the statement to be output on
the line printer.
This tracing continues until execution of a NO
TRACE.
If the statement is a data manipulation statement,
the value
which is produced and stored by the statement is printed on the
following line.
For example:
AT LINE 0200
000006
NOTE
Indiscriminate placement of TRACE statements may
cause excessive output to be printed on the line
printer - this is often as bad as no output at
all.
To use the TRACE statement to best advantage
an attempt should be made to
"pin down"
the
problem to a certain part of the program flow and
then, us ing cond i tional
(IF)
sta temen ts,
tr ace
only that part of the program.

1-51

Example:
If a certain DIBOL program works correctly on records 1-53 of a file,
but produces incorrect results on the 54th record and an inspection of
the program fails to reveal the reason, the following
statements can
be added:
In the data definition section:

RECNO,D6

at the beginning of the procedure
section:

RECNO=Q

after reading an input record:

INCR RECNO
NO TRACE
IF (RECNO.EQ.54)TRACE

This will provide a complete trace of the operations
processing record 54, and only that record.

1-52

performed

when

1.4.21

TRAP

The TRAP statement allows a DIBOL program to spool output to the line
printer while executing the program. The format of the statement i c •
TRAP label
where
label

is the label of a line printer routine.

Whenever the line printer becomes free,
DIBOL statement execution
temporarily terminates and a call is made to the label specified in
the last TRAP statement executed.
(If no TRAP statement was executed,
no call is made.) When a RETURN is made from this call, normal program
execution resumes.
The TRAP statement normally precedes a FORMS or
XMIT statement.
The following example prints numbers 1-500 on the line
some other task is being performed:
N,

RECORD A
D3
PROC
TRAP SUB
FORMS(6,0)

;START LPT
iPERFORM TASK

LOOP,
SUB,

IF (N.LT.SOO) GOTO LOOP
STOP
N=N+l
IF (N.GT.SOO) RETURN
Xr1IT (6 ,A)
RETURN

1-53

printer

while

Typically, the TRAP statement would be used in a program as follows:
RECORD LPREC
A80
RECORD
LINCT, D2
LPFLAG, Dl
RECORD HDG
A60, , ••••• '

PROC n

INIT (l,IN,'INFILE')
INIT (6,LPT)
CALL PGTOP
iPERFORM TASK
DONE,

IF(LPFLAG.EQ.O) GO TO DONE

iLPFLAG=O IF TASK IS
iCOMPLETED BEFORE
iPRINTING

STOP
LPTRTN, IF(LPFLAG.NE.O) RETURN
LINCT=LINCT-l
IF(LINCT.EQ.O) GO TO PGTOP
XMIT (l,LPREC,LPEOF)
TRAP
LPTRTN
XMIT (6,LPREC)
RETURN

PGTOP,

LINCT=50
iINITIALIZE FOR NEXTPG
TRAP
PGTOPI
iWHERE TO GO NEXT
FORMS(6,O)
iSKIP TO NEW PAGE
RETURN iGO BACK
PGTOP1, TRAP
PGTOP2
iWHERE TO GO NEXT
XMIT(6,HDG)
RETURN
PGTOP2, TRAP
LPTRTN
FORMS(6,3)
RETURN
LPEOF, LPFLAG=l
FORMS(6,O)
RETURN

1-54

NOTES
1.

If the line printer buffer becomes empty during execution of
an INIT,
XMIT, READ, WRITE, DISPLAY or FINI statement while
I/O is in progress the trap will be delayed until execution
of the statement is complete.

If the line printer buffer becomes empty during execution of
an ACCEPT statement the ACCEPT statement will be interrupted
and resumed.
The user typing
in response to the ACCEPT
statement
will
not
be aware that the statement was
interrupted and the ACCEPTed data will be correct provided
the TRAP subroutine takes no longer than approximately 150
milliseconds to execute.
In the example above the execution
time
of
the
TRAP routine is much shorter than 150
milliseconds if INFILE is on disk.
If INFILE were on DECtape
the TRAP routine might take up to one-half second to execute
and the ACCEPT statement might lose some input.

Always construct a TRAP subroutine so that output to the line
printer
is immediately followed by a RETURN statement (as in
the example).

Due to a lack of line printer buffer space in the Monitor,
users with TD8/E DECtape and Centronix line printer are
limited to output lines of 76 characters in length when using
the TRAP statement. Outputting longer lines will result in
the program spending all of its time servicing line printer
TRAPs, thus cancelling the advantage of the TRAP statement.

The TRAP statement will not work on Analex line printers.

In the above example, the main DIBOL program will be slowed
down by approximately 5 to 10% by the TRAP processor
providing that INFILE is on disk.

1-55

1.4.22

WRITE

A direct access WRITE statement allows a specified data record
to
moved from an area in core to a specified file and has the form:
WRITE

(channel,record,rec#)

where
channel

is a decimal expression with the value of 1-15
specifying
channel which 1 inks the WRITE
a
statement to the related
INIT statement.
(The
INIT statement must specify UPDATE as the COS
device.)

record

is the record from which data is output.

rec#

is a decimal expression specifying
number of the record to be written.

the

sequence

Example:
WRITE(5,REX,88)
returns the 88th record
from the core area REX to the device
associated with the channel specified with the INIT (5, ... ) statement.
NOTE
Because of the harm which the READ and WRITE
statements could cause if improperly used, several
restrictions have been placed upon their use:
1.
2.
3.
4.
5.

The channel involved must refer
to a mass
storage device.
The file to be accessed with READ or WRITE
operations must contain records of a uniform
size.
Only one reel of a multi-reel
file
(the one
currently
mounted)
may be accessed by a
READ/WRITE statement.
The record which is specified must be exactly
the same size as the records of the file being
accessed.
Attempting to READ
or
WRITE
over
the
end-of-file mark results in an error message
and termination.

1-56

In general, READing or WRITing a record which
comes after
the end-of-file results in an
error
message;
however,
a
fortuitous
configuration of words on the unit being
accessed may cause the operation to succeed
and results in garbled data (on a READ) or the
loss of the output record
(on a WRITE)
but
does not crash the cbs system.
Unless the user provides a FINI statement
before terminating his DIBOL programs which
have UPDATE files, the data from the last few
WRITE statements may not be output properly.

Refer to Appendix L for more
access techniques.

examples

1-57

and

discussion

direct

1.4.23

XMIT

The XMIT statement transfers a data record and is of the form:
XMIT (channel, record[, eof label])
where
channel

is a decimal expression whose value is 1-15
specifying a channel number which associates the
XMIT statement with the related INIT statement.

record

is a name previously used in a RECORD statement
which identifies the area in core to which or from
which data is to be transmitted.
It may be a
simple
or
subscripted variable or a record
literal.
One must be very careful
record names.

when

using

subscripted

A single subscript, such as REC(3) should only be
used
if there are consecutive records, beginning
with REC and all of the same length as REC, i.e.,
RECORD REC
,D6

; REC (I)

,AIO
RECORD
,D6

; REC (2)

,AIO
RECORD
,D6

; REC (3)

,AIO
If a double subscript form is used,
i.e.,
REC
(n,m), then n must be less than m-l, n must be odd
and m must be even (or the last character
in the
record) .
This double subscript form refers to
characters n-2 through m-2
inclusive in record
REC.
(If n=l, it refers to characters I through
m-2.) Whenever an XMIT occurs referring to the
record,
two characters before character n in the
record are destroyed.
This is the COS 300 word
count
(see Appendix G).
The user need not be
concerned about this if n=l.
eof label

is the label of a statement to which the program
Used wi th
branches if an end-of-file is read.
input files only.
(Refer to Statement Labels at
the beginning of this chapter.) Label is optional
but if not specified, an error message is output
when an end of file occurs.

Examples:
XMIT (3,INV,EOF)

1-58

transfers a record from the input file associated with the statement
INIT
(3,IN, .•• ), to the RECORD area in core labeled INV.
If
end-of-file is reached, control branches to
PROCedure
section
statement labeled EOF.
If the length of the record being read is
greater than the defined size, an error message is output at run time.
If
the length is less than the defined size, the record is
left-justified and padded with spaces on the right.
XMIT (l,CUST,NEXT)
transfers a record from the
input file
associated
with
the
INIT(l,I, •.. } statement to the RECORD area CUST. At end of file, it
branches to the statement labeled NEXT.
The statement
XMIT (2,BUFF)
takes a record from RECORD area
associated with the INIT(2, •.. }
initialized for output, LPT etc.}.

BUFF and
statement

puts it in the file
(assuming channel 2 is

operator's

XMIT (8, "HI THERE')
would output the message HI THERE
channel 8 was INITed to the TTY.

the

terminal

XMIT (8,CUST(1,7) ,EOF}
accesses the first five characters of the record area CUST.
Other examples:
Legal
XMIT(N,B2}
XMIT(FUNCT(I}+2,FOO}
XMIT(7,ROD,LAB3}
XMIT(9,BLKNUM,L}
XMIT(3,Bl}
XMIT(M,B23A,I}
XMIT(l,HHAA,FFPP}
XMIT(8,"MOUNT INVOICE FORMS'}
XMIT(15,REC(1,33)}
XMIT(CHAN,REQ(1,INVOC+6}}
The XMIT statement may be used with direct access READ and WRITE
statements for operations on files.
See Appendix L for details.

1-59

PART II
SYSTEM AND UTILITY
PROGRAMS

CHAPTER 2
MONITOR

The Monitor is the master control program
contains all the system I/O handlers:
.LA30 VT05, LA36, VT50
.DECtape
.RF08
.RK08
.RK8E
.RX08
.Line Printer
.Card Reader
.High-speed Paper Tape Reader
.High-speed Paper Tape Punch

for

the

COS

system.

(LA36, VT50, COS-3l0 only)
(COS 300 only)

(COS-300 only)
(COS-300 only)
(COS-300 only)

The Monitor enables the user to edit, compile, save, and execute user
programs.
It also maintains a directory of all programs stored on the
system device, labels files when required, and opens and closes files
as needed.
During program execution the Monitor produces the messages which
instruct the user
to mount files.
It also provides the means for
BATCHing several programs for sequential execution.
Table 2-1 lists the special key commands that are available in on-line
operations.
The editing feature of the Monitor can be used to create SYSGEN tables
and source files on the system device for later use.

2.1

OPERATING PROCEDURES

The Monitor must be loaded via a bootstrap routine each time the
system is restarted.
(Refer to Appendix B.) Monitor signals that it
is loaded into memory by displaying the message
COS MONITOR 5.05 (or current version number)
DATE?

2-1

Type the DATE command, a space, the current date, and the CR key in
the form
.DA mm/dd/yy.
For example .DA 1/25/72. The date must be
entered before proceeding. This date is used during program execution
to date reports, files,
and new programs created.
The Monitor
responds to the date with a dot to indicate it is ready to accept any
of the commands shown in Section 2.2.

2-2

Table 2-1
Monitor Key Commands
Key

Function

CTRL/C

Returns control to the Monitor.
The Monitor
displays a dot and awaits a command.
If already
in the Monitor, CTRL/C has the same effect as a
CTRL/U.

CTRL/N
(COS 300
only)

Turns on imbedded numeric keypad at terminal and
can be used
in a BUILD or DIBOL program for
numeric input.
Type CTRL/N a second time or
return to the Monitor and the feature is turned
off. When the feature is turned on certain keys
are interpreted as numbers as indicated below.
Key

Value

K
L
U
I

----3

4
5

0
NOTE

Keys 7, 8, and 9 have values 7, 8, and
These ten keys form a numeric keypad.

CTRL/O

Suppresses terminal echo of typed output.
If echo
is
already
suppressed,
CTRL/O restores the
terminal echo. CTRL/O is also used to halt and
resume output from an LI command or the compiler.
The echo always resumes the next time the dot is
printed. All suppressed output is lost.

CTRL/Q
(COS 310
only)

Resumes output to the terminal from the
which it was halted by CTRL/S.

CTRL/S
Halts output to console terminal.
(COS 310 only) are lost.

point

No output data

CTRL/U

Deletes the current input line.

CTRL/V

Suppresses echo of Monitor message COS Monitor
5.03.
The Monitor displays a dot (.) when CTRL/V
is typed.
If message is suppressed,
type CTRL/V
to enable message.
CTRL/V has the same effect as
CTRL/U.

2-3

TaOLe ~-L \con~.}
Monitor Key Commands
Key

Function

CTRL/Z

Signals the end of input and
returns control
to
the Moni tor.
Halts output of line numbers from LN
command.

RUB OUT

Erases the last character
typed and echoes the
deleted character.
RUBOUT cannot be used in all
instances;
if the character to be deleted does
not echo when RUB OUT is typed, type CTRL/U and
retype the line.

2-4

2.2

MONITOR COMMANDS:

The commands
BATCH
DATE
DELETE
DIRECTORY
PLEASE
RUN
SAVE
apply to Monitor functions.
Only the first two characters of the
command need be typed.
(R is sufficient for the RUN command.) Any
additional characters up to the first blank are ignored. All commands
must be terminated by typing the CR key. With the exception of RUN,
the Monitor commands have no effect on system programs
(COMP,
SORT,
etc.). Refer to Section 2.3 for editing commands.

2-5

2.2.1

BATCH

with the BATCH command a string of Monitor commands -can be executed
sequentially.
As soon as the execution of one program ends, another
is automatically started. The form of the BATCH command is:
.BATCH pronam
where
pronam

is the name assigned to a list of Monitor commands
previously created via editing commands and saved
with a WRITE command.

For example:
.0090

RUN COMP,JOBI

.0100

SAVE JOBI

.0110

RUN SYSGEN, SYSTAB

.0120

RUN JOBI

.0130

RUN JOB2

.0140

RUN SORT, SRCL

.0150

RUN SYSGEN,OLDTAB

.0160

DE JOBI/B

All programs batched must be on the system device.
The Batch
can contain another BATCH command only as its last line.

program

When a Monitor command is read from the BATCH file, it is displayed on
the terminal and executed. Type CTRL/C to terminate a BATCH program.
The BATCH must then be restarted at the beginning.
All of the necessary programs and data files must be available;
if an
error occurs, BATCH Monitor mode terminates, control
returns to
interactive mode, and a dot is displayed on the terminal.
Correct the error in the program being executed and restart the BATCH
program at the beginning or individually type each command left to be
executed.
When the BATCH program is finished, control returns to the Monitor and
a dot is displayed on the terminal.

2-6

2.2.2

DATE

The DATE command stores a date to be assigned to any program created
or reports printed until a new DATE command is issued or the system is
restarted. The DATE command is of the form:
.DA

mm/dd/yy

mm
dd
yy

is a number 1 - 12
is a number 1 - 31
is a number 72 - 79

where

If the date entered does not conform to
error message is displayed.

the

above

restrictions,

This command must be entered whenever the Monitor
is restarted.
It
may also be used whenever it is necessary to change the system date.
Examples:
.DA 1/25/72
.DATE 12/5/73

2-7

2.2.3

DELETE

The DELETE command removes the named source, binary, or system program
from the specified device directory.
The DELETE command is of the form:
.DE pronam[,dev]/X
where
pronam

is the name of the file to
directory.

dev

is the 3-character designation for
device where the file is stored
DECtape
disk
diskette

the

from

the

physical

DTO-DT7
DKO-DK3
RXO-RX5

If no device is specified, the
assumed.

removed

system

device

Is a 1- or 2-character code indicating whether the
file to be deleted is a source (S) ,binary (B) or
system (SV)
file.
This code is necessary to
differentiate between three files with the same
name but of different types.
The code SV is used
rather
than V to make
it more difficult to
inadvertently delete a system file.

Examples:
Legal

Illegal

.DE

JOB1, DT1/B

.DE JOBl

.DE

PROGA,DK3/S

.DE CONVEX/V

.DE

INV/S

.DE CONVEX/SV
See Chapter 3 for a description of device designations.

The types of programs
system are as follows:
V S B F M -

system
source
binary
binary
binary

the

COS-300

program
file
(CaMP)
(FORMC)-(COS-300 only)
(MCOMP)-(COS-300 only)

2-8

2.2.4

DIRECTORY

The DIRECTORY command prints a list of programs stored on
device or the label of the file stored on a logical unit.
the command is:
.DI
{

[physical device]
Ilogical unit#

a physical
The form of

[/T~
~

where
physical device

-is one of DTO-DT7, RXO-RX5 or DKO-DK3 and
must be preceded by a comma or space.
If not
specified the system de~ice is assumed.

logical unit#

is the number
(0-15)
of the logical unit
assigned with SYSGEN. A logical unit # must
be preceded by a I.
Specifying a logical
unit causes a data file label to be printed.
If the unit is a physical device 'which has
more than one logical unit, the DI command
must be repeated for each logical unit to be
labeled.

is an optional switch which causes
directory to be output to the terminal.

the

A directory contains the current date, names of programs,
types of
programs, length in 256-word blocks (LN) and the date the program was
stored.
Be sure the line printer is on-line before issuing
the DI
command.
A file label contains the file name, sequence number (if a multi-reel
file),
the date the file was created, file length in segments, and
logical unit number where the file was stored when the label was
requested.
Segments are 16 256-word blocks long.
See Chapter 3 for a description of
unit numbers.

device

designations

and

logical

Examples:
.DI DKO
causes a directory similar to the following to be printed on the
pr in ter :
DIRECTORY
NAME

TYPE LN

l5-FEB-72
DATE

CaMP
v 14 19-JAN-72
MORE
S 10 l5-FEB-72
<0006 FREE BLOCKS)
TST2
S 07 l2-FEB-72
<0007 FREE BLOCKS)
TST4
S 07 l5-FEB-72
GLOP
S 10 l5-FEB-72
<0579 FREE BLOCKS)
The command
.!.oDI/3
outputs a file label similar to one below.

****************************
*
*
NAME
DATE
SEQ.
*
*
*
*
DEP
#01 l8-NOV-75 *
*
*
*
LENGTH: 0046 UNIT: 3 *
*
*
*
****************************

2-10

line

2.2.5

PLEASE

The PLEASE command outputs a message to the terminal
program execution.

during

BATCH

The form of the PLEASE is:
PLEASE text string
The message is displayed exactly as entered and the terminal alarm is
sounded.
After
taking the action requested in the PLEASE message,
type anyone key to continue" the BATCH program. To make a two-line
PLEASE command,
the first line can be terminated with an AND and the
second line begun with another PLEASE. This lets the operator
know
more message is to follow.
For example:
0020
0030
0040
0050
0060

RUN JOBI
PLEASE PUT INVOICES IN LINE PRINTER AND
PLEASE TYPE 3 TO THE NEXT MOUNT MESSAGE.
RUN JOB3
PLEASE PUT REGULAR PAPER IN LINE PRINTER

When this BATCH program is executed, JOBI will be run,
the first
PLEASE message will be displayed and the terminal alarm sounded. The
system waits for a key to be typed in reply to the PLEASE message then
displays the next PLEASE message. When a key is typed in reply to the
message, JOB3 is executed and the last PLEASE message displayed.
Control returns to the Monitor when a key is typed in reply to the
last PLEASE message.
If a PLEASE command is given when in a non-BATCH mode, following
CR, the bell rings and the system waits for a key to be typed.

2-11

the

2.2.6

RUN

The RUN command loads and executes the named system or binary program
using
the named file.
This is the command which provides access to
all other system programs such as:
SYSGEN

To change or check system configuration

BUILD

To build data files

SORT

To sort data files

UPDATE

To update data files

PIP

To move information between physical devices.

COMP

To compile a user source program into
program.

the

binary

The RUN command has the for~:
.RU [pronam] [+chainl+ ..• +chainn] [,filnaml, •.. ,filnamn] [/xx]
where
pronam

is the name of the system program to be run or
pronam+chainl...
are binary files which are part
of a program and constitute one large program
broken up into several chains.
For example:
.RUN PROG+CHAINl+CHAIN2+CHAIN3
would execute program PROG.
PROG would
then
determine
whether programs CHAINl,
CHAIN2 or
CHAIN3 would next be run.
See section 1.4.4 for a
description
of the CHAIN statement.
If the
program name is omitted, Monitor
loads
and
executes the last compiled DIBOL program from the
compiler's scratch area.

filnaml, ... ,filnamn
are source files which must be on the system
device.
If one of the system programs is executed
via the RUN command and no source files are
specified as input,
the file
in the editing
input
(system programs
scratch area is used as
only) .
The maximum number of binary and source files per
program is eight. Multiple files are concatenated
and passed to system programs as one large file.
/xx

is one or a combination of the option switches
including /T which enables the TRACE feature and
/D which enables the DDT mode.

2-12

Other options are described with the programs to which they apply.
Refer
to the appropriate chapter
for more detailed information on
these programs.
If the program specified is not found, an error message is produced.
The RUN command is the only Monitor command which can
to one character, R.

abbreviated

Examples:
Command

Explanation

.RU

executes previously compiled DIBOL program from
the Compiler's scratch area .

. RU JOBI

runs program called JOBI

.RUN CaMP, CHECK

this command compiles the source program CHECK,
prints a compiler listing and temporarily saves
the binary on the system device .

. R BUILD, FD, IO,FILEl,FILE2
this command builds a data file
in the format
specified
in the Field Descriptor (FD) and I/O
section (10) of the BUILD control program with the
data in FILEI and FILE2.
MOUNT Messages
The Monitor outputs MOUNT messages to the terminal whenever an
or output unit must be specified.
These messages have the form:

input

MOUNT xxxxxx inn FOR INPUT:
MOUNT xxxxxx inn FOR OUTPUT:
where
xxxxxx

is the name of the data file
program currently executing.

inn

is a sequence number from 1 to 63.

desired

the

Answer the MOUNT message with the appropriate logical COS unit (1-15).
If an error is made in the reply, type CTRL/U and the correct reply.
The MOUNT message is displayed (and the sequence number
incremented)
whenever
the program reaches the end of the device and more
information remains to be read or written.
When default units specified in control programs are not available,
question mark precedes the MOUNT message, i.e.,
?MOUNT xxxxxx #01 FOR INPUT:
The message
REPLACE xxx xxx inn ?

2-13

is displayed when a file is already stored on the logical unit
specified to a MOUNT message. Answer REPLACE with a Y and the CR key
to replace the old file or any other key to keep the old file.
File
labels beginning with any character other than A-Z, [, ~, or ] are
considered to be temporary files, and no REPLACE message is displayed
when such files are about to be written over. The following error
messages can occur:
IN USE
ILLEGAL UNIT
NOTE
Occasionally, if output is to a scratch unit, a
garbled filnam or sequence number in the "REPLACE"
message may be typed. This is because the area on
the unit where the label should be is simply
random characters.
Just answer "YES" to the
REPLACE message as usual.

2-14

2.2.7

SAVE

This command stores the binary program from the compiler scratch
on the named device. The form of the SAVE command is:

area

.SAVE pronam[,dev] [/Y]
where
pronam

is the name to
binary output.

dev

is any mass storage device DTO-DT7, RXO-RX5 or
DKO-DK3 which has a directory. If no device is
given, the system device is assumed.

message and YES
is used to bypass the REPLACE?
would be
response
when
a
duplicate
name
encountered. Normally used in a BATCH mode to
bypass operator response.

assigned

the

compiler's

The SAVEd program must be moved to the system device before it can be
executed. Depending on the compiler used (CaMP, FORMC, or MCOMP), the
saved binary will be of type B, F, or M, respectively.

2-15

2.3

EDITING COMMANDS

insertions,
The COS Monitor contains a line editor;
that is, all
changes, and deletions are done with line numbers. All text input to
the Monitor is assigned a sequence of line numbers.
For example:
0100
0110
0120
0130
0140
0150
0160
0170
0180
0190
0200

START
RECORD A
AI,
A64
PROC 2
INIT (2, IN, , MINT' )
LOOP,
XMIT (2, A, EOF)
XMIT (8, A)
GO TO LOOP
EOF,
FINI(2)
STOP
END

The commands:
ERASE
FETCH
LIST
Line Number
Number Commands
RESEQUENCE
WRITE
apply to the editing functions.
These commands can be entered at any
time in response to the dot displayed by the Monitor. Only the first
two characters of the command need be typed and all commands must be
terminated by typing the CR key.

2-16

2.3.1

ERASE

This command erases text from the edit buffer
where the Monitor stores text being edited).
is:

(that portion of core
The form of the command

. ER [nl] [,n2]
where
nl

is the line number of the line to be erased or the
first line number of two which delimit the lines
to be erased.
If omitted, erasing starts at the
beginning of the text buffer.

is the second of two line numbers indicating where
erasing ends.
If n2 omitted, but the comma is
included, erasing continues to the end of the edit
buffer.

If no line numbers are entered, the ERASE command clears the entire
edit buffer.
Use this command to erase the edit buffer before
entering material to be edited.
Examples:
.ER

clears the entire edit buffer.

.ER 5

clears line 5.

.ER ,5

clears from the start of the buffer to 1 ine 5.

.ER 5,10

clears from line 5 to 1 ine 10, inclusive.

2-17

2.3.2

FETCH

The FETCH command loads the named source file into
specified device. The form of the FETCH command is:

core

from

the

.FE pronam[,dev]
where
pronam

is a previously created program
brought into core.

dev

is the 3-character designation of the physical
device, DTO-DT7, RXO-RX5 or DKO-DK3 where the
program is stored.
dev may be separated from
pronam by a space or comma.
If device
assumed.

omitted,

Monitor searches for the named program.
Monitor displays the message

the

which

system

device

If the program is not

found,

FILE NOT FOUND
Retype the command with the correct program name or device.
If another program is run while this program is in the active work
area, Monitor temporarily stores this file in the editing scratch area
on systems device, then returns it to core when program execution is
complete.
Examples:
.FE RICH

Moves program RICH from the system device to the
edit buffer •

. FE PAYROL,DT2 Moves program PAYROL from DECtape unit 2 to the
edit buffer.
.RUN XYZ

Runs a program called XYZ. On completion of this
program, the PAYROL source file is in the edit
scratch area.

The FETCH command erases the edit buffer before inserting the
program.

2-18

FETCHed

2.3.3

LIST

The LIST command outputs the specified lines or the entire edit buffer
to the high-speed punch, line printer, or terminal.
The LIST command has the form:
..!,.LI [nl] [,n2] [Ix]
where
nl

is the line number of the line to be listed or
the
first line number of two which delimit the lines to be
listed.
If omitted, listing starts at the beginning of
the text buffer.

is the second of two line numbers indicating where
listing ends.
If omitted, but the comma is included,
listing continues to the end of the edit buffer.

is the one
device.

IL
IP

letter

code

which

indicates

the

output

line printer
high speed punch (available on COS 300 only)
(line numbers are omitted) .

If no device is indicated, the list is
terminal.

output

the

If no line numbers are specified, the entire edit buffer is output to
the specified device. Type CTRL/O to stop output from an LI command.
Examples:
.LI

List entire edit buffer on line printer.

.LI 5

List line 5 on the terminal.

.LI ,5

List from start of buffer to line 5 on the terminal.

.LI 5,/P

List from line 5 to
high-speed punch •

• LI 5,10

List line 5-10, inclusive, on the terminal.

the

2-19

end

the

buffer

the

2.3.4

Line Number

The Line Number command automatically outputs line numbers
at
specified
increments so new lines can be entered without typing each
line number.
The Line Number command has the form:
. LN [ n] [, inc] [/ x]
where
n

is the starting line number.
If no
number is specified, 100 is assumed.

starting

line

If the comma after the starting number is omitted,
the
starting number and
increment number are the same,
i.e., LN 100 causes a starting number and an increment
.of 100.
If the command is LN 100, the start line number is 100
and the increment remains unchanged unless the Monitor
is read back into core;
at which time the
increment
returns to 10.
inc

is the increment between line numbers.
is specified, 10 is assumed.

is the input device (available with COS 300 only).
/R
/C
/T

If no increment

high-speed reader (All rubouts are ignored so that
paper
tapes with TABS followed by RUBOUTs will be
read correctly.)
card reader
low-speed paper tape reader.

If no device is specified,
the terminal
is assumed.
Line numbers are specified by the LN command and cannot
be part of the text read from the input device.
If Line Number mode is terminated and some editing has been done, type
the line number command, LN, with no arguments to display the next
number in sequence. When inputting from the card reader, all 80
columns are stored (40 columns with mark-sense cards).
The LN command does not clear the edit buffer.
Line Numbers 0 to 4095
are available. Under the default conditions (start at 100, increment
by 10), the program can be approximately 400 lines long.
The maximum number of characters on a line, including the line number
and space, is 120.
In automatic line number mode the line number and
space are counted as two characters. When the terminal line is full,
Monitor executes carriage return/line feed and since this is a
continued line does not display the next line number.
If the 120 characters limit is exceeded, the Monitor gives
message LINE TOO LONG and the line is lost.

2-20

the

error

If a CR is the first character typed, no line
is entered and
number increments.
To obtain a blank line, type space and CR.

line

Tabs can be used to increase the readability of a
program.
The
tab
key on most terminals is set to produce up to 8 spaces.
The first tab
goes ·to column 13 because the line number and
space
take
the
first
five positions.
Type CTRL/Z to indicate the end of input and halt the automatic line
numbers.
When input is coming into the Editor from the card reader, a
beep sounds to indicate no more cards.
Load more cards
into
the
reader
and type any character to continue the input or type CTRL/Z to
indicate the end of input.
Examples:
Requests line numbers starting at 100 with
increments
of 10.
The terminal is assumed to be an input device.

.LN

The terminal display would be
.0100
Type the first line of the program and
the CR key.
Editor displays
the next line number each time the CR
key is typed.
~0100

START

.0110
.LN 10,5/R Requests line numbers starting at 10 with increments of
5 to be assigned to input from the high-speed reader.
. LN ,100

Request line number
starting at 100
(default)
increment of 100.
Input is from the terminal.

.LN 10/C

Requests line number starting at 10 with
increment
10 (default).
Input is from card reader.

with
of

If an error is made when using automatic line numbers, type RUB OUT or
CTRL/U.
RUB OUT erases the last character typed and echoes the erased
character.
CTRL/U erases the entire line.
Monitor
redisplays
the
line number.
CTRL/C acts the same as CTRL/U.
If the edit buffer is full, the error message TEXT AREA FULL appears,
and
the last line entered is lost.
The edit buffer must be separated
into two or more source filees.
This is done by:
1.

WRiting the edit buffer as file B.

ERasing the last half of the edit buffer.

WRiting the edit buffer as file A.

FEtching file B.

ERasing the first hal f

WRiting the edit buffer as file B.

0 f

the edit buffer.

2-21

NUMBER COMMANDS

2.3.5

Number Commands

Any line beginning with a number inserts the line number and
the edit buffer.

text

The form is:
nnnn
or
nnnn text
where
nnnn

is a line number.

text

is the data to be inserted in the buffer.
The data
must be separated from the line number by a space
or tab.
(If tab is used, it becomes the first
character of text.)

If text is already at that line number, the new text replaces it.
Files are stored in increasing line number order and a new line number
is inserted in numerical order. If a CR is typed right after the line
number, data at that line number is cleared from the buffer.
The input line is limited to 120 characters including the line
which counts up to four characters.
Examples:
.40
erases text and number at line 40
745 RECORD inserts RECORD at line 45
Text before editing:
0035
0040

PROC
INIT{I,V,IN)

0047

XMIT(6,B)

0060

END

Editing commands:
.35
:40

~47

PROC 1
INIT{l,IN,'LABEL' ,2)

Text after editing
0035
0040

PROC 1
INIT{l,IN,'LABEL' ,2)

0060

END

2-22

number

2.3.6

RESEQUENCE

The RESEQUENCE command renumbers
additions and deletions.

the

program

lines

adjust

for

starting

line

The form of the command is:
• RE [ n] [, inc]
where
n

is the starting line number.
If no
number is specified, 100 is assumed.

If the comma after the starting number is omitted, the
starting number and increment number are the same,
i.e., RE 100
causes a starting number and
increment
of 100.
If the command is RE 100, the starting line
number is 100 and the increment remains unchanged
unless the Monitor
is read back into core; at which
time the increment returns to 10.
inc

is the increment between line numbers.
is specified, 10 is assumed.

If no increment

If line number exceeds 4095 on a resequence, the error message, LINE #
TOO LARGE,
results.
Enter another RESEQUENCE command with smaller
increments.
If this is not done, the text will be only partially
RESEQUENCEd and duplicate line numbers may result.
Examples:
.RE

Resequences line numbers of program in the edit
buffer using 100 as starting line number and 10 as
increment •

. RE 10,5

Resequences line numbers of program in edit buffer
using 10 as starting line number and 5 as the
increment •

. RE ,100

Resequences line numbers of program in edit buffer
using 100 or the last specified line number as the
starting line and 100 as the increment .

. RE 50,

Resequences line numbers of program in edit buffer
using 50 as the starting number and the increment
remains unchanged.

2-23

2.3.7

WRITE

The WRIrE command stores a new or modified source program on the
specified device so it can later be compiled or FEtched for editing.
The form of the command is:
• WR pro n am [ ,d e v] [/Y]
where
pronam

is the six character name assigned to the
. to be stored.

dev

is the three character designation of the physical
device
(DTO-DT7,
RXO-RX5 or DKO-DK3) where the
program is .to be stored.
I~
no device
is
specified, the system device is assumed.

is used to bypass the REPLACE?
message and YES
name
would be
response
when
a
duplicate
encountered. Normally used in a BATCH mode to
bypass operator response.

program

If the name specified is a duplicate name, Monitor asks
REPLACE?
Type Y or YES and the CR key to replace the old file with the new
file.
Type N or any other· character and the CR key to leave the old
file and return to the Monitor.

2-24

2.4

EDITING EXAMPLE

.FE

ACC08S

.LI
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110
0120
0130
0140
0150
0160
0170
0200
0210
0220
0230
0240
0250
0260
0270
0280
0290

0500

iLIST PROGRAM TO BE EDITED
START
iACC08SiTHIS PROGRAM UPDATES A MASTER FILE OF STOCK RECORDS, WITH A
iTRANSACTION FILE OF RECEIVALS AND WITHDRAWALS.
iA NEW MASTER FILE AND A REPORT ARE PRODUCED.

RECORD MASTI
iMASTER INPUT RECORD
D9
DI0
iUNIT COST
DI0
D6
iDATE OF LAST TRANSACTION (MMDDYY)
BLOCK,X
MIWHOL, A35

MIPART,
MICOST,
MIQTY,
MIDATE,

RECORD
D9
DI0
DI0
D6

MASTO

MOPART,
MOCOST,
MOQTY,
MODATE,

RECORD
Al
D9
D7
D6

TRAN

TYPE,
TPART,
TQTY,
TDATE,

iMASTER OUTPUT RECORD

iTRANSACTION FILE
iTYPE (R=RECEIVAL, W=WITHDRAWAL)
iPART NUMBER
iQUANTITY
iDATE (MMDDYY)

END
jEditing

.150
.230
.290
.0090
.0110
.0228
.0229

MIPART, D9
iPART NUMBER
MIQTY,
DI0
iQUANTITY ON HAND
BLOCK,X iREDEFINITION
MOWHOL, A35

2-25

.LI

iLIST EDITED PROGRAM

0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110
0120
0130
0140
0160
0170
0200
0210
0220
0228
0229
0240
0250
0260
0270
0280

START
iACC08S
iTHIS PROGRAM UPDATES A MASTER FILE OF STOCK RECORDS, WITH A
iTRANSACTION FILE OF RECEIVALS AND WITHDRAWALS.
:A NEW MASTER FILE AND A REPORT ARE PRODUCED.

MIWHOL,

0500

END

MIPART,
MICOST,
MIQTY,
MIDATE,

MOPART,
MOCOST,
MOQTY,
MODATE,
MOWHOL,
TYPE,
TPART,
TQTY,
TDATE,

RECORD MASTI
iMASTER INPUT RECORD
09
iPART NUMBER
010
iUNIT COST
010
iQUANTITY ON HAND
;DATE OF LAST TRANSACTION (MMDDYY)
06
BLOCK,X
A35
;MASTER OUTPUT RECORD
RECORD MASTO
D9
010
DIO
D6
BLOCK,X
iREDEFINITION
A35
RECORD TRAN
iTRANSACTION FILE
Al
iTYPE (R=RECEIVAL, W=WITHDRAWAL)
D9
iPART NUMBER
07
iQUANTITY
06
iDATE (MMDDYY)

2-26

2.5

BATCH EDITING

As the following example illustrates, editing can be done via a
program .

BATCH

• ERASE
.LN 5,5/R
• LIST

iAUTOMATIC LINE NUMBERS START AT
INPUT FROM
iPAPER TAPE READER.
iLIST THE FILE READ IN .

is AND INCREMENT BY 5.

0005 FETCH TEMP
0010 LIST
0015 ERASE 4
0020 ERASE 30,34
0025 35 PROC 1
0030 40 INIT(1,IN,'LABEL',2)
0035 47 FORMS (6,0)
0040 RE 5,5
0045 LIST
0050 WRITE TEMPA
0055 DELETE FILE/S

iBATCH Program

• WRITE FILE

iWRITE BATCH PROGRAM CALLED "FILE" •

.BATCH FILE

iRUN BATCH PROGRAM "FILE"

FETCH TEMP

iFIRST COMMAND OF BATCH PROGRAM LOADS

iTEMP FOR EDITING
LIST
0000 RECORD A
0004
,01
,AS
0005
0010
,A6
0015
,D6
0020
,010
,A25
0025
0030
,02
0031 RECORD B
,02
0032
,AI
0033
0034
,02,
70
0035 PROC
0040
INIT(l,V,IN)
0045
XMIT (1 ,A, EOF)
0046
XMIT (6,A)
0047
XMIT (6, B)
0050 EOF,
FINI (1)
STOP
0055
0060
END

SECOND COMMAND LISTS TEMP

iTEMP PROGRAM

2-27

ERASE 4
ERASE 30,34
35 PROC 1
40INIT(I,IN,'LABEL',2)
47 FORMS (6,0)
RE 5,5
LIST
0005 RECORD A
0010
,AS
0015
,A6
0020
,D6
0025
,DI0
0030
,A25
0035 PROC 1
0040 INIT(I,IN,'LABEL',2)
0045
XMIT (I,A,EOF)
0050
XlvlIT (6,A)
o0 5 5 FORM S ( 6 , 0)
0060 EOF,
FINI (1)
0065
STOP
0070
END
WRITE TEMPA
DELETE FILE/S

iTHIRD COMMAND ERASES LINE 4.
iFOURTH COMMAND ERASES LINES 30-34.
iFIFTH COMMAND INSERTS NEW TEXT
iAT"LINE 35.
iSIXTH COMMAND INSERTS NEW TEXT
iAT LINE 40.
iSEVENTH COMMAND INSERTS NEW TEXT
iAT LINE 47.
iEIGHTH COMMAND RESEQUENCES LINE
iNUMBERS STARTING AT 5 AND
iINCREMENTING BY 5.
iNINTH COMMAND LISTS EDITED TEMP.

iSTORE TEMPA ON SYSTEM DEVICE
iDELETE SOURCE FILE OF BATCH
iPROGRAM

2-28

2-29

2.6

ERROR MESSAGES

Monitor Errors
Message

Explanation

BAD COMPILATION

User tried to SAVE
that had errors.

BAD DIRECTORY

compiled

binary

Directory does not
example, requesting
data tape).

look right
a directory

(for
of a

BAD LABEL

Tape has no
incorrect.

form

DT n?

The expected tape is not available on
DECtape drive.

label,

its

For example,
if following
a
MOUNT
message the operator has specified that
a file should be written on unit 3, and
unit 3 is not selected, or is WRITE
LOCKed, this message is output.
CTRL commands are not operative at this
point,
i.e., CTRL/C does not work.
To
recover, set switch to ENABLE the unit
or set the dial selector to the proper
number. Then type any key to continue.
On LINCtapes there is no message.
ENTER "DATE MM/DD/YY"

Typed an unrecognizable date.

ERROR IN COMMAND

Miscellaneous.

FILE NOT FOUND

After, for example, FETCH FILEX.

LINE TOO LONG

Greater than 120 characters.

LINE # TOO LARGE

Greater than 4095.

REPLACE?

Duplicate
replace.

TEXT AREA FULL

Greater than 8,150 characters.

?ILLEGAL UNIT

The unit specified is not
DKO-DK3,
RXO-RX5 or DTO-DT7;
for example,

file

names.

Type

WRITE FILE,XY3
WRITE FILE, RK9
!*!MEMORY FAIL!*!

A hardware error has occurred.
the program.

2-30

Restart

Message

Explanation

?NO FILE TO SAVE

Nothing in the edit work area when WRITE
command is issued.

*NO INIT DONE

Program attempted to read or write on a
device that was not opened by the system
program.

Run-Time Error Message
BAD DIGIT

A character other than +, -,
space, or
the digits 0~9 was encountered
in an
alpha to decimal conversion.

*BAD PROGRAM

Attempting to run a binary program which
contains a compilation error.
Check
compilation listing for error flags.
Correct flagged errors, and recompile.

*DIBOL FILE NUMBER IN USE

In INIT, the channel number
associated with a device.

already

*DIBOL FILE NUMBER NOT INITED
An attempt was made to XMIT,
READ, or
WRITE with a channel number that was not
associated with a device.
END OF FILE

I/O ERROR ON XX, RETRY?
(COS 300 only)

I/O ERROR ON XXN, RETRY?
(COS 310 only)

The last record of an input file has
been read and the end of file mark
encountered, but no EOF
label
was
specified in the XMIT statement.
System failed in three attempts to read
from or write to a disk.
XX=DT or OK.
Type Y if retry is desired;
any other
character
if no retry and return to
Monitor is desired.
In either case,
type only one character.
Where XX is the device type and N is the
unit number. Type N for no retry. Any
other input causes the device handler to
re tr y.

*ILLEGAL DEVICE

Attempt to WRITE on a file that was not
INITialized for
UPDATE or attempt to
READ from
a
file
that
was
not
INITialized for INPUT or UPDATE.

*ILLEGAL SUBSTRING

A DIBOL PROCedure section statement has
attempted to access a data field, Fl
(m,n)-, but m=O or m>n.

ILLEGAL RECORD #

Record number is 0, past the end of the
physical unit or the length specified in
the record header word does not match
the length of the XMIT block (records in
data file are all not the same length) .

2-31

Message

Meaning

LINE TOO LONG

An input line
(record)
block it was read into.

*NO BUFFERS LEFT

Not enough core available
for
I/O
buffers. An I/O buffer of some multiple
of 512 characters is set up for
each
active
mass
storage file.
Another
possibility:
too
few
files
were
specified in the PROe statement.

NO FILE

No file specified in RUN statement
satisfy INIT (SYS) command.

NUMBER TOO LONG

A decimal field longer
than
was used in a calculation.

*PROGRAM TOO BIG

Binary program does not fit in available
core.
Reduce program size.

*PUSHDOWN OVERFLOW

Either (1) a statement is too complex or
(2)
subroutines are nested to a depth
greater than 50, or a combination of the
two.

*RETURN WITHOUT CALL

The program tried to execute a RETURN,
but there was no place to go:
there was
no corresponding CALL statement.

*SUBSCRIPT TOO BIG

Program
attempted
to
destroy
the
run-time system or
itself by using a
subscript larger ,than that defined
in
the Data Definition section.
Note that
the run-time system does not detect all
illegal
subscripts;
only those which
would cause the user's program or
the
system to be destroyed.

ZERO DIVISOR

The program attempted to divide by zero.

overflowed

digits

NOTES
1.

The messages marked with
an
asterisk
(*)
cannot be checked
with an ERROR statement.

AT LINE nnnn is displayed under
all
run-time error messages.
nnnn is a DIBOL source program
line number
in which the error
occurred.
If
COMP/O
were
specified for this program, nnnn
is meaningless.

2-32

the

CHAPTER 3
SYSGEN
SYSGEN is a conversational program used to reconfigure the COS-300 and
COS 310 software systems to handle available hardware, to change the
logical unit assignments, and to print a table of those assignments
for reference.

3.1

COS-300 OPERATING PROCEDURES

To load SYSGEN type:
.RU SYSGEN[, filnam]
or
.RU SYSGENlx
where

3.2

filnam

is optional and is a file previously created and stored
with the Editor which contains the table of logical unit
assignments to be made.

is one of three options used with SYSGEN.

change the hardware handlers in the system.

take new logical unit assignments from terminal.

print table of current logical unit assignments.

COS-300 SYSTEM SOFTWARE CONFIGURATION

As shipped, the COS-300 software system is set
following hardware configuration on the PDP-8:
TC08 with 8 DECtapes (or TD8E for the 8/E)
RK8E, RK08 or RF08 disks
Terminal
LP08 80-column Line Printer
Card Reader
High-speed Paper Tape Read and Punch

3-1

handle

the

or
LINCtape (PDP-12)
RK8E, RK08 or RF08 disks
LP08 80-column Line Printer
Terminal
Card Reader
High-speed Paper Tape Reader and Punch
on the PDP-12.
NOTE
The SYSGEN program cannot be used when
the
COS - 300 Foreground Data Entry
Program is running.
NOTE
No change is necessary if your
COS 300
hardware configuration matches the one
described here.
If, however, the system has either a 132 column line printer or an
Analex lin~ printer or the system device is to be changed, type the
command
.RU SYSGEN/C
and the CR key.
(Refer to Figure 3-1 for a flow chart of the question
and answer sequence.)

SYSTEM IS
RECONFIGURED
AND CONTROL
RETURNS TO
MONITOR

Figure 3-1
SYSGEN Flow Chart

3-2

For the COS-300 System, SYSGEN displays the following question on
terminal:

the

DO YOU WANT TO CHANGE THE SYSTEM DEVICE?
Type NO and the CR key to keep the current
skips to the question on the line printer.

system

SYSGEN

device.

Type YES and the CR key to change the system device from DECtape to
Disk or Disk to DECtape. This is the only way to move the Monitor and
DIBOL run time system (RSYS).
If the system device is changed,
the
system must be restarted with the bootstrap appropriate to that device
(see Appendix B or Chapter 9, BOOT). SYSGEN then asks:
WHAT IS THE NEW SYSTEM DEVICE?
Type the three character designation of the new system device, DKO-DK3
or DTO-DT7.
NOTE
Since only an RK08, or RK8E, or RF08
disk may be resident on the sytem, disks
are commonly addressed as DKO-DK3.
An
RF08 disk must always be addressed as
DKO.

DO YOU WANT TO TRANSFER YOUR FILES?
Answer YES and the CR key to move the rest of the system programs
(PIP, COMP, etc.)
to the new system device.
Transfer of the system
program destroys anything stored on the new system device.
Answering
NO and the CR key will zero the directory of the new device.
DO YOU HAVE AN ANALEX PRINTER?
NOTE
The Analex printer must be configured
into the system with SYSGEN before it is
turned on or it will not
function
properly.
Answer YES and the CR key and SYSGEN skips to the next question.
IS EVERYTHING CORRECT?
Answer NO and the CR key and SYSGEN assumes an LP08
and the question

LS8E

printer

132 COLUMN PRINTER?
is displayed. Answer YES and the CR key if there is a 132 column line
printer.
Answer NO and the CR key and SYSGEN assumes an 80 column
line printer. The next question is:
IS EVERYTHING CORRECT?

3-3

Check all answers made to the above questions.
Type YES and
the CR
key if the
answers are correct.
Type NO and CR key to correct any
errors.
SYSGEN repeats the questions starting at the beginning.
SYSGEN/C does not repeat the logical unit assignment to
reflect the
new area occupied by the system on the disk.
Run SYSGEN/T to reassign
logical units if the system is moved to a disk or the system may be
destroyed by a data file.

3.3

COS-310 OPERATING PROCEDURES

To load SYSGEN type:
.RU SYSGEN[, filnam] [/x[y]

where
filnam

is optional and is a file previously created and
stored
with the Editor which contains the table of logical unit
assignments to be made.

is one of the following four options used with SYSGEN.
x=

/xy

takes changed
information
(There is no print out) .

from

the

edit

change the hardware handlers in the system device.

take new logical
unit assignments
(There is no printout).

print table of current logical unit assignments.

from

buffer.

terminal.

is an additional option from the system.

where

x = B or T as described above
y

L to give printout logical units table
immediately after
first option
table) .

(which

changes

If no file or options are specified SYSGEN takes
the
information
stored
in
the edit buffer
as the new logical unit assignments and
prints them on the line printer.
explanation

Refer
to
the appropriate section for
further
reconfiguration and logical unit assignments.

3.4

COS-310 SYSTEM SOFTWARE CONFIGURATION

As shipped the COS~310 software system
is
following hardware configuration on a PDP-8.

3-4

set

handle

the

RX8E
VT50
LA36

Dual diskette
terminal
DECwriter
NOTE
SYSGEN/C should be run immediately upon
installation of the COS-310 Software or
upon transfer of system. Failure to do
so may result in inconsistencies within
the system.

To run COS-310 SYSGEN type:
..:.RU SYSGEN/C
(Refer to Figure 3-2 Flow Chart of question and answer sequence.)

HOW MANY LINES
PER PAGE?

SYSTEM IS
RECONFIGURED AND
CONTROL RETURNS TO
THE MONITOR

Figure 3-2
Question and Answer Sequence

3-5

For the COS-3l0 system, SYSGEN displays the following questions on the
terminal:
DO YOU WANT TO CHANGE THE SYSTEM DEVICE?
Type NO and the CR key to keep the current system
skips to the questions concerning the line printer.

device.

SYSGEN

Type YES and the CR key to either change the system device from
DECtape to Disk (i.e., diskette) or change the system device from Disk
to DECtape. This is the only way to move the Monitor and DIBOL run
time system (RSYS).
If the system device is changed, the system must
be restarted with the bootstrap appropriate to that device
(see
Appendix B or Chapter 9, BOOT). SYSGEN then asks:
WHAT IS THE NEW SYSTEM DEVICE?
Type the three character designation
DKO-DK3, DTO-DT7, or RXO-RX5.

the

new

system

device,

DO YOU WANT TO TRANSFER YOUR FILES?
Answer YES and the CR key to move the rest of the system programs
(PIP, COMP, etc.)
to the new system device. Transfer of the system
programs destroys anything stored on the new system device. Answering
NO and the CR key will zero the directory of the new device.
If a VT50 is not present, the system then asks:
DO YOU HAVE AN LA36 AS TERMINAL?
Answer YES and the CR key to have this
Answer NO and the system answers an LA30.

system

provide

for

LA36.

The

DO YOU HAVE AN LA36 AS A LINE PRINTER?
Answer YES and CR key to have
question:

system

provide

for

HOW MANY LINES PER PAGE?
If the answer to the previous question was No, the above printout is
bypassed and the system answers in LP08 or LS8E printer and then asks:
132 COLUMN PRINTER?
is displayed. Answer YES and the CR key if there is a 132 column line
printer.
Answer NO and the CR key and SYSGEN assumes an 80 column
line printer. The next question is:
IS EVERYTHING CORRECT?
Check all answers made to the above questions. Type YES and the CR
key if the answers are correct. Type NO and the CR key to correct any
errors. SYSGEN repeats the questions starting at the beginning.
SYSGEN/C does not reset the logical unit assignments to reflect the
new area occupied by the system on a disk.
Run SYSGEN/T to reassign
logical units if the system is moved to a disk or the system may be
destroyed by a data file.

3-6

If transferring from a disk containing more than 46 segments to a
DECtape, SYSGEN/C will
not permit the transfer due to insufficient
room on the DECtape.
This is manifested by the DECtape rocking back
and forth at the end of tape.
Reboot the disk system and run SYSGEN/C
(without transferring files) onto two DECtapes.
Then run PIP,
option
E, and put source files on one DECtape and binary files on the other.
NOTE
LINES-PER-PAGE may be controlled by use
of the program LINCHG (see chapter 16).

3.5

LOGICAL UNIT ASSIGNMENT FOR THE COS SYSTEM

The assignment of logical
units to physical mass storage devices
provides greater utilization of the storage area and a certain device
independence.
The device independence is available at run time;
any
mass storage device can be specified for I/O and the program executes
using the specified devices.
The COS system handles storage using the following hierarchy
2 characters
256 words
16 blocks

1 word
1 block
1 segment

Refer to sections 3.4.1 and 3.4.2 for background on logical
unit
assignment before proceeding.
Devices not currently part of the
system configuration can not be assigned a logical unit.

3.5.1

How Logical Units are Assigned by SYSGEN

On the system device, disk or DECtape, logical units are assigned from
the end working
toward
the beginning of the device, but in reverse
order.
For example, a disk with 3 logical units would be arranged,
starting
from
the beginning as:
Directory;
Monitor;
System
programs; User programs;
Unused space;
Logical Unit 1;
Logical
Unit 2;
Logical Unit 3.
If more logical units are assigned, then
logical unit 1, etc. will be moved closer to
the beginning of the
device.
On non-system devices, disk or DECtape, logicat units are assigned
in
sequential order
starting at the beginning of the disk or DECtape.
For example, a disk with 4 logical units might be arranged:
Logical
Unit 1;
Logical Unit 2;
Logical Unit 3;
Logical Unit 4;
Unused
space.
There is a total of 406 segments available on an RK05 disk and
46 segments on DECtape.

3-7

For example, with the following log ical unit assignments:
Logical unit

Device

Size (Segments)

1
2
3
4
5
6

DKO
DKO
DKO
DKI
DKI
DKI

40
20
10
60
40
20

would cause an RK05 system disk (DKO) and non-system disk (DKl)
organized as follows:
System Disk

Non-system Disk

DK.0

DKI

LU3
LU2
LUI

UNUSED
SPACE

USER
PROGRAM
(SOURCES &
BINARIES)

UNUSED
SPACE

co
N

M
M

SYSTEM
PROGRAMS
LU6
MONITOR
LUS
DIRECTORY

LU4

3-8

3.5.2

COS-300 Logical unit Assignment

To make logical unit assignments from the terminal, type
~RU

SYSGEN/T

and the CR key.
Type the 3 character device code, a comma and length
in segments in the order they are to be assigned to the logical units.
The comma and length in segments can be omitted and SYSGEN assigns all
the space left on the specified device to the next logical unit
(DECtape = 46 segments;
RK08 Disk,
203, segments;
RK8E,
406
segments;
RF08 disk, 64 segments times the number of platters). The
first device entered is automatically assigned to logical unit 1,
the
second to unit 2, etc. For example,
.RU SYSGEN/T
DKO,20
DKO,20
DKO,20
DKl,lOO
DKl,lOO
DKl,lOO
DKI
END
The result of the above SYSGEN will be to assign logical units 1,
2,
and 3 to RX8E drive 0; and logical units 4, 5, 6, and 7 to drive 1.
Logical unit 7 will be 106 segments long (406-300).
If an error
is
made when typing the table, type CTRL/C then retype the RUN SYSGEN/T
command ahd start the table over again.
When all desired logical
units have been specified, type CTRL/Z.
Logical assignment can also be accomplished by typing the table into
the Edit Buffer. This method allows error correction with the Editor
commands.
If the following logical unit assignments are typed to the
Edit Buffer,
0010
0015
0020
0025
0030
0040
0050
0060
0070
END

DTI
DT2
DT3
DT4
DT5
DT6
DT7
DKO,46
DKO,46

and the command
. RUN SYSGEN
and the CR key is typed, SYSGEN makes the logical unit assignments
according to the table in the edit buffer.
Logical units 1-7 are
DECtape units 1-7 with a length of 46 segments and logical units 8 and
9 are portions of RK08 disk (DKO), each 46 segments long. A listing
of the SYSGEN assignments is always printed on the line printer upon
successful completion of the SYSGEN program.

3-9

The table created with the Monitor can also be stored with
command for future use. Use the command
~RU

the

WRITE

file.

For

SYSGEN, file

to assign the logical units
example

with

previously

created

.RU SYSGEN,TABLI
where TABLI contains
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
END

DTI
DT2
DT3
DT4
DT5
DT6
DT7
DKl,202
DK2,101
DK2,101

Occasionally, it is desirable to check the logical
table. To obtain a printout of the table, type

device

assignment

.RUN SYSGEN/L
and the SYSGEN table of the current logical
printed on the line printer. For example:
LOGICAL
UNIT #
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15

device

assignments

LENGTH
(SEGMENTS)

DEVICE
NAME
DTI
DT2
DT3
DT4
DT5
DT6
-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-

0046
0046
0046
0046
0046
0046

This table should be made available for reference by all persons using
the COS system.

3-10

3.5.3

COS-310 Logical Unit Assignment

For use with the RX8E diskette (COS 310 only)
...:..RU SYSGEN/T
RXl,41
RX2,41
RX3,41
RX4,41
END
/T option takes input from the terminal, but, produces no output •
. RU SYSGEN (or .RU SYSGEN/B)
RXl,41
RX2,41
RX3,41
RX4,41
END
/B (or the absence of a switch) option takes input from
and produces no output.

Edit

Buffer,

After /T or /B option has been run the following may be performed:
..:.RU SYSGEN/L
LOGICAL
UNIT #

DEVICE
NAME

LENGTH
(SEGMENTS)

01
02
03
04
05
06
07
08
09
10
11
12
13
14
15

RXI
RX2
RX3
RX4
-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-

41
41
41
41

/L puts out the contents of logical units table.
It may be combined
with /T, /B,
or filename option of SYSGEN to perform both functions
with one command.
The /L option must always follow the other option.
~RU

SYSGEN, filename/L

RXl,41
RX2,41
RX3,41
RX4,41
END
The following is_put out on the line printer terminal.

3-11

LOGICAL
UNIT fi:
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15

DEVICE
NAME

LENGTH
(SEGMENTS)

RX1
RX2
RX3
RX4
-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-

41
41
41
41

DEVICE
NAME

LENGTH
(SEGMENTS)

---

.RU SYSGEN/TL
RX1,41
RX2,41
END
LOGICAL
UNIT#
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15

-------

RX1
RX2

41
41

-UNDEFIN~D-

-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-UNDEFINED-

Physical mass storage devices are limited by the hardware~
but can be
subdivided
into logical units.
Logical units are composed of
segments~
the number of segments per unit is up to the user.
COS
allows one file per logical unit.
NOTE
The COS-300 system is shipped
with
logical units 1-7 assigned to DECtape
units 1-7 and each unit is originally
assigned 46 segments.
Logical units
8-15 are left unassigned.
The user may
assign and reassign any of the 1-15
logical units as desired.
Logical unit
o is always the system device and cannot
be reassigned.

3-12

The COS- 310 system is shipped with
logical units 1-7 assigned to RX8E units
1-7.
Each unit is initially assigned 41
segments.
It is advisable to create logical units of varied sizes since file
lengths vary and a short file in a long logical unit wastes storage.

MEDIA PHYSICAL SIZES IN SEGMENTS
device

segments
202
406
64/PLATTER
46
41

RKOl
RK05
RF08
DECTAPE
RXOl

3.5.4

DECtape Users

Although more than one logical unit can be assigned
to a physical
DECtape unit, it is advisable to assign one logical unit per DECtape.
The use of two or more units per DECtape may cause excessive tape
spinning;
for
example,
if reading
from one unit and writing to
another located on the same physical device.
In the same example,
if
the output unit filled and had to be replaced, it might be necessary
to dismount the tape, ~hereby also removing the needed input unit.
A DECtape is 46 segments long so the logical
unit assigned
physical DECtape unit should be assigned 46 segments.

3.5.5

the

Disk Users

The easiest method of assigning logical units to a disk is to think in
terms of a sequence of DECtape.
A DECtape is 46 segments long.

Dectape
A RF08 platter is 64 segments;
segments.

I
another

The RK8E disk cartridges are each 406 segments and there can be up
four drives (0-3).

and a second platter adds
128

64 65
RF08 platter

RF08 platter

3-13

406
DKO

406
DKl

406
DK2

406
DK3

The RK8E disk cartridges are each 406 segments and there can be up to
four drives
(0-3).
If the disk is the system device, approximately
200 blocks (or 12 segments)
must be left unassigned to hold the
operating system and system program.
In addition, space should be
left on the system device to store source programs, control programs
and binary programs.
The remainder of drive 0 may be used for this
purpose.
However, fifty segments are
suggested
which
leaves
approximately
342
segments
(140
for
RK08)
for logical unit
assignments.
Logical units on the system device are automatically
assigned at the end of that device (the opposite is true of all other
devices).
For example:
DKO (assuming RK08)
12

406

SYSTEM
Space for user
PROGRAMS source and
binary files
Log ic al Un itO
which leaves approximately 342 segments for logical unit assignments.
A sample assignment might be
DKO (assuming RK8E)
1

406
SYSTEM

User source
& Binary
Files

logical unit

46
SEGMENTS

46
46
SEGMENTS SEGMENTS

Log ical
unit 8

Log ical
unit 9

-------Log ical
unit 10

The area between logical unit 8 and the system can be left unassigned
for
system program overflow, if desired.
On devices other than the
system device SYSGEN starts assigning logical units at the beginning

3-14

of the device.
SYSGEN Error Messages are located in Appendix J.

3-15

CHAPTER 4
COMP

The Compiler takes a DIBOL language program and converts
it to a
binary language program which can be executed by the run time system
(RSYS).
In the process of creating
the executable program,
the
Compiler outputs a listing of the source program and a storage map of
the records and fields which are used by the program.
The Compiler sets aside storage space for the constants, variables and
statements
used by the program.
Refer
to Appendix C for
an
explanation of the amount of code generated by the Compiler.
Be sure to document any new programs completely and accurately and
to
prepare a set of instructions for execution including any necessary
changes to logical device assignments, names of files, and where each
is to be mounted, the format of any entries to be made to the program,
restart procedures and any other essential information.

4.1

OPERATING PROCEDURES

The command to run the Compiler has the form:
. RU COMP [, filel, ... , file7] [/xx]
and the CR key.

(Be sure the line printer is turned on.)

Where
filel, ... ,file7

are files on the system device to be compiled into
one binary.
These files contain segments of one
program, not several programs.
If no files are
specified, the program in the edit scratch area is
compiled.

/xx

is one or a combination of the following options:
suppresses the output of
listing and storage maps.

compiles the source program and if no errors
are detected, executes the binary program
with no user intervention. /N is implied.

4-1

the

Compiler

The
message
LOADING is displayed when
compiling is successfully completed.
The
program must have an END statement to be
compiled and executed with the IG option, if
IN IT SYS is used in the program.

11-7

specifies a physical unit (DECtape only) for
temporary storage during compilation.
The
unit specified should contain a scratch
tape.
File is recopied on the system device
when compilation is complete.
Use
of
temporary storage units minimize tape motion
during compilation.
enables the TRACE function and
IG option. For example,

implies

the

RUN COMP/T2

performs an implied IG option and then
chains
to
the
DDT
(DIBOL
Debugging
Technique)
mode which is
explained
in
Appendix H.

compiles the source program in less than the
amount of memory space required by the other
options, at the expense of eliminating
the
TRACE feature
(if stated in the source
program)
and accurate
error
reporting.
Execution speed of the compiled program is
increased by as much as 20%.
This option
can be combined with either the IN, IG, 10,
IT, or 11-7 option.

Use the SAVE command to store
Compiler.

the

binary

program

prepared

To calculate how much memory space is saved by using
the following rules apply:

the

Saves 1 location for each executable statement.

Saves 1 additional location for each label.

Costs 1 location for each ON ERROR statement.
ERROR statements are executable.)

the

option,

(However,

Since the TRACE statement in a program compiled with the 10 option has
no useful effect, and Slnce any run-time error reported in such a
program will be reported to have occurred in the PROC statement, it is
suggested that the 10 option only be used on those programs which have
been thoroughly debugged.
Unless the IN or IG option is specified
in the RUN command,
the
Compiler produces a listing of the source program (Figure 4-1) and a
storage map (Figure 4-2) on the line printer or the device specified
in the START, PROC or END statements. Most errors detected by the
Compiler are indicated by underscoring the number of the line where
the error exists and inserting a caret ( ) pointing in most cases to
the error. Other errors are listed in the symbol map.
These errors
must be corrected before an attempt is made to execute the program.

4-2

The number of errors detected by the
terminal as
nn ERRORS

4-3

compiler

displayed

the

COS OlBOL

01-MAR-73 THUR
COMP1~ATI0N LISTING
V12;2 A PAGE ~1
DATA UIVISION
S QUA R E ROO T S U 8 R 0 UTI N E

ROO T

k1~10
k102~
~~3e1

t~ECORD

004~

A15
A15

k1~50

0k:160
0fC7Vl
~"8~
009k)

010~

SfJRTX,
TEMPX,

011~

PR I NT

~H.. CORD

815
f)15
D15
0.5, i2l(~ 1

0120

Figure 4-1
Compiler Listing
(sheet 1 of 2)

4-4

SUBROLJT

1 ~E

COS Ol80L
01~MA~-73 T~UR
PRDCEDURE DIVISION
~+30

COMPI~ATION LISTI~G

V12128

PROC 1

~+4"

0+5kl

INIT(l,l..PT)

0160

X=\J .. ~0eJ0000
CALL SQRT
A=N

f(J+7~

"'1810

S = SQRTX,'XXX;XXX'

ellCJ0

INCR N
XMtT(l,P R INT)
IF (\.LE!99) GOTD ~onp

0~00
C121~
1!122~
el23~

STOP

el24~

'-"250
026~

0270

SQRT,
TEMpX=X/2
iTRIAL VALUE
SQLOOP, SQRTX=(TE~PX+(X/TlMPX»/2
JNEWTON~S METHOD
IF(SQRTX·TEMPX ,EQ~ ~) RETURN

"'28~;

Tt.~pX=SQRTX

"29~

';OTO SQLOOP

el.500

END

Figure 4-1
Compiler Listing
(sheet 2 of 2)

4-5

PAGE ~2

~QS
1#

"001
0"fC'I2

0003
00214
"005
212106
"0217

0010

""11

0180l..

NAME

PRINT
A
S

RECORD
AL,PHA
AI..PItfA

01
""1
01

Oe:CMA~

OECMAL.
OECMA\.

N
t',l

OECMA~

,,-OOP
!

"013

SQRT

21"1'
21016
02117
02120

• ,99

21 °16

STORAGE MAP L.lSTlNG
DIM SliE
ORIGIN

SQRTX
TEMPX

00l,2

""14

rca"MAR",73 THUR
TYPE

',1000

'.', XXX".
. '.2

SQI.OOP
, ", 0

OECMAI..
L.ABEL.
OECMAL.
L.ABEL,.
AI..P~A

oE:CMAI,.
DECMAL.,
I. ABEl..
OECMAL

32
l'

15
15

ft:J1
01

2"""2

20021
221042
221061
20101lJ

221117

20122
UUl71

20123

0el

10132

2014:1.
2121143
1121141
20144

01
01

07
02

PAG! "3

221000

01
01

'1:1.2128

201.32

~YM80k,S

NQ ERRORS DETECTEO~

08 K CORE REQUIRED (3940 rREE 1.0CS • 14 ayrrERSJ

Figure 4-2
Storage Map Listing

4-6

The storage map lists the names of all the records, fields, and labels
as they were processed by the Compiler. The information is arranged
in six columns as follows:
this column contains the internal number of the symbol named
in column 2 and may be ignored.
NAME

Column 2 gives the name of any symbol
(field name, record
name, program label)
or constant used
in the compiled
program.
Constants are
decimal
constants
or
alpha
characters which appear
in the PROC section of the source
program. Only the first six characters of a symbol name are
used;
any characters after the sixth are ignored. Only the
first four characters of a constant are used and each
constant is preceded by two periods ( .. ) to distinguish it
from a symbol.
Constants with four characters or less
appear only once on the symbol map even though they may
occur more than once in the program.
Constants with more
than four characters are listed each time they occur in the
program.
Record literals begin with a" •. ".

TYPE

Column 3 states the symbol type.

The types are:

ALPHA

Symbol is an alpha field name or an alpha constant
(character string).

DECMAL

Symbol
is
constant.

RECORD

Symbol is a record name or record literal.

LABEL

Symbol is a program label.

REDEF

Symbol is multiply defined
(redefined).
All
attempts at definition after the first are flagged
as errors in the source listing.

UNDEF***

Symbol is a label referenced by the program but
not defined.
For example, GO TO TAGI in a program
where TAGI does not appear as a label.

decimal

field

name

decimal

This error is output to the line printer even
though the IN option is in effect. The line
number displayed is the line where the symbol
is
first used.
DIM

Column 4 contains the array dimension (number of fields)
of
the alpha or decimal symbols. The column is meaningless for
other types of symbols.

SIZE

Column 5 lists the size of the symbol in characters.
The
size of a RECORD symbol is the number of the characters of
all its symbols plus 2.

ORIGIN

Column 6 gives the 15-bit (octal)
byte address where the
symbol appears in core and for most users should be ignored.
Octal addresses are shown for labels.

The number of symbols used, number of errors detected, core required
and free locations are listed at the bottom of the symbol map.

4-7

Maximum number of symbols allowed in an 8K system is 365;
larger systems, 511.
4.2

12K

CONDITIONAL COMPILATION PROCEDURE

The Conditional Compilation Procedure (CCP) is a feature which permits
the DIBOL programmer to include statements in a source program which
will be compiled by the compiler only if the programmer elects for
those statements to be compiled.
Statements included in a program for conditional compilation
enclosed with angle brackets as in the following example.

are

RECORD A
Bl,
05
Cl,
A4
PROMPT, 01
RECORD N
NAME, A6
PROC
<PROMPT
XMIT(8,"ENTER NAME:")

>
XMIT(7,N)
STOP
END
Notice that the left angle bracket is followed by a control variable,
in this case PROMPT.
Until
the control variable is turned "on"
somewhere in the program before the left angled bracket is encountered
by the compiler, those statements between the angled brackets will be
ignored.
Notice too, that the right angled brackets appears on a line
by itself.
The command to turn on a control variable is as follows:
=control variable
The example above shows a program that requires the operator
to
type
in a name on the keyboard.
If the program is recompiled with the
control variable PROMPT on, it produces a DIBOL program which first
displays a message to the operator.
RECORD A
Bl,
D5
Cl,
A4
PROMPT, Dl
RECORD N
NAME, A6
PROC
=PROMPT
;TURN ON PROMPT
<PROMPT
XMIT(8,"ENTER NAME:")

XMIT(7,N)
STOP
END

Conditional compilation can also
be
used,
for
example,
to
conditionalize debugging statements into a source file.
Once the
program has been tested, the control variable can be shut off by
deleting the command to turn it on, effectively removing the debugging
routines from the source.

4-8

CCP can also allow you to combine several similar
source programs into one source.

(but not

identical)

If the control variable used in a CCP statement is undefined,
the
compiler will automatically set aside space for it.
This, however, is
wasteful of space and it is to the programmer's advantage to use,
for
CCP symbols,
symbols that are already being used for some other
purpose.
The CCP value of a symbol
ordinary DIBOL value.

(on or off)

is independent of

the

symbol's

If a CCP variable is used in the middle of a record definition (in the
record section of a DIBOL program) it must be a variable that has
previously been defined, because if not, the compiler will allocate
space for
the variable in the middle of your record (where not
desired) .
CCP sections may be nested to any depth. Any sections inside a
CCP
section that is off will be completely ignored by the compiler.
This
includes other CCP statements.
To indicate that certain statements
are being ignored (conditionalized out), the compiler listing will not
print the line number for that statement.
There must be a matching >
for each < used.
If this condition is not met, the compiler generates
a CCP ERROR message.
This error is fatal if angled brackets do not
match by the end of the program.

4.3

ERROR MESSAGES

Most Compiler error messages are printed on the source listing
directly after the line in which the error occurs. A caret ( ) in the
error message points to the approximate location of the error.
Other
errors are listed in the symbol map.
The Compiler error messages are:
Message

Explanation

BAD ALPHA VALUE

Initial
value
in
an
alpha
data
specification statement did not begin or
end with a single quotation mark.

BAD DECIMAL VALUE

The initial value for
a decimal data
specification was incorrectly formed.

BAD PROC #

The number in a PROC statement was not a
digit from 0 to 7.

BAD RELATIONAL

An illegal relational occurs in an
IF
statement.
For example, a .GX.
instead
of a .GT.

4-9

Expl ana tion

Message

one

was

Matching
missing.

COMMA MISSING

No comma
expected.

DATA INITIALIZATION MISSING

No data initialization followed a comma
in a data specification statement.

EXPECTED LABEL IS MISSING

A required label is missing.

EXPRESSION NOT ALLOWED

A complex expression or
occurs to the left of an
a variable is allowed.

EXTRA CHARS AT STMNT END

Extra characters occur at the end
legal statement.

FIELD TOO LARGE OR 0

In a data description statement, the
dimension was 0 or more than 3 digits
long, or the field size was 0 or larger
than 511.

ILLEGAL OPERATOR

A bad character was encountered in an
expression where an operator would be
expec ted.

ILLEGAL STMNT

The
statement
was
not
a
data
manipulation statement (it had no =) nor
did it start
with
a
recognizable
keyword.

INITIAL VALUE WRONG SIZE

The
initial
value
in
a
data
specification statement had a length
different from the field size specified.

LABEL NOT ALLOWED

A symbol in an expression was not of
type alpha or decimal, or a record or a
symbol which had been redefined was
used.

MISSING CLOSE PAREN

No close parenthesis occurred where
was expec ted.

one

MISSING OPEN PAREN

No open parenthesis occurred
was expected.

one

MISSING OPERAND

A
binary
operator
occurs
in
an
expression with no operand following it;
or no expression at all occurs where one
is expected.

MISSING OR BAD DEVICE

The device in an INIT statement was
missing
or started with an illegal
character.

4-10

angled

(

CCP ERROR

bracket

appeared

where

bad

character

= or where only

where

Message

Meaning

MISSING QUOTE

The statement contained an odd number of
quotes (').

MISSING RELATIONAL

No relational
statement.

NAME PREVIOUSLY DEFINED

The name used was previously defined and
this statement tries to redefine it.

NOT A OR D

A character other than A or D occurred
in a data specification statement where
A or D was expected.

NOT LABEL

A symbol which was not
a
'label'
occurred where a label was required.

PROGRAM TOO BIG

Binary output too big for
the binary
scratch area. Remedy: Run PIP with In
option, (refer to Chapter 5).

RECORD TOO BIG

A named record exceeded
in size.

STMNT TOO COMPLEX

The statement generated too much code
and
overflowed
the Compiler's code
buffer, or it had too much nesting and
overflowed
the
Compiler's push-down
stack.
Remedy: break up the statement
into smaller parts.

SUBSCRIPT ERROR

No comma or close
after a subscript.

SUBSCRIPT NOT DECIMAL

The type of a subscript was not decimal.

TOO MANY SYMBOLS!

A fatal error message. Only 365 symbols
allowed in symbol table in 8K system,
and only 511 symbols allowed in larger
systems.
The compiler stops compiling;
no storage map can be produced.

TOO MUCH DATA

Program's
bytes.

UNDEFINED NAME

A name is used which was
in the data section.

WRONG DATA TYPE

Mixed modes occurred in an expression;
or an argument which was supposed to be
decimal was not or one of the three
arguments
in
a
data
manipulation
statement was of the wrong type.

4-11

Data

appeared

510

characters

parenthesis

division

occurs

exceeds
never

32K

defined

CHAPTER 5
PIP

PIP (Peripheral Interchange Program) handles file transfers between
two logical devices,
i.e., moves data, source and binary files, and
system programs between two file-oriented devices;
copies
the
contents of one device to another, and consolidates files to remove
free blocks.
PIP can also be used to allocate more space to the
binary scratch area.

5.1

OPERATING PROCEDURES

Type the Monitor command:
• RU PIP [/n]

and the CR key to load PIP. /n is optional and is a number 0-9 which
indicates the number of segments to allocate to the binary scratch
area. The /n switch is used in conjunction with the E option.
Refer
to section 5.1.4 for further information.
PIP responds to the RUN command with:
PIP 5.05 (or current version number)
OPT=requesting the type of transfer and I/O devices. Answer with one of
the options, B, C, D, E, S, V, X, described on the following pages.
After the option is specified, PIP displays the IN and OUT questions
requesting the input and output files.
If an incorrect character is
entered when responding to the OPT, IN or OUT questions, type CTRL/U
and the correct response. Type CTRL/C to return to the Monitor.
NOTE
When PIP is used under a COS-3lO system
PIP will accept either "RX" or "DT" as a
legal mnemonic, depending on the type of
device on the system.

5-1

OPT-B
5.1.1

Transfer Binary File

Type B to move a binary program between two file-oriented devices.
Answer IN with the name of the binary program to be transferred and,
optionally, a comma, and the input device, DTO-DT7, DKO-DK3. If no
device is specified, the system device is assumed.
Answer OUT with the name to be assigned to the output file, and,
optionally, a comma, and the output device, DTO-DT7, DKO-DK3. If no
device is specified, the system device is assumed.
If an attempt is made to transfer to or
device, the IN or OUT message is repeated.
Example:
B

TEST,DTl
TEST,DKl
(CTRL/C typed here)

5-2

from

non-file-oriented

OPT-C
5.1.2

Copy Device

TYPE~

C in reply to OPT- to copy the contents of an entire device onto
another similar device, i.e., disk to disk or DECtape to DECtape.

Answer IN
DKO-DK3.

with

the

three-character

physical

device,

DTO-DT7,

Answer IN
DKO-DK3.

with

the

three-character

physical

device,

DTO-DT7,

Answer OUT with one of the physical devices as explained for IN.
Example:
OPT- C
IN- DTO
OUT- DT7
OPT-

(X or CTRL/C typed here)

..!..

NOTE
Do not use
this
option
to
copy
diskettes.
A new program,
RXU,
is
available for
this
purpose.
(See
Appendix M for details) .
NOTE
PIP option C cannot be used when the COS
300 Foreground Data Entry Program is
running.

5-3

IOPT-DI
5.1.3

Transfer Data Files

Type D to move data files
OUT.
Answer
IN with a
must be alphanumeric), an
by a slash or comma, or a

IR
IC
IK
IF

between the devices specified after
IN and
label (up to six characters, first character
optional logical unit number (1-15) preceded
device switch:

for paper tape reader (COS 300 only)
for card reader (COS 300 only)
for keyboard
BCD paper tape in high speed reader (COS 300 only)

If IR is specified as the input device, records stored are variable
length.
If IC, the record produced is the same size as the card even
if a smaller number of columns are punched.
If a label is specified without a logical unit number or if both
specified and a data file of another name is found, PIP asks
MOUNT

xxx xxx

are

#01 FOR INPUT:

answer with the appropriate logical unit number
(1-15).
Answer OUT
with a label, number or device as described for IN.
Output devices
are:

IL
IP
IT

line printer
paper tape punch (COS 300 only)
terminal

If a label is specified without a default
message
MOUNT xxxxxx #01 FOR OUTPUT:

unit

specification,

is displayed. Answer with the appropriate logical unit number
If a file with a different name already exists, Monitor asks
REPLACE

the

(1-15).

xxxxxx#nn?

Type Y and the CR key to destroy old file and replace with new.
Type
N and
the CR key to return to the OPT- question.
If output device
specified is a logical unit (1-15) and the unit is filled,
an error
message is produced.
When at the end of the input file or card deck, PIP asks
MORE?
Answer with the CR key if there is no more input, or
label, logical unit or device.

5-4

specify

another

NOTE
PIP transfers data in alphanumeric mode
only.
A negative number is treated as
the letter which has the equivalent
code.
(Refer
to
Appendix A,
COS
Character Codes.)
Examples:
Explanation
OPTINMOUNT
1
OUTMORE?
OPT-

Dump EMPNAM
file on logical
unit 1 on the
line printer.
CTRL/C is typed to
. . eturn to the
Monitor.

EMPNAM
EMPNAM #01 FOR INPUT:

II
N

HRPAY,2
PAYFIL/I

Combines 2
data files i'nto
one output
file.
CTRL/C is typed
to return to the
Monitor.

SALPAY,3
N

OPTINOUTMOUNT
10
MORE?

.!l!::

MOUNT
11
MORE?
OPT-

LISTI
LISTI

#01 FOR OUTPUT:

Input from
card reader which
creates a
mul ti-un it
file.

IC
LISTI

#02 FOR OUTPUT:

5-5

Log ical un it 10
is full.
CTRL/C is typed
to return to the
Monitor.

5.1.4

Eliminate Spaces in Directory

Type E to consolidate the free blocks of the file on the input device,
and store the consolidated files on the output device. In addition,
it is possible to eliminate one or two of the types of files (source,
binary, and system) during the consolidation. These free blocks are
shown in the file directory and occur when a file is deleted from the
directory. The Monitor is not copied.
Answer IN with DTO-DT7, or DKO-DK3 and the CR key.
Answer OUT with DTO-DT7, or DKO-DK3 and the CR key.
When consolidating the system device onto itself, PIP
free space as shown below:

eliminates

the

SYS
f

before

files

r
e
e

files

f
r
e
e

files

log ical
unit

logical
unit

log ical
unit

logical
unit

SYS
after

files

free

However when consolidating a device (other than the system
onto another device PIP eliminates the free space as follows
DKO
before

files

f
r
e
e

files

f
r
e
e

log ical
unit
8

log ical
unit

DKI
after

files

free

Example:
OPTINOUT-

E
DTO
DTO

The message
TYPES OF FILES TO BE SKIPPED (S,B,V,F,M):

5-6

device)

is displayed. Answer with the CR key if all files are to be
transferred, or answer by typing one or a combination of two of the
characters S, V, B separated by a comma and followed by the CR key if
one or two of the types of files are to be eliminated.
For example:
~R

PIP../
PIP 5.05
OPT- E..,..J
IN- DTO.-J
OUT- DKO..)
TYPEs OF FILES TO BE SKIPPED (S, V,B): S ~
This would consolidate free space and eliminate source files located
on DTO and store the consolidated files on DKO assuming there was a
directory on DKO.
NOTE
PIP option E cannot be used when the COS
300 Foreground Data Entry Program is
running.
NOTE
All control keys are ignored until
option E operation is completed.

the

Binary Scratch Area Modification:
As mentioned in section 5.1, the E option can be used in conjunction
with In to change the size of the binary scratch area. The binary
scratch area, normally two segments (32 blocks) long, is adequate for
most applications but extremely large DIBOL programs (containing many
statements) may need more space.
A typical sequence for expanding the binary scratch area is:
.,!.R

PIP/2.J

PIP 5.05
OPT- E~
INDTO..)
OUT- DKO..)
TYPES OF FILES TO BE SKIPPED (S,V,B) : . /
This would allocate 64 blocks to
doubling its original size.

the

binary

scratch

area

DKO,

In general,
1.

If the output device is not the input device,
the binary
scratch area on the output device equals 2+n segments.

If the output device is the input device, the binary scratch
area is set either to the size of the current area or to 2+n,
whichever is less. That is, compressing a device onto itself
can shrink the binary scratch area, if 2+n is less than the
size of the original area. The binary scratch area cannot be

5-7

expanded if a device is being compressed onto itself since
that would require writing over existing files.
3.

If no (n is sp~cified, the size of the binary scratch area on
the input device is assumed.

5-8

5.1.5

Transfer Source Files

Type S to transfer source files between two file-oriented devices.
Answer
IN with the name of the source file to be transferred and,
optionally, a comma, and the output device DTO-DT7, or DKO-DK3.
If no
device is specified the system device is assumed.
Answer OUT with the name to be assigned to the output file and,
optionally, a comma, and the output device, DTO-DT7, or DKO-DK3.
If
no device is specified, the system device is assumed.
If an attempt is made to transfer to or
device, the IN or OUT message is repeated.

from

Example:

S.J

TEST,DTl~

TEST,DK1...,J
(X or CTRL/C typed here)

5-9

non-file-oriented

IOPT-V I
5.1.6

Transfer System Program

Type V to move a system program between two file-oriented devices.
Answer IN with the name of the system program to be transferred and,
optionally, a comma, and the input device, DTO-DT7 or DKO-DK3.
If no
device is specified, the system device is assumed.
Answer OUT with the name
optionally, a comma, and
device is specified, the
made to transfer to or
message is repeated.

to be assigned to the output file, and,
the output device, DTO-DT7 or DKO-DK3.
If no
system device is assumed.
If an attempt is
from a non-file-oriented device the IN or OUT

Example:
V~

SORT,DT1..}
SORT,DT4..)
(X

or CTRL/C typed here)

NOTE
PIP option V cannot be used when the
Foreground Data Entry program is running.

5-10

COS

300

5.1.7

Return to Monitor

Type X in response to OPT- and
return to the Monitor.

the

system

will

terminate

PIP

and

This feature is particularly useful when PIP is included in a string
of Monitor commands in a BATCH program. The X option signals the end
of the PIP program and the next Monitor command in the BATCH program
is executed.

5-11

5.2

ERROR MESSAGES
Message

Explanation

BAD DIRECTORY

Attempt to reference or store a file on a
device with no directory (or a directory
that has b€en destroyed) .

ILLEGAL DEVICE SWITCH

A switch was specified that was not IR,
IC
or IK for input or IP, IL or IT for output.

NO ROOM

Attempt to store a file on a full device.

PIP INITIALIZATION ERROR

PIP unable to identify the type of devices
resident in the system.
Following this
message, PIP attempts to return control
to
Monitor.

5-12

CHAPTER 6
BUILD

BUILD is a utility program used to create data files containing fixed
length records.
Creating data files requires careful consideration of
their proposed usage.
(Designing BUILD records is discussed
in
Appendix D.)
BUILD allows input from the keyboard, card reader (COS
300 only), high-speed paper tape reader
(COS-300 only), or system
device.
A BUILD control program must be written before BUILD can be run.
This
control program provides the field description and order for input and
output of the data to be used to create the file.
Be sure to prepare
a set of instructions on input preparation so data input will be
according to specifications in the control program.
During execution, BUILD checks each data input line to be sure the
data entered conforms exactly to specifications in the control
program.
If no errors are found, the input line is written on the new
data file and maybe printed on the line printer.
If an error is
detected, the record is not written to the data file.
BUILD outputs a listing of the control file and data records created
on the line printer, if desired. A summary of the number of records
written, number of records not written and any hash totals kept is
printed on the line printer for each file created.

6.1

BUILD CONTROL PROGRAM

The BUILD control program is created and written
device with editing commands.

mass

storage

This program conveniently falls into three sections, Field Descriptor,
INPUT, and OUTPUT.
Briefly, the Field Descriptor section defines all
the fields to be used, the type and size and any initial values to be
assigned.
.
The INPUT section describes the order
in which the data will be
received.
The input format specified in this section must be followed
when preparing data for input. The OUTPUT section describes the order
of the output of data.
There may be up to seven output formats for
one set of input so several different output files can be created with
one input file.

6-1

Figure 6-1 is an example of a simple BUILD control
explained in detail in the following sections.

program

which

Errors in the BUILD control program cause a return to the Monitor.
Spaces and tabs may be inserted anywhere in the program which is
explained in detail in the following sections.
Errors in the BUILD control program cause a return to the Monitor.
Spaces
and tabs may be inserted anywhere in the program for
readability. As with any COS program, comments inserted on a line
must be preceded by a semicolon.
The control program may be stored as two separate source programs:
the Field Descriptor section as one and the INPUT and OUTPUT sections
as the other •
;AUTOMATIC LINE NUMBER COMMAND,
;START,AT 10, INCREMENT OF 10

• LN ,10
.0010 DEFINE
.0020 Fl,D6
.0030 F2,A30
.0040 F3,D7
.0050 F4,D6
.0060 F5,D6
.0070 Tl,DlO
.0080 Cl,D2,25
.0090 C2,D6,50000
.0100 INPUT CDR
.0110 PART
.0120 Fl
.0130 F2
.0140 F5
.0150 DATE
.0160 F4
.0170 VENDOR 1
.0180 F3
.0190 OUTPUT 1 PRTFIL
.0200 F3 C
.0210 Fl C
.0220 F2
.0230 F5
.0240 F4
.0250 END

;PART NUMBER
iPART NAME
;VENDOR NUMBER
; DATE
;WHOLESALE PRICE
;TEMPORARY STORAGE
;CONSTANT FIELD
;CONSTAN FIELD
;INPUT FROM CARD READER
; KEYWORD
;PART NUMBE
;PART NAME
;WHOLESALE PRICE
; KEYWORD
; DATE
; KEYWORD
;VENDOR NUMBER
;OUTPUT TO PRTFIL FILE
iVENDOR NUMBER
;PART NUMBER
;PART NAME
;WHOLESALE PRICE
;DATE

Sample input lines for above control program
: PART 126789
'HEX BOLT' 200
:VENDOR 1476116
: PART 176117 WASHER 100
: VENDOR 7346177
:PART 176118
:VENDOR 3177615
:PART 126788
'#6 MACHINE SCREW' 300
:VENDOR 2061311
Figure 6-1
Sample BUILD Control Program

6-2

Records created according to control program in Figure 6-1 would look
like:
1476116126789HEX BOLT
7346177176117washer

6.1.1

22 spaces
24 spaces

000200030772
000100030772

Field Descriptor Section

The Field Descriptor section specifies all fields to be referenced
in
later sections of the Control program.
It gives their type (alpha or
decimal), size (1-510) and initial values, if any.
The format for this section is
DEFINE
fieldx, typesize, initial value

where
DEFINE

is the section heading
statement in the program.

fieldx

is an F, T, or C and a number.
The F,
T or C
indicates data fields (F), temporary fields (T), and
constant fields
(C).
These names are for
the
convenience of the user and are treated equally by
BUILD. All F fields must be listed first, all T
fields second, and all C fields last. The number
(x) must be a positive, non-zero decimal number not
greater
than 511 and may begin at any number.
Numbers may be skipped but within a field group
(F,
T or C)
must be in ascending order. Temporary
fields are used to accumulate hash totals. Constant
fields are used to enter a default value or range
check.

and

must

the

first

Remember, however, that Fl is assigned to the first
field,
F2 to the second, etc.
Lf the field names
assigned are, for example,
Fl
F5
F6
Blank fields will be left in core.
would cause an error.

References to F2

typesize

is A for alphanumeric fields or D for decimal
fields.
Use A to describe fields which contain
other than numbers. Size is an unsigned decimal
number in the range 1-510.

initial value

Initial values can be assigned to any fields.
Alphanumeric values consist of any sequence of legal
COS characters (excluding single quotation marks)
and are enclosed in single quotation marks.
Initial
decimal values consist of a sequence of digits

6-3

optionally preceded by a + or - sign or followed by
a - sign. These initial values must be of the same
size and type as the field for which they are
specified. The sign and quotation marks do not
count in the size of the initial value.
If an
initial value is not specified, the field
is
initially set to 0 if decimal or blank if alpha.
Remember, there is a difference between decimal and
alpha constants consisting of the same digits and if
a constant is to be referenced as alpha and decimal,
it must be defined both ways. For example,
Cl,D3,234
C2,A3,'234 1
Field description statements must be entered on consecutive lines and
each line is terminated with the CR key.
Use the editing commands to
create this file.
Example:
.0010
.0020
.0030
.0040
.0050
.0060
.0070
.0080
.0090

6.1.2

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
Tl,DlO
Cl,D2,25
C2,D6,500000

iSTARTS BUILD CONTROL PROGRAM
iPART # FIELD, DECIMAL, 6 DIGITS
iPART NAME FIELD, ALPHANUMERIC, 30 CHARS
iVENDOR # FIELD, DECIMAL, 7 DIGITS
iDATE FIELD, DECIMAL, 6 DIGITS
iWHOLESALE PRICE
iTEMPORARY FIELD
iCONSTANT FIELD CONTAINING VALUE OF 25
iCONSTANT FIELD CONTAINING VALUE OF
500000

INPUT Section

The INPUT section specifies the device (i.e., card reader
(COS
only), keyboard, etc.) and arrangements of the data to be input.
format of the INPUT section is

300
The

INPUT dev
keyword, output format#
fieldx

where
INPUT

is the section header and must
statement of the INPUT section.

the

first

dev

is the 3
device:

for

the

input

character

designation

KBD - terminal keyboard
CDR - card reader (COS300 only)
RDR - high-speed paper tape reader
only)

6-4

(COS300

SYS - a file on the system device
In the case of CDR, RDR and SYS, the input data is
prepared in advance and stored on cards, paper
tape, or the system device.
keyword

is the name assigned to a group of fields which
will make up One input line to the BUILD program.
Keyword is an alphanumeric character string of any
length but only the first six characters are
significant.
Field names cannot be keywords.
Keywords provide the connection between the data
entered at BUILD run time and the data format
specified in the control program.

(comma)

is optional and specifies comma as the field
delimiter
on
input lines.
This allows the
alphanumeric data input for
that keyword
to
contain embedded spaces and tabs.
In most cases,
however, spaces are suggested as delimiters.
For example,
if control program specifies
PART,
then the input line is
PART,126789,HEX BOLT,200
However, if the control program specifies
PART
then the input line is
PART 126788 '#6 MACHINE SCREW' 300

output format#

is a number (1-4 for 8K systems or 1-7 for
12K
systems)
which specifies the output format to be
used for this input line.
This number corresponds
to an OUTPUT section number (described in section
6.1.3).
If an output format number
is
not
specified, there is no output for fields under
this keyword.
If an output # is specified, no
output
takes
place
until
input
for
the
corresponding fields is completed.

fieldx

is one of the previously described fields
(F,
T,
or C)
and a number.
These fields must be listed
in the order in which they will be input.

To enter the BUILD input line without keywords,
(*) in place of a keyword. For example,
INPUT

specify

asterisk

listed

before

CDR

*1
Fl
F2
All field descriptions beginning with the
those beginning with keywords.

6-5

must

NOTE
BUILD input lines without keywords must
be entered in the exact order specified
in the INPUT section of the control
program because there are no keywords to
relate an input line to the fields
described in, the control program.

Data is inserted in

* fields in order of input.

The INPUT section is terminated by the OUTPUT
section of the control program.

heading

for

the

Example:
.0010
.0020
.0030
.0040
.0050
.0060
.0070
.0080
. ·0090
.0100

DEFINE
Fl,D6
F2.A30
F3,D7
F4,D6
F3,D6
Tl,DlO
Cl,D2,25
C2,D6,500000
INPUT CDR

.0110
.0120
.0130
.0140

PART
Fl
F2
F5

.0150
.0160
.0170

DATE
F4
VENDOR 1

.0180

iINPUT IS FROM CARD READER (COS 300
only)
ilST INPUT LINE HAS KEYWORD PART
ilST FIELD OF PART LINE IS PART NUMBER
i2ND FIELD OF PART LINE IS PART NAME
i3RD FIELD OF PART LINE IS WHOLESALE
PRICE
i2ND INPUT LINE HAS KEYWORD DATE
ilST FIELD OF DATE LINE IS DATE
i3RD INPUT LINE HAS KEYWORD VENDOR
iAND ALL INPUT IS TO BE OUTPUT
iACCORDING TO FORMAT 1
iIF VENDOR IS MISSINGi NO OUTPUT OCCURS
;lST FIELD IS VENDOR NUMBER

OPTIONS
The INPUT section of the control program also allows assignment of
default fields,
skips, or errors for missing fields, flag fields,
range checks, hash total fields, and checkdigits.

6-6

These options may be specified for any field and have
format shown:

the

order

and

field~~~~idx'\1[flag] [RANGE ([lfldx] ,[ufldx] )] [HASH (hfldx)] [CHECK]
l;RROR .~

Spaces may be used to separate the options for
readability.
Any
combination of options can be used but they must appear in the order
shown above.
Use one of the SKIP, default (dfldx) or ERROR options to specify an
alternative for a missing field in a BUILD input line. SKIP causes
it
the value previously stored in fieldx to be used again, assuming
was not cleared.
=dfldx assigns the current value of the default field
(dfldx)
to
fieldx.
The default field (usually a C field) must have the same type
and size as fieldx.
ERROR causes an error message on the terminal at
BUILD run time.
If none of these is specified, SKIP is assumed when
fieldx is missing from the BUILD input line.
A constant field
(C)
cannot be used as fieldx.
Use the flag option to set the flag field to 1 when fieldx is present
and zero when not present. The flag field can be checked later in the
user's DIBOL program.
The flag field must be a decimal field.
Use the RANGE option to specify a high and low limit gov the data in
the destination field.
The destination field and the fields specified
as the limits must be decimal fields.
The low limit field
is
specified first and must be less than or equal to the size of the
destination field.
If the low limit field is omitted, there is no
lower limit.
The high limit field is specified second and must be
larger than or equal to the size of the destination field.
If the
high limit field is omitted, there is no high limit check. The value
of the data entered at run time must be within the range specified.
Use the HASH option to request a hash total on the data
in the
destination field.
The destination field and the hash total field
(hfldx) must be decimal. Since the hash total field
(usually a T
field)
is used to store an accumulating sum of the destination field
data, it cannot be smaller than the destination field and should be
larger.
If the hash total exceeds the storage field, the high order
(leftmost) digits are lost. The field used to store the hash total
should not be used for any other purpose within the BUILD control
program.
Use the CHECK option to request a checkdigit calculation on the data
entered in the destination field.
The destination field must be
decimal.
The checkdigit is the rightmost digit and
is not stored.
For example,
if the field is a D6 and the number 1234561 is entered,
the rightmost digit (l) is the checkdigit and is not stored.
Using
the same field size, D6, if the number 12345 is entered, 5 becomes the
checkdigit. The checkdigit formula is calculated on 001234 and 5 is
not stored.
If the checkdigit is not valid, an error results.
The
formula for calculating the checkdigit is described in Appendix E.

6-7

Example:
DEFINE
F1,D6
F2,A30
F3,D7
F4,D6
F5,D6
F6,D1
T1,D15
C1,D2,25
C2,D6,500000
INPUT CDR
PART,
iPART INPUT LINE HAS COMMA AS DELIMITER
F1 ERROR
iERROR MESSAGE IF PART NO. NOT ENTERED
F2 F6
iF6=0 IF F2 NOT PRESENT
F 5 RANGE ( C1 , C2 ) HASH T1
iWHOLESALE PRICE MUST BE
DATE
iWITHIN RANGE SHOWN IN C1 AND C2. STORE
F4
iHASH TOTAL IN T1
VENDOR 1
F3 CHECK
iDO CHECKDIGIT ON VENDOR NUMBER

0010
0020
0030
0040
0050
0060
0065
0070
0080
0090
0100
0110
0120
0130
0140
0150
0160
0170
0180

EXPLANATION
RANGE check on amount $.25 to $5,000.00 HASH total
on amoun t.
If there is not a part number, output an error.
Do a checkdigit on vendor number. Data input for
keyword PART must be separated by commas.

6.1.3

OUTPUT Section

The OUTPUT section specifies the name and sequence of the data to be
stored on a mass storage device. The OUTPUT section is composed of
OUTPUT blocks, each of the following form
OUTPUT output format# label/default unit
fie1dx

where
OUTPUT

is the heading of the last section of the control
program and must appear even if no output is
desired. Be sure that OUTPUT is not used as a
keyword in the control program.

output format#

is the number (1-4 for 8K systems, 1-7 for
12K
systems)
which links a keyword in the INPUT
section to a particular OUTPUT for mat.
Th i s
number must be in ascending order (beginning with
1) and must appear after OUTPUT.

label

is the name to be assigned to the newly created
data file.
The name can be any unique sequence of
characters (only the first six are significant).

6-8

default unit

is an optional entry specifying the logical unit
(1-15)
where the output file is located.
(If the
number specified exceeds 15, it is taken modulo
16.)
If the file is not found on the default unit
when the control program is executed, a MOUNT
message is displayed.

fieldx

The F, T or C fields in the order
are to be output.

which

The OUTPUT section ends with the specification of another
section (up to 4 for 8K, 7 for 12K) or an END statement.

they
OUTPUT

If another OUTPUT section is specified, its format number must be the
next number in sequence. Different OUTPUT format sections can contain
the same fields so that several data files can be created from one
input. All fields between OUTPUT and END or OUTPUT and another OUTPUT
form one record;
record length must not exceed 510 characters.
OPTIONS
The OUTPUT section of the control program has two
which are specified as follows:

options,

OUTPUT 1 datal

FlO
T50

T75

C
+
+C5

Use C to clear a field after output of the record.
Decimal fields are
cleared to zeros, alpha fields to blanks. During input the contents
of fields not cleared with C options can be repeated
in following
records by skipping fields in subsequent input lines.
Use + as a counter or sequence check. The field to be incremented is
usually a temporary field
(T)
and must be decimal.
This field is
output as part of the record being created. + increments the field by
one;
to increment by any other number specify an increment field
(+fieldxx). The increment field (usually a C field) must be decimal
and must not be larger than the destination field.
If the increment
sum exceeds the destination field, the high order digit is lost.
The
C and + options cannot be used on the same field.

6-9

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110
0120
0130
0140
0150
0160
0170
0180
0190
0200
0210
0220
0230
0240
0250

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
Tl,DlO
Cl,D2,25
C2,D6,500000
INPUT CDR
PART,
Fl ERROR
F2
F5 RANGE (Cl,C2) HASH Tl
DATE
F4
VENDOR 1
F3 CHECK
OUTPUT 1 PRTFIL
:OUTPUT FORMAT 1: FILE NAME, PRTFIL
F3 C
iVENDOR NUMBER
Fl C
iPART NUMBER
F2 C
iPART NAME
F5 C
iWHOLESALE PRICE
F4
iDATE
END
EXPLANATION
Clears fields
written.

6.1.4

with

"C"

each

time

record

Storing the BUILD Control Program

When the BUILD control file is complete, it should be stored on a mass
storage device for future use. Use the command
-!...WRITE BLDSPC
for example, to store the control program.

6-10

6.2

OPERATING PROCEDURES

To load the BUILD program, type
-=-RUN BUILD[, filel, ••• ,file7] [/xx]
in reply to the dot displayed by the Monitor
where
filel, ••• ,file7

are the BUILD control programs.
If the BUILD
control file specified the INPUT device as SYS,
then the rest of the files are names of data files
on the system device to be input.
If no files are
specified, the file in the edit buffer is used to
create a master file.
Multiple files are concatenated and passed to
program as one large file.
are optional switches which affect the
printed by BUILD. The first switch is:

/xx

list the BUILD control program, or

do not list the control program.

the

listing

The second switch is:
/L

list the data records created

do not list the data records

If a second switch is not specified, a data
records list is not produced.
(/N is assumed.) If
the no list condition is
specified,
errors
encountered in the control program or data records
are listed on the line printer as is the summary
of hash totals, records created, etc.
Examples:
..:.. RU BUILD, BLDSPC/NL
lists the data records and summary
program.
~RU

totals

but

not

BUILD,BLDSPC/L

lists the control program and summary of totals, while
.!...

RU BUILD, BLDSPC/LL

outputs both listings and the summary.

6-11

the

control

When BUILD is loaded, the message
MOUNT xxx xxx #01 FOR OUTPUT:
is displayed. xxxxxx is the output file name specified in the control
program and #01 is the sequence number.
Continuing the example, the message is
MOUNT PRTFIL #01 FOR OUTPUT:
Answer with the number (1-15) of the logical device where the output
file
is to be stored. When BUILD is complete, the Monitor DIRECTORY
command can be used to obtain a label of the data file.
The next
steps depend on the device (KBD, CDR, RDR, SYS) specified in the INPUT
statement of the control program.
If the data output exceeds the
storage capacity of the logical
unit, the following message is
displayed if default messages were not specified:
MOUNT xxx xxx #02 FOR OUTPUT:
where xxxxxx is the label of the output file.
Mount the new output
device if necessary.
Type the number of the logical unit where the
rest of the output is to be stored.
BUILD can be aborted at any time by typing CTRL/C, to return to the
Monitor.
All output is lost. The BUILD program must not be run if
the line printer is inoperative.

6.2.1

KBD - Data Input from Keyboard

If the INPUT statement
command with

specifies

KBD,

BUILD

responds

the

RUN

DO YOU PREFER PROMPTING?
Answer YES and a CR key to
automatically by BUILD.
To
of successive input lines
exactly the order in which
the control program with no

have the
input line keywords printed
use prompting, the data input must consist
whose keywords
(if present)
appear
in
they are presented in the INPUT section of
omissions. Monitor displays

after each keyword to indicate it is ready for input.
Answer NO and a CR key to allow each input line keyword to be typed.
BUILD displays a colon (:) when it is ready to accept input.
Input
lines can then be typed in any order and BUILD will match the input
line keyword to the keyword in the INPUT section of the control
program.
If the reply is YES or NO, the dialogue is over and
begin.
BUILD assumes errors will be edited on line.

data

input

can

When all the input lines have been entered, type CTRL/Z to
indicate
the end of data input.
BUILD closes the output files and prints
number of records written, records not written and any hash totals
requested and returns control to the Monitor.

6-12

6.2.2

SYS, CDR or RDR - Data Input from System Device,
Card, or Papertape

If the INPUT statement specifies SYS, CDR or RDR,
The question
DO YOU WISH TO EbIT ON-LINE?
is displayed. Answer NO and the CR key and future errors are printed
on the line printer and all fields in the bad input line will become
undefined.
Records with undefined fields are not written to the
output file.
They remain undefined unless cleared in the output
section or redefined with subsequent input.
Answer YES and the CR key to have an error message displayed on
terminal whenever an error is encountered.
BUILD displays
incorrect input line, the error message and the message

the
the

RETYPE:
Type the correct input line. For details on data
input refer
to
section 6.3.
If just the CR key is typed, the field in error becomes
undefined.
When BUILD reaches the end of the paper tape in the reader, it asks
MORE?
----Answer NO and the CR key if there is no more input.
Otherwise, put
additional paper
tape
in the reader and type Y and the CR key to
continue reading data.
When input is complete, BUILD closes the output data files and prints
the number of records written and records not written and any hash
totals requested and returns control to the Monitor.
When BUILD reaches the end of cards in the reader, the terminal alarm
sounds.
If there are more cards put them in card reader and type any
key to resume operation. Otherwise type CTRL/Z to terminate input.

6.3

BUILD INPUT LINE

The data input line is free form, i.e., specific data fields need not
be in· prespecified column positions.
However, the fields must be
input in the same order within a keyword line as specified
in the
INPUT section.
Data input lines entered at the keyboard mayor may
not contain keywords depending on the answer to the prompting
question.
Data input lines prepared on paper tape or cards must
contain keywords, unless an asterisk was specified in the control
program.
The following conventions must be observed when entering input lines.
1.

The first field in an input line may be preceded by spaces or
tabs if desired.

Each input line must be terminated with the CR key.

6-13

The fields, if more than one is entered per line, must be
separated from each other by spaces, tabs, or commas.
For
example, an input line described in the control program as
DEFINE
FI,D6
F2,A30
F5,D6
INPUT KBD
PART,
FI
F2
F5
can be entered as
PART,126789, HEX BOLT, 500
or if described as:

PART
FI
F2
F5
can be entered as
PART~126789~'HEX BOLT'~500

A field can be omitted by actually omitting it if there are
no fields to the right or by typing two consecutive commas if
there are fields to the right. For example
PART,126789,HEX BOLT
PART" HEX BOLT,lOO
Omitting a field may cause an error depending on the options
(SKIP, default field or ERROR) specified for the field in the
control program.

Decimal data must be entered in one of the following forms:
digits
+digits
-digits
digitsThe plus sign is optional and minus can be placed at the
beginning or end of the data.
Decimal data cannot have
embedded blanks, i.e., 12 34 is illegal.
The number of
digits must be less than or equal to the field size specified
in the ~ontrol program in the corresponding field.
If the number entered is smaller than the field size,
right justified and zeros are inserted on the left.
number is too long, an error results.

6-14

it is
If the

Alphanumeric data must be entered in
forms:
'data'

one

the

following

data

In the first form ('data') any legal COS character (including
spaces and tabs) is permitted except a single quotation mark
(apostrophe). BUILD treats tabs inside quotes as single
spaces, but echoes them as backslashes on the terminals.
In the second form any legal COS character except spaces,
tabs, and commas is permitted. The first character, however,
must not be a single quotation mark.
If the keyword PART did
not have a comma, then input would be
PART~175~WASHER~100

The number of alphanumeric characters must be less than or
equal to the size of the corresponding field specified in the
control program.
If the data is smaller than the field,
it
is left justified and padded with blanks.
If the data is too
large, an error results.
7.

If comma was specified as the field delimiter
in the INPUT
section of the control program, then alphanumeric data not
enclosed in single quotation marks may have embedded spaces
and tabs but no commas and must be separated from the next
field by a comma.
For example, if the control program
specified:
DEFINE
Fl,A30
F2,A30
F3,A30
INPUT KBD
NAME,l
Fl
F2
F3
the input line would be typed as
NAME, JOHANSEN'S MARKET, 2 MAIN ST.,

'MAYNARD, MASS.'

Extra text at the end of a data record is illegal except
comments which must be preceded by a semicolon.

6-15

for

The arrangement of * and keyword records is immaterial at run
time.
BUILD inserts records in the asterisk fields by the
order of input and in keyword records by matching keywords.
For example,
Descriptor

Input Lines

Keywordl .••

Keyword2 •••

*
*

*
*
etc.

Keywordl

Keyword2
There is a current asterisk record pointer in BUILD
keeps track of the next asterisk record to be filled.
inputting the last * record, BUILD goes back to the
one.

6.4

which
After
first

ERROR MESSAGES

BUILD error messages are output to the line printer as the errors
encountered.

are

BUILD checks each section of the control program and
for errors.

data

the

input

If editing is being done on line, the line in error, error message,
and RETYPE message are displayed on the terminal. The keyword is
displayed if prompting is in effect. Otherwise, the keyword must be
retyped.
If no input editing is being done, data input errors are
logged on the line printer in the form
input line

****
error message
Failure to correct a field in error causes the field to become
undefined. The record containing an undefined field is not written on
a storage device.
Error messages followed by an explanation point (1)
indicate fatal
errors.
The error messages for the Field Description section of the
control program are:

6-16

Message

Meaning

BAD DIGIT IN DECIMAL INITIAL VALUE
Alpha character
val ue.

decimal

initial

DATA TABLE OVERFLOW!

Too
much
characters.

data.

Maximum

DESCRIPTOR TABLE OVERFLOW!

Too many F,
T
Maximum = 160.

and

EXTRA CHARS AT STMNT END

Characters not relating
appear with statement.

FIELD NUMBER MISSING OR 0

Field number or default unit number
is
missing,
is 0 or greater than or equal
to 512.

INITIAL ALPHA VALUE DOESN'T BEGIN WITH QUOTE
Beginning quotation
initial alpha value.

fields

defined.

mark

960

statement

missing

for

INITIAL VALUE TOO BIG

The initial value specified
than the field size.

larger

INITIAL VALUE TOO SMALL

The initial value specified
than the field size.

smaller

specified

where

MISSING CLOSE QUOTE ON ALPHA INITIAL VALUE
Quotation mark
necessary.

not

MISSING INITIAL VALUE

Comma was inserted after type and size
but initial value was not specified.

NO COMMA AFTER FIELD NAME

No comma or a character other than comma
was specified after the field name.

NOT A OR D

A character other than A or D occurred
in a data specification statement where
A or D was expected.

NOT BUILD FILE!

File
did
sta temen t.

NOTHING AFTER FIELD NAME

Field type and size are not
after field name and comma.

not

start

with

DEFINE
specified

NUMBER REPEATED OR OUT OF ORDER
A field sequence number
is used more
than once or is out of ascending order
sequence.

6-17

Error messages for the INPUT/OUTPUT sections of
are:

the

control

program

Meaning

Message
BAD END STATEMENT

Missing or misspelled END statement.

BAD DEFAULT FIELD

Default field does not agree
or type of object field.

BAD DEPICTOR

OUTPUT format in error.

BAD DEVICE

Not KBD, SYS, RDR or CDR.

BAD FIELD NUMBER

Field number specified is
2047.

BAD INPUT STATEMENT

Statement missing or misspelled.

BAD OPTION

Something other than C or + is specified
in OUTPUT section.

BAD OUTPUT FORMAT #

Format number is wrong
(lor 7).

BAD RANGE CONSTRUCTION

In RANGE portion of statement an open or
close parenthesis or comma is missing.

C-FIELD NOT PERMITTED

C-type fields cannot be specified in the
INPUT section.

with

size

greater

than

out-of-range

DESTINATION FIELD NOT DECIMAL
The destination field specified
defined as decimal.

not

hash

bigger

than

FIELD NOT DECIMAL

Field not defined as decimal.

HASH FIELD MISSING

No field specified for storage
total.

INCREMENT TOO BIG

Length of increment field
length of output field.

MISSING FIELD NAME

Field name is missing.

OUT OF ROOM!

Too many descriptors in INPUT section.
Maximum
keywords, approximately 121.
Approximate maximum fields
for
all
keywords
192.
Too many fields in
OUTPUT section. Approximate maximum
128.

OUTPUT RECORD TOO BIG

Record exceeds 510 characters.

6-18

Meaning

Message
UNDEFINED FIELD

Field referenced was not defined in
Field
Descriptor section of control
program or a decimal field was being
equated to an alpha field, or an alpha
field was specified for a flag.

UNDEFINED FORMAT

Format number referenced but not defined
(error given on END statement) .

UNEXPECTED END OF FILE

Missing END statement.

Error messages for data input are:
Meaning

Message
BAD ALPHA FIELD

Illegal entry in alpha field.

BAD CHECKDIGIT

Checkdigit calculated by BUILD does
match the one entered.

BAD DECIMAL FIELD

Illegal entry in decimal field.

BAD DELIM

Fields are separated
delimiter, i.e., by a

by an
or 1.

BAD KEYWORD

Keyword entered not
control program.

described

EXTRA CHARS AT END OF DATA

More data was entered than was defined
in the control program, i.e., 4 fields
entered when only 3 were defined.

FIELD OUT OF RANGE

Data entered is not within the
specified in the control program.

range

NECESSARY FIELD MISSING

Field specified as necessary in
program was not entered.

control

TOO BIG

Data entered is greater than the size of
the field as defined
in the control
program.

6-19

not

illegal
in

the

CHAPTER 7
SORT

SORT arranges COS data files in order according to a key specified in
the SORT control program. The key consists of certain fields or parts
of fields which appear in each record of the file being sorted.
The
data records can be sorted in ascending or descending order based on
the key.
Before SORT can be run, a SORT control program must be written.
The
control program defines the fields of the records to be sorted,
specifies labels for input and output files, and the key to be used in
the sort.
SORT handles one file per run and there must be a separate SORT
control program for each file sorted.
Files to be sorted must contain
fixed length records.
If a multi-reel file is to be sorted, the SORT program sorts each reel
separately.
A separate merge pass of the same SORT program must then
be done to combine the file.
To sort a multi-reel DECtape file on
disk without going through a merge phase, PIP the data file onto disk
before sorting.
(See PIP option D.)

7.1

SORT CONTROL PROGRAM

The SORT control program is created and written on a mass storage
device with editing commands.
It consists of a file Descriptor
section and an INPUT/OUTPUT section.

7.1.1

File Descriptor Section

The File Descriptor section has the form:
DEFINE
Fx,typesize

7-1

where:
DEFINE

is the section heading and
statement in the program.

should

the

first

is the first field in the record to be sorted.
All fields in the record to be sorted must be
defined in the File Descriptor section in the
order
they appear in the record. The number (x)
must be a positive, non-zero decimal number not
greater than 511 and should begin at 1 with no
numbers skipped.

typesize

is an A for alphanumeric fields or a D for decimal
fields and an unsigned decimal number in the range
1 - 510.

File Descriptor statements must be entered on consecutive lines and
each line is terminated with the CR key.
Use the editing commands to
create the control program.
NOTE
The BUILD Field Descriptor written previously to
create the data file can be used as the SORT File
Descriptor if the input and output fields are
exactly the same and if it was stored as a
separate program. T and C fields and initial
values are ignored by SORT.
Example:
0010
0020
0030
0040
0050
0060

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6

This example is based on the
prog r am.
(See Chapter 6.)

7.1.2

iPART NUMBER
iPART NAME
iVENDOR NUMBER
;DATE
iWHOLESALE PRICE
file

created

with

the

BUILD

control

INPUT/OUTPUT Section

This section specifies the input and output file labels and the number
of work units available for the SORT.
These work units become part of the
SORT run time.

7-2

MOUNT

message

displayed

during

Format of the INPUT/OUTPUT section of the control program is:
INPUT filnam/d,filnam/d
SORT n/wl,w2,w3
KEY fieldx(n,m)-,fieldx(n,m) .••
OUTPUT filnam/d
END
where
INPUT filnam/d,filnam/d

is the name of the file containing the
records to be sorted.
If no name is
specified, SORTIN is assumed.
/d is
optional and is the unit (1-15) to check
for the named file.
If the control
program is used for a separate merge
operation, the additional filnams are
the files to be merged.

SORT n/wl,w2,w3

is the number (n) of logical units to be
used as scratch areas during the sort.
n is any number from 3 to 7.
There must
be at least three units to execute a
sort.
If the SORT statement is not
present 4 units are assumed.
It is
recommended that the size of the scratch
units be as large as one "reel" of the
input file.
/wl,w2,w3 are
optional
default
work
units to be used
if
available.
Using default units bypasses
the mount work unit message.

KEY fieldx(n,m)-, ••.

fieldx is the field name and number of
the field
to be used as the SORT key.
(n,m) are optional and specify the nth
through mth characters of the field.
If
no characters are specified, the entire
field is used as a SORT key.
- requests
a SORT on that field
in descending
order.
If - is not specified, ascending
order is assumed (+ may also be used to
indicate ascending order).
Up to eight
fields or parts of fields
can
be
specified for the sort key.
There is no
restriction on the size of the sort
field as long as the total size of the
fields which comprise the key is not
larger than 510 characters. The sort is
done left to right on the key using the
COS codes shown in Appendix A.
The
leftmost key is most significant, and
the leftmost character in each key field
is
most
significant
for
sorting
purposes. nand m are unsigned integers
with n less than or equal to m.
Each m
must be less than or equal to the length
of the record.

OUTPUT filnam/d

is the data file name to be given to the
sorted records.
If this statement is
missing SORT assigns the name SRTOUT to

7-3

the output.
For multi-reel files the
names $TMPOO
$TMPnn are used.
/d is
the default unit to check for the output
file.
If the file is not there, the
mount message is displayed.
NOTE
For a multi-reel input file,
SORT sorts the first reel onto
the output file and then asks
for the second reel of input.
For example, when sorting a
3-reel input file, the output
files (which are to be input
to MERGE)
are called $TMPOO,
$TMPOl, and $TMP02.
END

Terminates control program.

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
INPUT PRTFIL
SORT 4
KEY Fl-,F4

0100
0110

OUTPUT PRTFIL
END

7.2

iDATA FILE NAME
;WORK UNITS
;SORT KEYS = PART
iNUMBER IN DESCENDING ORDER
iDATE IN ASCENDING
; ORDER
;SORTED DATA FILE NAME

SORT OPERATING PROCEDURES

SORT is called via the Monitor command:
.RU SORT[, filel, ••• ,file7]
where
filel, •• ,file7 are the SORT control program which can be stored
as more than one file.
If none are specified, the
file in the edit scratch area is used.

7-4

Mount the input file and n-l (where n is the number of logical units
to be used for SORT) of the scratch units.
The Monitor outputs the
message:
MOUNT

filnam #01 FOR INPUT:

where filnam is the input file name specified
in the SORT control
program.
Type the number of the logical unit (1-15) where the data
file is stored. Monitor outputs the messages:
MOUNT $WORK2 #01 FOR OUTPUT:

MOUNT $WORKn #01 FOR OUTPUT:
Type the number (1-15) of the logical unit where all but the last of
the SORT scratch units are located. The Monitor checks the labels of
the files currently on these units.
If a label exists, Monitor asks:
REPLACE xxxxxx inn?
Type YES and the CR key to use the unit as a scratch area and destroy
what 1S currently stored there. Type NO and the CR key to save the
present label and specify another unit for a work area.
When the first sort phase is complete, the input is no longer
Monitor displays the message:

needed.

MOUNT $WORKI #01 FOR OUTPUT:
In a tape SORT the input tape can be dismounted and the last work unit
mounted or the input file can become the last scratch area.
(Monitor
will ask REPLACE?
if this is done;
type YES and the CR key) .
Type the number (1-15) and the CR key of the logical
last work area is located.

unit

where

the

When the main part of the SORT is complete, Monitor displays:
UNIT xx IS FREE
MOUNT filnam #01 FOR OUTPUT:
xx is one of the units previously assigned as a work unit.
SORT no
longer needs the unit and it may be used for the output unit if
desired.
Mount the output file.
Type the logical unit number (1-15) where the
output is to be stored.
SORT reads the sorted records from the work
units and stores them on the specified output unit.
If a default output unit is specified in the OUTPUT line of
control, the MOUNT .•• FOR OUTPUT message may not appear.
NOTE
Sometimes the SORT program does not request that
the last work unit be mounted - it types UNIT xx
IS FREE where xx is the unit where the input file
is mounted, and asks that the output file be
mounted. Do not be alarmed
SORT has decided

7-5

the

SORT

that
the
file
is
sufficiently
small and
well-ordered that the main part of the sort is
unnecessary.

7.3

RUNNING SORT AS PART OF AN UPDATE PROCEDURE

If the SORT control program has a line the first six non-blank
characters of which are UPDATE, then the contents of the INPUT and
OUTPUT parameter lines are ignored and the SORT program uses the label
$UPDOO for both its input and its output file.
(See Chapter 8 for
UPDATE.)
7.4

MERGE OPERATING PROCEDURE

To run SORT as a merge only operation, type:
.RU SORT[, filel,file2, .•• ,file7]/x
where
filel, ••. ,file7

are the control program (possibly stored in two or
more segments).

is one of the following switches:
/A

specifies that names of files to be merged are to
be entered from the terminal in answer to the
INPUT FILE LABELS:' message. The output data file
and default unit name are specified in the OUTPUT
clause of the control program.

specifies that names of files to be merged are
listed in the INPUT line of the SORT control
program. This option bypasses the INPUT FILE
LABELS: message.

specifies that the name of the files to be merged
is listed in the INPUT line of the SORT control
program. Checks for n files of the same name on
n(1-6)
default units as specified.
If the number
of units specified is more than the number of
units shown on the SORT control INPUT line, a
MOUNT message is displayed for those files not on
the INPUT line.

can be used with any of the above
lists
the SORT control program
printer.

When called with the /A option, SORT requests the names
files to be merged with the message:

switches and
on the line
of

the

data

INPUT FILE LABELS:
Enter the 1-6 character data file names separated by commas up to a
maximum of six and default units if desired. At least two names must
be entered or the error message NO INPUT is displayed.

7-6

For example:
PRTFIL/lO,FILEl/3
If default units are not specified, Monitor outputs the message:
MOUNT filnam #01 FOR INPUT:
(where filnam is the data file name)

for each data file specified.

For example:
MOUNT PRTFIL #01 FOR INPUT:
Type the logical unit number (1-15) where the data file is located.
Example:
Sample Control Program called PAYKEY
0100
0110
0120
0130
0140
0150
0160
0170
0180

DEFINE
Fl,A6
F2,D5
F3,Dll
INPUT PAYROL/4, PAYI/2
SORT 3/1,2,3
KEY F2
OUTPUT PAYROL/6
END

To SORT the data file described in the control program PAYKEY,
~RUN

SORT, PAYKEY

The input data file is PAYROL located on logical unit 4.
There are
three sort work units, logical units 1, 2 and 3.
The output data file
is PAYROL and the sorted file is to be put back on unit 6.
The /A option in SORT is used as follows:
~RUN

SORT, PAYKEY /A..)
INPUT FILE LABELS:
PAYROL/4, PAYl/2,PAYO~
MOUNT PAYO #01 FOR INPUT:
10..)

7-7

The input data file names to be merged are:
PAYROL on logical unit 4
PAYl on logical unit 2
PAYO on logical unit 10
The output data file, PAYROL, is put on logical unit 6.
The 1M option in SORT is used as follows:
.RUN SORT,

PAYKEY/M~

The input data file names to be merged are found in the INPUT line
the control program:

PAYROL on logical unit 4
PAYl on logical unit 2
The output data file, PAYROL, is put on logical unit 6.
The In option in SORT is used as follows:
~RUN

SORT, PAYKEY/2'/

The input data files to be merged are:
PAYROL on logical unit 4
PAYROL on logical unit 2
7.5

RULES FOR USING DEFAULT UNITS

If you are using default units in your control programs for SORT,
BUILD, etc., the data files should be mounted on those units. If the
Monitor cannot find the input file on the default unit the message
?

MOUNT xxxxxx #xx FOR INPUT:

is displayed. In the above MOUNT message the?
indicates that
input data file "xxxxxx" was not on the default unit specified.

the

MOUNT PAYROL #01 FOR OUTPUT:

6./
If output of a data file using a default unit is attempted and another
file already exists, the REPLACE message is displayed:
REPLACE LIST #02 ?
Y~

If output of a data file onto a logical unit
used for input or output is attempted,
displayed.
IN USE
MOUNT PAYROL #01 FOR OUTPUT:
10~

7-8

that is already being
the message IN USE is

7.6

ERROR MESSAGES
Message

Explanation

File descriptor errors:
BAD DIGIT IN DECIMAL INITIAL VALUE
Alpha character
value.

decimal

initial

BAD WORK UNIT COUNT

Number of work units not in range 3-7.

EXTRA CHARS AT STMNT END

Characters not relating
appear on statement line.

FIELD NUMBER MISSING OR 0

Field number is missing or
greater than or equal to 512.

ILLEGAL SORT KEY

Bad syntax on KEY statement, KEY
complex, or KEY statement missing.

ILLEGAL UNIT

Default unit is 0 or greater than 15.

INITIAL ALPHA VALUE DOESN'T BEGIN WITH QUOTE
Beginning quotation
initial alpha value.

mark

statement
is

missing

or
too

for

INITIAL VALUE TOO BIG

The initial value specified
than the field size.

larger

INITIAL VALUE TOO SMALL

The initial value specified
than the field size.

smaller

specified

where

MISSING CLOSE QUOTE ON ALPHA INITIAL VALUE
Quotation mark
necessary.

not

MISSING INITIAL VALUE

Comma was inserted after type and size
but initial value was not specified.

NO COMMA AFTER FIELD NAME

No comma or a character other than comma
was specified after the field name.

NOT A OR D

A character other than A or D occurred
in a data specification statement where
A or D was expected.

NOTHING AFTER FIELD NAME

Field type and size are not
after field name and comma.

specified

NUMBER REPEATED OR OUT OF ORDER
A field sequence number is used more
than once or is out of ascending order
sequence.
TOO MANY FILES

Merge only,
specified.

7-9

than

input

files

Explanation

Message
UNRECOGNIZABLE LINE

Parameter line did not start with a good
keyword.

SORT Errors:
BAD RECORD SIZE

File contains
length.

NO INPUT

Input file is null or not enough
files specified for a merge.

OUTPUT ERROR

'Indicates a system malfunction.

UNIT xx IS FREE

NOT AN ERROR - merely an informative
message to aid the operator. xx is a
COS logical unit number.

7-10

records

variable
input

CHAPTER 8
UPDATE

COS UPDATE is a file maintenance program
insertions and deletions to data files.

which

makes

changes,

UPDATE must also have a control program before it can be executed.
UPDATE operates in two passes, passl reads the control program and
UPDATE commands and checks for errors.
It then outputs a modified
version of these commands to a scratch file for
the user
to sort.
Pass2 uses the UPDATE commands read from the scratch file to modify
the input file, and creates a new output file.
The UPDATE commands are listed on the line printer as they are
executed.
A summary of the number of records read, changed, inserted
and deleted is printed when the UPDATE is complete.

8.1

UPDATE CONTROL PROGRAM

The control program consists of a file dscriptor section followed
UPDATE, INPUT, SORT, KEY, OUTPUT and END statements in that order.

8.1.1

File Descriptor Section

This section defines the fields as they appear
in the file to be
updated.
The restrictions are the same as for the SORT program. T
and C field names are accepted but are ignored by UPDATE.
Initial
values are retained unless specifically changed by an UPDATE command
(see section 8.2). Be sure not to skip any numbers when assigning
field names.
The file descriptor section must start with a DEFINE
statement.
Example:
0010
0020
0030
0040
0050
0060

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6

iPART NUMBER
iPART NAME
iVENDOR NUMBER
iDATE
iWHOLESALE PRICE

8-1

This file descriptor section is based on the file created
in Chapter 6.

8.1.2

with

BUILD

UPDATE Statement

The form of this statement is:
UPDATE filnam[/unit]
where
filnam

is the 1-6 character name of the data file to be
updated.
This name may not contain spaces, tabs,
or commas.

/unit

is optional and specifies the logical unit (1-15)
which contains the file.
If the number specified
exceeds 15, it is taken modulo 16.
If the unit is
omitted or if the file is not on the unit
specified, Monitor outputs
the
usual
MOUNT
message.

Example:
0010
0020
0030
0040
0050
0060
0070

8.1.3

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3

;DATA FILE TO BE UPDATED IS
iPRTFIL LOCATED ON LOGICAL
iUNIT 3.

INPUT Statement

The form of this statement is:
INPUT dev
where
dev

is the physical device (KBD, RDR, CDR, SYS) to
used to input the UPDATE commands
KBD RDR CDR SYS -

terminal keyboard
paper tape reader
card reader
system device file

8-2

Example:
DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3
INPUT SYS

0010
0020
0030
0040
0050
0060
0070
0080

8.1.4

jUPDATE COMMANDS
jARE STORED
jIN A FILE ON THE
jSYSTEM DEVICE.

SORT Statement

The form of this statement is:
SORT n
where ri

is the number (3-7) of logical scratch units which
are to be used if the updated records must be
SORTed. This is an optional statement.
If the
UPDATE commands are sorted, no SORT statement is
needed. There must be at least three units to
execute
a
SORT.
If no scratch units are
specified, 4 is assumed. If no SORT statement is
specified, UPDATE executes a combined passl and
pass2, reading one UPDATE command at a time and
performing the update on the input file directly.
No scratch unit is needed in this case.

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090

8.1.5

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3
INPUT SYS
SORT 4

jTHERE ARE FOUR SCRATCH
jUNITS FOR THE SORT

KEY Statement

The form of this statement is:
KEY fieldx(n,m)-[, ••• ,fieldx(n,m)-]
where
fieldx(n,m)-, •••

is the SORT key used to sort the data file.
(See
SORT control program.)
There can be up to eight
fields or parts of fields in the SORT key.

8-3

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100

8.1.6

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3
INPUT SYS
SORT 4
KEY Fl-,F4

iTHE KEY FIELDS
iARE PART NUMBER IN
iDESCENDING ORDER
iAND DATE
iIN ASCENDING ORDER.

OUTPUr Statement

The form of the statement is:
OUTPUT filnam[/unit]
where
filnam

is the 1-6 character name to be assigned to the
updated file.
If the statement is omitted the
output file has the same name as the input file.

/unit

is the logical unit (1-15) where the output will
be stored.
If no unit is specified, a MOUNT
message is displayed.

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110

8.1.7

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3
INPUT SYS
SORT 4
KEY Fl-,F4
OUTPUT PRTFIL/4

iCALL UPDATED FILE PRTFIL
iAND STORE ON UNIT 4

END Statement

The form of the statement is:
END
END is the terminating statement of the
must be present.

8-4

UPDATE

control

program

and

Example:
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110
0120

8.2

DEFINE
Fl,D6
F2,A30
F3,D7
F4,D6
F5,D6
UPDATE PRTFIL/3
INPUT SYS
SORT 4
KEY Fl-,F4
OUTPUT PRTFIL/4
END

iTERMINATING STATEMENT OF
iCONTROL PROGRAM
.

UPDATE COMMANDS

The UPDATE commands can be input from the keyboard at UPDATE run time
or can be prepared previously on paper tape, cards or the system
device.
UPDATE commands have the general form:

{g}. .

ke y [Fn=valuel, •••

where
I,D,C

denote Insert, Delete, and Change

key

represents the data in the fields used to sort the
record.
The character string typed must be
exactly the same length as specified for the SORT
fields.
The string is continuous; no spaces,
etc. can separate fields if more than one is in
the sort key.
If the data file is sorted on an
alpha key (A12) all trailing spaces must be typed
on an UPDATE command line, for example,
C...... JONES JOHN ..................F6= •••

is a field name and sequence number as specified
in
the
UPDATE File Descriptor.
Fields are
processed from left to right and Field numbers may
appear in any order.

=value

is the data to be inserted in the field named
(Fn) ~
This data must agree in type with the type
specified for Fn and must be less than or equal in
size.
If the value specified is smaller than the
field size, the alpha field data is left justified
and padded with blanks;
decimal field data is
right-justified and padded with zeros.
On an insert, missing fields are set to their
initial value if one wa~ specifie~, otherwise the.
field
is set to blanks
(alpha)
or
zeroes
(dec imal) .

8--5

Decimal values have the format

[:]ai9itS[-l
Alpha values have the format
'characters'
No single quotation marks may appear within the
alpha value.
If no character is a space, tab or
comma then the surrounding quotation marks need
not be present.
A command which is too long for one UPDATE command line can be
continued on the next line. Type a comma at the end of each line to
be continued. There can be any number of continue lines.
Example:
Iu998877377l76l Fl=001122,.)
F2='SCREW,#11' ,F3=100,F4=777l777,F5=32072
An error in an UPDATE command line causes the lirie to be ignored and
an error message is output to the line printer.
Execution continues.

8.3

UPDATE EXAMPLE

A file called ACCT has a mistake in the zip code of an employee whose
employee number is 63411. The following control file is prepared:
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110

DEFINE
Fl,A40
F2,D5
F3,A30
F4,A5
F5,D8
UPDATE ACCT
INPUT KBD
KEY F2
OUTPUT ACCT
END

iNAME
iEMPLOYEE NUMBER
iSTREET ADDRESS
:ZIP CODE
; SALARY

This file describes the fields of the file ACCT, specifies that it is
sorted on field2
(employee number) and says that input commands are
from the keyboard. No SORT statement is included because only one
update command is needed and so no sorting is necessary.
If more than
one change is to be made in the file and if the changes are typed with
their keys not in order then a SORT statement would be necessary.
The user then types:
.RUN UPDATE/IL
to run passl and get a listing of the control file.
If no SORT is required, UPDATE asks that the input and output files
(both named ACCT) be mounted.
If a SORT statement had appeared, the
mount message would have requested the scratch tape $UPDOO.
(ACCT

8-6

would not be mounted until pass2).
be typed at the keyboard:

The single UPDATE command can then

C 63411 F4=1169l
specifying the new zip code of 11691 for
the record whose key is
63411.
All other fields remain the same. The user then types CTRL/Z
to specify the end of the input from the keyboard.

8.4

OPERATING PROCEDURES

UPDATE is called with the Monitor command:
~RUN

UPDATE[,filel, ••. ,file7] [/s(1)s(2)]

where
filel, ••• file7

are the names of the files which contain the
UPDATE control program (possibly stored in two
segments, FDR and I/O) and the UPDATE commands (if
previously stored on the system device) •
If no files are specified, UPDATE uses the file in
the edit buffer.

/s(1)s(2)

are switches
The first switch is 1 or 2 specifying
phase.
1
read control program
commands and check for errors, put
area.
2
read scratch area after
perform UPDATE.
If neither 1 nor 2 is
passl is assumed.

the UPDATE
and UPDATE
in scratch
sorting and
specified,

The second switch is
/N

for no listing (default)

for a listing of UPDATE commands, (pass2) or
control program (passl).

If no errors are found in the control program and the UPDATE commands
are sorted, UPDATE continues automatically to pass2. Otherwise, SORT
must be used to sort the UPDATE commands into the same order as the
file.
Type RU SORT,UPCON,UPCOM (control program). SORT asks that
$UPDOO be mounted. When the SORT is complete, call the second phase
of UPDATE.
For example,
.RU UPDATE,UPCON,UPCOM/2
NOTE
If the input device specified was KBD, there is no
indication from UPDATE that it is ready to accept
input from the keyboard. To terminate input from
the keyboard, type CTRL/Z.

8-7

8.5

ERROR MESSAGES

indicates a fatal error.
Message
BAD DEVICE

Explanation
Illegal device
statement

specified

BAD DIGIT IN DECIMAL INITIAL VALUE
Alpha character
val ue.

input

decimal

initial

BAD INPUT STMNT

Input" statement is
order.

missing

BAD KEY STMNT

Key in KEY statement has bad syntax.

BAD UPDATE COMMAND

Bad syntax, probably missing
comma.

BAD UPDATE STMNT

Update statement is incorrect.

BAD VALUE

Value in update
incorrect.

DATA TABLE OVERFLOW!

Too much data, maximum = 960 characters.

DESCRIPTOR TABLE OVERFLOW!

Too many F,
T
maximum = 160.

EXTRA CHARS AT STMNT END

Characters other than A or D occurred in
a data specification statement where A
or D was expected.

FIELD NUMBER MISSING OR 0

Field number or default unit
missing or is 0 or > = 512.

I RECORD ALREADY EXISTS

Tried to
present.

command

and

insert

INITIAL ALPHA VALUE DOESN'T BEGIN WITH QUOTE
Beginning quotation
initial alpha value.

too

fields

long

defined,

number

record

mark

out

al ready

missing

for

INITIAL VALUE TOO BIG

The initial value specified
than the field size.

larger

INITIAL VALUE TOO SMALL

The initial value specified
than the field size.

smaller

specified

where

MISSING CLOSE QUOTE ON ALPHA INITIAL VALUE
Quotation mark
necessary.
MISSING INITIAL VALUE

not

Comma was inserted after type and size
but initial value was not specified.

8-8

Explanation

Message
MISSING OR OUT OF ORDER

On C or D, record was
Probably because scratch
sorted.

not
there.
file was not

MISSING UNIT

Unit number specified is a
16.

multiple

NO COMMA AFTER FIELD NAME

No comma or a character other than comma
was specified after the field name.

NO DEFINE!

Control
DEFINE.

NO END STATEMENT

END statement
specification.

after

OUTPUT

NONEXISTENT FIELD

Tried to update Fn where no Fn
in file description section.

appeared

NOT A OR D

A character other than A or D occurred
in a data specification statement where
A or D was expected.

NOT I,D, OR C

Bad first character on update command.

NOTHING AFTER FIELD NAME

Field type and size are not
after field name and comma.

file did not start with
a
Returns to Monitor immediately.
missing

specified

NUMBER REPEATED OR OUT OF ORDER
A field sequence number
is used more
than once or is out of ascending order
sequence.
OUT OF ORDER

SORT statement
missing
commands are not in order.

UNEXPECTED END OF FILE

Missing END statement.

8-9

and

update

CHAPTER 9
BOOT
BOOT is used to bootstrap the system from one device to another, i.e.,
if the system has been moved from DECtape to disk, BOOT may be run to
start the system on disk.

9.1

OPERATING PROCEDURES

The command to run the BOOT program has the form:
.:..RUN BOOT/xx
and the CR key:
where
is the switch which indicates the device where the
system to be bootstrapped resides.

/xx

(COS-300 only)

/RK

is the RK08 or RK8E disk unit O.

/RF

is the RF08 disk.

/DT

is the TC08 DEC tape unit O.

(COS-300 only)

/TD

is the TD8E DECtape unit

(COS-300 only)

/LT

is the LINCtape unit

/RX

is the RX8E diskette unit

(COS-300 only)

a (COS-310 only)

If no switch or an illegal switch is specified, BOOT displays the
error message NO and control returns to the Monit6r.
Do not try to bootstrap onto a device which is not ready or does not
exist.
Other options are available for experienced programmers:
/PT

loads the high-speed RIM loader and BIN paper tape
loader into field O.

/ZE

zeroes all of core in field O.

BOOT may also be used to bootstrap OS/8

9-1

CHAPTER 10
PATCH

PATCH is used to fix (or patch) system programs or the Monitor on a
COS system tape.
The data to make the changes is provided by Digital
Equipment Corporation as patches in the form of dialogue with the
PATCH program with the user input underlined.
It would be necessary,
therefore, to type only the underlined data exactly as shown and to be
sure that the computer output corresponds with what is illustrated.
If everything matches,
the patch is correctly made.
To better
understand
the workings of PATCH, it is important to note that both
system programs (files) and the Monitor consist of blocks of numerical
information
(program code;
machine language
instructions) on the
system device.
Each block consists of words of 4-digit numbers.
These words are numbered from 0 to 377 (base 8).
PATCH reads one of
these blocks, allows user examination and/or change
(patch)
of
individual words in this block and writes the new block back out to
the system device.

10.1

OPERATING PROCEDURES

Be sure the system tape has the current version of the program
patched.
The command to run the PATCH program is of the form:

.RU PATCH
and the CR key.
PATCH replies:
COS PATCH SYSTEM VERSION 5.05 (or current version number)
and
FILENAME:
Respond with one of the following:
(a)

The name of the file to be patched.

(b)

/N and/or a carriage return, to indicate a patch for the
Monitor not for a program.
The system responds PATCHING
MONITOR.

10-1

(c)

CTRL/Z or IX.
To indicate end of PATCH operation.
The
message EXIT is printed and control returns to the
Monitor (just as if CTRL/C had been typed.)

When typing in letters or digits, a typing mistake can be corrected
(before typing the CR key) by typing CTRL/U and retyping the entire
line correctly.
EXIT Routine
The next question asked by PATCH is:
LOCATION:
If no further patching in block is required, type:
END
PATCH now asks:
BLOCK:
If no further blocks are to be patched, type:
END
PATCH now asks:
FILENAME:
If no further patching is required type control/Z

tz
and exit to Monitor
NOTE
If it is discovered at anytime before
exit
routine, that the wrong block
number was typed, type R (for Restart)
and
PATCH
returns to the FILENAME
Question. Reenter the file name, then
type
the
correct BLOCK number and
continue from there.
The next PATCH question is:
BLOCK:
Type the block number specified on the patch information sheet as sent
out by DEC.
This specifies which BLOCK is to be patched.
PATCH displays:
LOCATION:
Type in the number provided by DEC.
This specifies which word in
block is to be patched. PATCH responds with:
OLD VALUE:

nnnn

where nnnn is the four-digit old (current) value of this location.

10-2

the

It is very important to verify that this value is
said it should be.
If it is not, do not proceed;

the one that DEC
something is wrong.

Check that everything previously typed was letter perfect.
If the wrong BLOCK number was typed, type R and restart at
FILE NAME.
If the wrong LOCATION was typed, type CR key in
answer to the NEW VALUE: question. This makes no change to
the location specified.
Then type the correct location
number in answer ~o the LOCATION question.

If everything typed was correct, perhaps an old version of
the Monitor or the system program in question is being used.
Before starting PATCHING, the version of the program in
question should be verified.

If everything seems correct but the dialogue doesn't agree
with what DEC predicted, save all output and consult your
software specialist.

When the program asks:
NEW VALUE:
enter the new value as supplied by DEC. This new value replaces the
old value in core.
The block is not changed on the system device
until all patches are made to this block.
If the wrong new value is typed, retype the location, verify that its
current (old) value is the one typed in by mistake, and then enter its
new corrected value.
If a new value is entered for the wrong location, reenter that
location and as the new value, type its original value (determined by
looking back on the listing for the OLD VALUE at the time this
location was first referenced).
If a scope display is used and the
original value cannot be determined, type R and redo the entire block.
Answer NEW VALUE: with the CR key or CTRL/Z if the location was
examined and no new value is to be entered.

just

When all patches to a block are complete, type CTRL/Z or END to the
LOCATION:
question.
At this point, PATCH asks for the RELATIVE
CHECKSUM (similar to a hash total). Type the relative checksum, as
supplied to you by DEC.
and the CR key. If everything is correct,
the new corrected block is written on the system device, and the
message:
NEW BLOCK PATCHED OK
is displayed. The question BLOCK:
is displayed again to allow
patching of another block. Type in the next block number in this file
to be patched or type END or CTRL/Z to terminate patches for the file.
In the latter case, PATCH displays the message:
nn BLOCK(S) PATCHED IN THIS FILE
FILE NAME:
If an error is detected, the message:
BAD CHECKSUM
LOCATION:

10-3

is displayed and the faulty block is not written to the system device.
The new block as changed is still in core. Review the numbers typed
in and possibly examine the locations changed to see if they are
correct.
If the error is found, fix it and then type END or CTRL/Z to
the LOCATION: message.
If an error is not found, type R to restart
the program and try patching the entire block again •

•

10-4

10.2

ERROR MESSAGES
Message

Explanation

BAD CHECKSUM

An attempt was made to write
which was incorrectly patched.

BAD DIRECTORY

Logical unit 0 was not a system device.

BAD NUMBER

A number with more than 4 digits, a
non-digit or containing the digits 8 or
9 (these are never used) was typed.

BLOCK TOO BIG

An incorrect block number was typed.
It
cannot be larger than the length of the
file being patched.

FILE NOT FOUND

The program with the name specified was
not found on the system device.
It must
be there.
If it is on another tape,
it
may be mounted on logical unit 0 when
PATCH first starts up;
but thereafter
DO NOT CHANGE UNITS.

LOCATION TOO BIG

A location greater than 377 was typed.

NO CHANGE IN BLOCK

An attempt was made to write a block but
no changes were made in it.
If this was
on purpose, fine, otherwise make the
changes to the block again.

10-5

block

CHAPTER 11
DAFT

The DAFT (Dump and Fix Technique) program is similar in function to an
editor but is used for data records. DAFT provides the capability to
search for, examine and change records and to list records or parts of
records on the line printer or terminal. DAFT may be used to make
minor changes to a data file.
DAFT allows one input and one output file to be open at the same time.
These two files may be the same file when in update mode. At any
given time, there is a record from the input file in core known as the
current record.
This record may be modified by the CHANGE command
before it is written on the output file. An output file need not be
open if records from the input file are only being examined.
11.1

OPERATING PROCEDURE

The command to run DAFT has the form:
~RU

DAFT[, filel, ••• ,filen]

and the CR key.
Where
filel, ..• ,filen

11.2

are previously created files which contain a
string of DAFT commands to be used to dump or fix
the data file. These files are optional and if
not
specified, commands are entered via the
terminal. After the last command in the last file
is executed, additional commands may be entered
from the keyboard.

COMMANDS

DAFT commands begin with a keyword (consisting of any number of
non-blank characters, of which only the first one is significant).
Some commands take arguments, in which case the keyword is followed by
one or more spaces and the arguments.
These arguments may be
separated from each other by one or more spaces.
In Table 11-1, an expression in brackets
optional.

11-1

[]

indicates

that

The following symbols are used in Table 11-1 to show the form of
commands:

DAFT

represents an unsigned non-zero positive integer.
If it is optional in a command and is omitted, n=l
is assumed.

<a,b>

represents a key field consisting of character
positions a through b inclusive. a and b must
each be unsigned non-zero positive integers and b
must not be smaller than a. If this is optional
in a command, and is omitted, the key specified in
the KEY command is used.

label

is a data file name.

unit

represents an integer in the range 1-15. If this
is specified, the system looks for the specified
file on this logical unit before issuing a MOUNT
message.
If given, /unit must immediately follow
the file name.

indicates that before a record is read from the
input file, the current record in core (if there
is one) is written on the output file. The + sign
need not have a space before it unless it is the
only argument.

data

represents a piece of data.
form:

Alpha

data

has

the

, characters ••. '
and decimal data has the form:
[-] digits .••
Before data is used in executing the command, it
is adjusted to the same length as the key field.
If the data is smaller and alpha, it is left
justified in the field and padded with spaces on
the right. If it is smaller and decimal, it is
right justified and padded with zeroes on the
left.
If data is larger and alpha,
excess
characters on the right are ignored. If it is
larger and decimal, excess characters on the left
are ignored.
Table 11-1
DAFT Command Summary
Command
Advance [ n]

Function
[+]

Backspace [ n]

Advances the input file n records.
is omitted, 1 is assumed.

If n

Backspaces n records if the input file
was opened with the UPDATE command. If
n is omi tted, 1 is assumed.

11-2

Table 11-1
DAFT Command Summary (Con't)
Command

Function

Change [<a,b>] data

Replaces the data in the current key
field of the record currently in core
with the data specified.
If the key,
<a,b>,
is used in this command,
it
temporarily overrides the key specified
by the KEY statement.

Display en]

Sets the width of the print line of the
listing
device
(terminal
or
line
printer) to n characters (maximum 130).
If n is omitted, this command turns on
the grid if it was off and turns it off
if it was on.

Exit

Returns control to the COS
Monitor
providing no output file is open.

Fini [+]

Closes the output file.
If
+
is
specified, the current record and the
remainder of the input file are first
copied to the output file.

Goto n [+]

Makes record n the current record.

Help

Displays a summary of DAFT
the terminal.

Input label [/uni.t]

Opens the specified file
(label)
for
input.
The first record is read and
becomes the current record.

Key a,b

Sets the current key
to
character
positions a through b inclusive.

List en]

[<a,b>]

[+]

commands

Lists on the lineprinter, n consecutive
records
beginning
with the current
record. The key, <a ,b>,
if spec ified,
represents those consecutive characters
in the record that are to be considered.

Output label [/unit]

Opens the
output.

Put en]

Writes n copies of the record
in core on the output file.

Query

Lists information about the
current
status of the program on the terminal.
This includes the names of the input and
output files, the units where the files
are located, the record in core and the
version number of the DAFT program.

Rewind

Rewinds the input file.
becomes the current record.

11-3

specified

file

(label)

for

currently

Record

Table 11-1
DAFT Command Summary (Con't)
Command

Function

Search [<a,b>] data [+]

Searches the current record and then
succeeding records for an occurrence of
the specified data appearing in the
current key field.

Type [ n] [< a ,b >] [+]

Same as list except
terminal.

Update label [/unit]

Opens the specified file (label)
for
updating.
This command can only be
specified for a file with fixed length
records since direct access I/O is used
to move records.

Version

Displays the version number
program on the terminal.

Wr i te [n]

Performs same function as ADVANCE [n] +,
i.e., n-l consecutive records, beginning
with the current record are written on
the output file. The nth record (after
the original one) becomes the current
record.

Lists the record number and size of the
current record on the listing device
(either
terminal
or
line
printer
depending on whether the last record was
listed via a TYPE or LIST command). The
line printer is the initial listing
device.

11.3

output

the

DAFT ERROR MESSAGES
Message

Explanation

o NOT ALLOWED

In some commands, 0 is not a permissable
argument.

BAD DIGIT IN DATA

In a CHANGE or SEARCH
command,
a
character other than digits or a minus
sign is contained in a numeric data
field (a field without quotes) •

CANT BACKSPACE PAST BEGIN OF FILE
Attempted to backspace past beginning of
file.
The first record becomes the
current record.
CANT BACKSPACE WITH SEQUENTIAL INPUT
Attempted to backspace
input.

11-4

with

sequential

Message

Explanation

END OF INPUT FILE AT RECORD nnnn
An attempt was made to read past the end
of file mark on the input file.
This is
not necessarily an error. nnnn was the
last record read.
The input file is
closed.
EXCESSIVE GRID SIZE

The grid
(printer width)
may
greater than 130 characters.

EXTRA CHARS

Extra characters were found
logical end of a command.

not

after

be
the

ILLEGAL RECORD - CLOSING FILE The file being updated contains a bad
record
(one not the same size as record
1) •
Only fixed length records
are
permitted on such files. The file is
closed.
KEY ENTIRELY PAST END OF RECORD
The key specified in a LIST or TYPE
command began with a character greater
than the record size.
KEY EXTENDS PAST RECORD END
A change or search was attempted with a
key that extends past the end of a
record. However, a list with such a key
is possible.
In such a case the list is
terminated at end of record.
KEY TOO BIG

The range of the key cannot

exceed

100

characters~

NO DATA

Data must be specified in
SEARCH command.

CHANGE

NO INPUT FILE

The command requires an input
one is not open.

file

but

NO LABEL NAME

The filnam was omitted in
OUTPUT or UPDATE command.

NO OUTPUT FILE

The command requires an output file but
one
is
not open.
The command is
terminated at the point just prior to
writing the current record on the output
file.

OUTPUT FILE ALREADY OPEN

A request was made to open an output
file while one was already open. Only
one output file may be open at a time.
The request is ignored.

11-5

INPUT,

Message

Explanation

OUTPUT FILE STILL OPEN

An EXIT cannot be made when the output
file is still open. The output file can
be closed with the FINI command or
CTRL/C may be typed.

PUSH DOWN OVERFLOW

The program will abort with this message
when too many errors are made.

11.4

DAFT OUTPUT

Records may be listed with a grid above them. This grid consists of
two lines of numbers showing the character positions. The lower of
the two lines represents the units digits of the column counts.
The
upper line represents the tens digits. The tens digits are printed
for the first and last character in the record (or part of the record)
listed or whenever
it changes. If there is a hundreds digit, it is
printed in column 1 or whenever it changes.
Refer to Figure 11-1 for a sample of a DAFT operation.
COS MO~ITOR

2.0606

• R DAFT

* 'i ELP
A DVA'\]CE \1+
BAC:{SP~CE

C:..rA\1GE <A, P> DATA
DI SPLA'l \1
EXIT
FI\11 +
G OTO ~+
H ELF
I \JPUT LAREL/TNIl
I{ E'l A, B
LIST \J :{E'l+
OUTPUT LABELl U\JI T
P tTl \J
QUER'l
R E\d \JD
5 E'~F:C~ < A, P> DATA +
TYPE '1 < A, P>+
UPDATE LAEFL ItNI 1
V FRSI 0\1
WBI TE \J

* VEHSI 0'1
D~FT

VERSIO~

6.06

DAFT VERSIO\1 6.06

* VERt

DAFT VERSIO\1 6.06

* I\1PUT MAIL\JG
MOU'\lT MAIL\lG #01

FOR

I\JPUT:

* T¥PE
Figure 11-1
DAFT Sample

11-6

RFCORD 000001 OF FILE :-tAIL'JG,

RFCORD L E'JGli= 204 CHABAC1E&S

DI GI TAL EQUI PM£1'J T CORP.

D. F.

12-3

146

MAY'JARD

P~v'LOC:{

~AI ~

ST.

0012345Al1111

MA011540S/~-1

06211

2003SUPERvrSOR

* 01 SPLAY 10\

*T 1

RECORD 000001 OF FILE MAIL~G,
01 GI TAL EQUI PM EN T CORP.

RECORD LE"JGT~=204 CHARACTERS

146 MAI~ ST.
111

12- 3

D. F. PAVLOCK
MA'f'JARD

MA011540S/3-1

0012345Ali

062112003SUPERVISOR

*0
T 2

RECORD 000001 OF FILE "1AIL'JG,

RECORD LE:-.JG1~=204 C-IAF.ACTEFIS

2 3 4

1234561~901234561~901234561q901234567~901234561~9012345613~01234561390

Dr GI TAL EQUIPME"JT CORP.
1

D. F. PAVLOCK

12-3
1

1234561q901234567~901234561~9012345673J01234561q)012345613901234567~90

146 ~AI'J ST.
14

MAi'JARD
6

0012345A11

~A011540S/-3-1

12345613901234561q90123456139012345613901234561~901234561890123

111

062112003SUPERVISOR

RECORD 000002 OF FILE MAIL'JG,

RECORD LE"'1GT-I=204 C'1ARACTERS
3 4 5

1234561q9012345678901234561~90123456139012345613901234 561B901234561~90

DIGITAL EQUIPME'JT CORP.
1

RIC-IER

12-3
2

12345618901234561g901234561g9012345673901234567~~01234561g)01234561890

146 MAI'J ST.
14

MA'f'JARD
6

MA01154COS 300 0001912T 3
3

123456189012345618901234561~90123456189012345618901234561390123

053112002WRITER

* T 2< 25, 50>

RECORD 000002 OF FILE MAILNG,

RECORD L E'JG l;.{= 204

234
5
56189012345613901234561890
K. RI C'-tER

Figure 11-1
DAFT Sample (Cont.)

11-7

C~ARACTER S

RECORD LE~GT....f=204 CHABACTERS

RECORD 000003 OF FILE MAILI\1G,

56789012345673901234567390
S. RABI lIJOWI TZ

*T 2
RECORD 000003 OF FILE MAIL1IJG,
DI GI TAL EQUI PMENT CORP.

RECORD LENGT-i=204
RAEI NOWI TZ

12- 3

MAYNARD

MA01754COS 300

146 MAIN ST.
345

CiARACTERS

OO)99~9CI2

062172009COMPILER

RECORD 000004 OF F1 L E MAl L~G,

RECORD LENGTri=204 CHAfiACTERS

R. LABY

01 G1 TAL

12- 3

MAYNARD

146 MAIN ST

MA01754

061272999 SORT

* KEY 1, 50
* T2

; SET

PERMA~ENT

RECORD 000004 OF FILE

:{EY

MAIL~G,

RECORD LE~GT-{=204 C>-iARACTERS

1
2 3 4
5
123T56789012345673901R3456739012345678901234567890
DI GI TAL
R. LARf
*T 2

RECORD 000004 OF FILE MAIL'JG,

RECORD LENGT"i=204 Ci-iARACTERS

1
2 3 4
5
12345678901234567890123456789012345673901234567890
DI GI TAL
R. LAR,(

RECORD 000005 OF FI L E MAl L'JG,

RECORD L E~GTH=204

CHARACTERS

1
2
3
LI
5
12345673901234567890123456789012345678901234567890
DEC
S. G. WELLCOME

*D
*T

Figure 11-1
DRAFT Sample (Cont.)

11-8

0000000

RECORD 000005 OF FILE MAIL"JG ..
DEC
*Q
I'1PUT FILE: MAIL"JG
UNI T: 00
OUTPUT FI L E: /\1 0\1 E/
UNI T: 00
K EY=< 00 1.. 050>

S. G.

RECOflD LENGTii=204 CHARACTERS
WELL COME

OPEN

RECORD 000005 OF FILE MAIL"JG ..

RECORD LE'VGT-{=203 CiARACTERS

DAFT VERSIO'1 6.06

*K
*Q
I'1PUT FILE: i'wfAIL\1G
U\1I1: 00
OUTPUT FILE: /"JO\1E/
UNI T: 00
K E'{=<OOI .. 510>

OPE\}

RECORD 00000 5 OF FI L E MAl L \JG ..

RECORD LE\JGTi=2C4 CiARAC1ERS

DAFT VERSION 6.06
; SHOULD CAUSE A\1 EREOR MESSAGE
CA"JT BACKSPACE WI~ SEQU~TIAL I\JPUT

Figure 11-1
DRAFT Sample (Cont.)

11-9

CHAPTER 12
COS-OS/8 CONVERTER
The COS-OS/8 Converter program, CONVEX, provides the capability to
convert source or data ASCII files created with OS/8 to COS source or
data files.
COS data files can be converted to OS/8 ASCII files.

12.1

OPERATING PROCEDURES

The command to run CONVEX has the form:
..:..RUN CONVEX
and the CR key:
CONVEX replies with the program version number:
VERSION 5.05

(or current version number)

And the message:
INPUT MODEAnswer wi th the lett.er code:
A
D

(OS/8 ASCII file mode)
(COS Data file mode)

and the CR key.
It then displays the message:
FILE NAMEThe answer has two forms depending on the mode.
1.

For mode A (OS/8 ASCII) type the OS/8 file name (one to six
characters optionally followed by a period and up to a two
character extension) followed by a comma and then the device
name which is one of:
DTn
DKn
where n is a single digit referring to the physical unit #
(DKO-DK3, DTO-DT7, or RXO-RX5).
If no device is specified,
CONVEX uses the system device.

12-1

Examples:
ABC,DTS
RICH.TM,DKO
2.

For mode D (COS data) type the COS data file name (one to six
characters)
optionally followed by a slash and the logical
unit number (I-IS).
If a logical unit is specified, the
system first checks that logical unit for the specified file
before issuing a MOUNT message.
Examples:
CUSTIS
LISTII
PROSPT
Illegal:
CLIENT,DTI

If everything is ok, the program displays the message:
OUTPUT MODE Type one of the following one character codes:
A
D
S

(OS/8 ASCII file)
(COS Data file)
(COS Source files)

The next question is:
FILE NAME If the option
explained in
CONVEX asks:

specified was
steps I and 2.

A or D,
reply following
the rules
If an existing file is to be destroyed,

REPLACE?
Type Y for YES, N or any other character for NO and the CR key.
If the option specified was S, type the filename, optional
COS device e.g. GROUPA,DT3.
This device must be physically the same as the
tape to tape or disk to disk.

system

comma

and

device,

i.e.,

specified.

Type

When the conversion is complete the message:
INPUT MODE is displayed again and another
CTRL/C to return to Monitor.

conversion

12-2

can

NOTE
If instead the message:
OUTPUT MODE is displayed again after successful output has
occurred
(MODE A), CONVEX needs more space to
continue. Specify an additional output file.
NOTE
OS/8 files are always generated so as to be
multiples of 16 blocks long. For this and other
reasons, the resulting OS/8 files may be longer
than necessary.
To recapture the wasted space,
use OS/8 PIP to copy the file back to itself using
the /A option.

12.2

Error Messages

In some cases, illegal or bad syntax on input causes the
be reasked.

question

Other errors generate one of the following:
Message

Explanation

BAD CHAR

Warning message logged for each attempt
to convert an OS/8 character you tried
to convert to COS which is not in COS
character set.
Characters such as ~,
CTRL/Q, etc. Note:
is translated to
<, tabs are converted to proper number
of spaces. Vertical tabs, line feeds,
form feeds,
rubouts etc. are ignored.
Lower case characters are converted to
upper case.

BAD DEV

Specified an output device with
-S
option and output device was not same
type physical device as system device.
e.g.
FILE,DKI when system device is
DTO.

BAD DIRECTORY

Attempted to reference or store a
on a device with no directory
directory that has been destroyed.

FULL

Specified an output device with
-S
option and source file being created was
filled.

NO END

Entire OS/8 ASCII input file has been
read but no end of file (CTRL/Z) was
found.

12-3

file
or a

Message

Explanation

NO ROOM

Directory of output device was full;
room for specified output file.

NONCE ERROR

User tried to use
a
feature
not
currently supported or documented (but
might be implemented in some future
release) .

NOT FOUND

File with specified name was not found.

12-4

CHAPTER 13
FORMAT PROGRAMS

There are four format programs provided with
mark disk or DECtape for use with COS.

the

COS

package

which

These programs are:
RK8MRK

Formats an RK08 disk.

RKEMRK

Formats an RK8E disk.

TDMARK

Formats a DECtape on a TD8E drive.

DTMARK

Formats a DECtape on a TC08 drive.

These programs are described in detail in diagnostic documents
(RK08
Disk Formatter, DEC-08-DSKB; RK8E Disk Formatter, DEC-08-DHRKD-A-D;
TD8-E DECtape Formatter, DEC-8E-EUZC-Di . TCOl
TUSS
DECtape
Formatter, DEC-08-EUFB-D) provided with the system. The procedures in
the following paragraphs cover the use of these programs with the COS
system.
NOTE
The Format programs cannot be used when
the
COS
300 Foreground Data Entry
program is running.

13-1

RK8MRK
13.1

RK8MRK

Mount the system tape or disk and the disk to be formatted.

13.1.1

Operating Procedure

The command to run RK8MRK is:
~RUN

RK8MRK

and the CR key.
RK8MRK replies with the question:
DRIVE NO.?
Type the unit number of the RK08 disk to be formatted.
Be sure that
the unit specified is not a good system device since formatting
destroys the contents of a disk.
DRIVE NO.?

RK8MRK displays the question:
WRITE STANDARD FORMAT?

(TYPE "Y" FOR YES)

Type Y to format the disk for use with COS. Any other reply
a non-standard format which can not be used with COS.
When formatting is

co~plete

provides

RK8MRK displays the messages:

DISK FORMATTED
and
DRIVE NO?
At this point specify another drive to be formatted or type CTRL/C
return to the COS Monitor.

13.1.2

Error Messages

All error messages displayed contain the track
number where the error occurred. For example:
TRACK 0104 SECTOR 02
PARITY ERROR

13-2

number

and

sector

Message

Explanation

CONTROL BUSY ERROR

Disk lOT issued when control
which would effect operation.

DATA RATE ERROR

The processor was busy and did not
respond to a data break request within
the 13 micro seconds required.
The
transfer is terminated immediately.

EXEC WORD ERROR

An error occurred while
EXEC. word of the sector.

reading

the

FORMAT ERROR

An error occurred
header word.

reading

the

FORMAT PARITY ERROR

A parity error occurred in either the
Header word or Exec word while reading
the sector.

PARITY ERROR

A bit in data, parity or timing has been
picked up or dropped on read. Transfer
continues to the end of the sector where
the error occurred. Word count, current
address information can be used
to
identify the error.

SECTOR NO GOOD

The program attempted to read or write
data on a sector whose header words
indicated a bad sector. The transfer is
terminated immediately.

TIME OUT ERROR

The control
did
not
complete
operation after 32 revolutions.

TRACK ADDRESS ERROR

Track, surface or secto~ address read
from the disk did not agree with the
address count registers or the disk
drive
electronics
indicated
track
position 000 and the track counter did
not agree.
The transfer is terminated
immediately.

while

was

busy

TRACK CAPACITY EXCEEDED ERROR
The program attempted to read
beyond sector 17.
WRITE LOCK ERROR

write

The program attempted to write a section
that was write protected.
The write
operation is terminated immediately.

Error Recovery
If the program halts with the program counter = 1337 indicating
non-existent drive or device not adjusted correctly. If CTRL/C does
not work, restart the COS system.
(Refer to Appendix B.) If an error
reoccurs, the disk may be bad.

13-3

At the end of each track in which an error occurred on the
read operation the message:
RE-WRITE TRACK (Y FOR YES)?
is displayed.

Type Y to rewrite the track.

13-4

write

and

13.2

RKEMRK

Mount the system tape and the disk to be formatted.

13.2.1

Operating Procedure

The command to run RKEMRK is:
....:...R RKEMRK
and the CR key.
RKEMRK replies with the following:
RK8E DISK FORMATTER PROGRAM
FOR ALL QUESTIONS, ANSWER Y FOR YES OR N FOR NO.
The following questions are asked:
FORMAT DISK O?
FORMAT DISK l?
FORMAT DISK 2?
FORMAT DISK 3?
ARE YOU SURE?
After answering Y or N to each of the above questions the
formatted and the following message is printed:

disk(s)

RK8E DISK FORMATTER PASS COMPLETE
FORMAT SAME DISK(S) AGAIN?
Answer Y or N. At this point each of the FORMAT DISK n?
questions
are printed again.
Exit to Monitor by typing CTRL/C or continue
procedure for formatting disks.

13.2.2

Error Messages

The following error messages may be displayed
disk for COS:

when

marking

RK8E

DISK DATA ERROR
READ STATUS ERROR
WRITE STATUS ERROR
RECALIBRATE STATUS ERROR
After the error header mentioned above is displayed, the terminal will
print some of the following error
information pertaining to the
failure.
PC:

PROGRAM LOCATION OF FAILURE

CD:

EXPECTED INFORMATION

13-5

ST:

CONTENTS OF THE STATUS REGISTER

CM:

SOFTWARE COMMAND REGISTER

DA:

SOFTWARE CYLINDER, SURFACE, AND SECTOR REGISTER

AD:

ADDRESS OF DATA BREAK

DT:

DATA FOUND DURING DATA BREAK

After the error information is displayed, the terminal will display
one of the following questions asking the error recovery desired.
1.

If the error was a recalibrate error, the following
will be displayed:

question

TRY TO RECALIBRATE SAME DISK AGAIN?
Typing a Y for yes will result in a repeat of the recalibrate
sequence on the disk in error. Typing N for no will result
in progressing to the next available disk.
2.

If the error was a write error, the following
be displayed:

question

will

TRY TO FORMAT SAME CYLINDER AGAIN?
Typing Y for yes will result in a repeat of the write
sequence on the current cylinder.
Typing N for no will
result in progressing to the next sequential cylinder.
3.

If the error was a READ or
question will be displayed:

check

error,

the

following

TRY TO CHECK SAME CYLINDER AGAIN?
Typing a Y for yes will result in a repeat in the READ and
check sequence on the current cylinder. Typing a N for no
will result in progressing to the next sequential cylinder.

13-6

13.3

TDMARK

TDMARK records the timing and mark tracks on a DECtape mounted on
TD8E transport.

13.3.1

the

Operating Procedures

The command to run TDMARK is:
..:..RUN TDMARK
and the CR key.

TDMARK replies with the question:

UNIT?
Before replying to this question, mount the DECtapes to be marked with
just enough turns on the right hand reel to provide a grip. Make sure
that no two tape units are set to the same unit number; set the WRITE
ENABLE/WRITE LOCK switch to WRITE ENABLE and the REMOTE/OFF/LOCAL
switch to REMOTE. Set the WTM (Write Timing Mark) switch to on
(in
down posi tio"n) •
This red swi tch is located on the upper right hand
side of Module 868
(printed circuit card with wires)
within the
processor (the black box behind the lights and switches).
Answer the UNIT question with one or two unit numbers, corresponding
to the units of the tapes to be marked. For instance, if tapes are
mounted on units 0 and 1, type:

o I..}
Spaces are ignored, so it makes no difference if spaces are typed
between the unit numbers.
Only one specification of a unit is
significant, i.e., typing 000111 has the same effect as typing 01.
Only tapes on units 0 and 1 may be formatted.
If an error is detected
in the response to this question, the question is repeated.
TDMARK
replies with:
FORMAT?
Type MARK and the CR key and the program assumes 201 words 2702 blocks
(standard PDP-8 format). Any other reply produces a format which is
not compatible with COS.
TDMARK displays the message:
0201 WORDS,2702 BLOCKS.OK?(YES OR NO)
If NO is typed, the program reverts to the FORMAT?
the program displays:

question.

Y~S,

SET SWITCH TO WTM
Type the CR key and if the switch is set, the tape on the
specified begins to move.

13-7

first

unit

Once all of the tapes specified have been marked, the message:
SET SWITCH TO OFF
is displayed. Reset the WTM switch to normal {up position}
and type
the CR key to start the second pass. Note that during the second pass
with multiple DECtape units, as soon as one tape stops, the next tape
starts;
the first tape -is completed and may be replaced with a fresh
tape in preparation for formatting additional tapes. When formatting
is completed, the message:
FORMAT
is displayed. Typing SAME and the CR key repeats the entire process
with the same format.
The new DECtapes must be mounted and ready
before the CR key is typed in response to the SET SWITCH TO WTM
message.
When all tapes are formatted, type CTRL/C to return to the
COS monitor.
13.3.2

Error Messages

The following error messages may be displayed when marking a tape
COS.

for

Error messages for response to SET SWITCH TO WTM:
Message
SETUP

Explanation
Indicates an error in the DECtape setup.
One of the units specified is in WRITE
LOCK position, not selected, or the
write flip-flop is unable to be set, or
there may be a timing error.
{After
UNIT message is displayed.}

SWITCH NOT SET TO WTM OR SINGLE LINE FLAG
SWITCH TO WTM.

FAILED

SET.

SET

Indicates the switch on the Ma68 module
is not set to the WTM position or the
timing generator for writing the mark
and timing tracks is not setting the
single line flag.
If the switch was not set to
WTM
position, set the switch and type the CR
key.
If the switch was set to WTM position
and this error occurred, try again or
examine the timing generator circuit.
Error messages for marking and verifying a tape:
PC
PC
PC
PC
PC
PC

xxxx
xxxx
xxxx
xxxx
xxxx
xxxx

BLOCK NUMBER ERROR PHASE n
CHECKSUM ERROR PHASE n
DATA ERROR PHASE n
MARK TRACK ERROR PHASE n
TIMING ERROR PHASE n
WRITE ERROR PHASE n

13-8

where xxxx equals the program counter (PC) at the time of the failure.
n equals the pass in which the error occurred.
(See Recovery.)
Recovery:
Although an error should cause doubt concerning the entire process. a
restart may be made (except in phase 0) by typing RETRY
RETRY
causes the program to go back to phase 1. Type RESTART
to return to
the UNIT? question.
Phase 0:
Phase 1:
Phase 2:
Phase 3:
Phase 4 :
Phase 5:

write timing and mark track forward.
Reads mark track reverse.
write data, forward block and reverse block
numbers forward and writes the checksums.
Displays block numbers in AC (acc umul ator)
reverse.
Reads data, forward block and reverse block
numbers forward and calculates the checksum.
Reads reverse block numbers in reverse.

The entire program may be restarted at 0200 any time.

13-9

DTMARK

13.4

DTMARK

DTMARK records the required timing and mark tracks on a DECtape
mounted on the TCOl-TU55 DECtape unit.
Up to eight DECtapes may be
formatted at a time, assuming the system has eight tape transports.

13.4.1

Operating Procedures

The command to run DTMARK is:
~UN

DTMARK

and the CR key.

DTMARK replies with:

.QTh!

Mount the DECtapes to be marked on the tape transports, with just
enough turns of tape on the right hand reel of each transport to
provide a grip. Make sure that no two tape units are set to the same
unit number.
Set the WRTM (WRite Timing Mark) switch located on the
hardware frame to the down position (on). Set the WRITE ENABLE/WRITE
LOCK switch to WRITE ENABLE and the REMOTE/OFF/LOCAL switch to REMOTE.
Reply to the DTA? question with a unit number or series of unit
numbers, corresponding to the DECtape units where the tapes are
mounted.
For instance, if tapes are mounted on units 2 and 3,
type
2 3 and the CR key.
Spaces are ignored, so it makes no difference if
spaces are typed between the unit numbers.
Only one specification of a unit is significant, i.e., typing
has the same effect as typing 2 3~.

2 2 3

The program displays:
DIRECT?
Type MARK and the CR key to format the DECtape for use with COS.
Any
other reply causes a format which is incompatible with COS.
(Refer to
TCOl-TU55 DECtape Formatter Writeup.)
The program assumes 201(8)
words,
2702(8)
blocks
(standard PDP-8 format)
and displays the
message:
0201 WORDS,2702 BLOCKS.OK?

(YES OR NO)

If a NO~ is typed, the program reverts to the DIRECT? question.
YES~, the tape on the first unit specified begins to move.

Once all of the tapes specified have been marked, the message:
SET SWITCH TO NORMAL
is displayed.
Return the WRTM switch to normal (up position) and type
the CR key to start the second pass. Note that during the second pass
with multiple DECtape units, as soon as one tape stops, the next tape
starts.
The first tape is completed and may be replaced with a fresh

13..;.10

tape in preparation for rerunning the
tapes.

program

format

additional

When DTMARK has formatted the tapes specified, the message:
DIRECT?
is displayed again. Type SAME and the CR key to repeat the entire
process with the same format.
The new DECtapes must be mounted and
ready before SAME.) is typed.
(0 means DECtape unit 8).
When all tapes are
Monitor.

13.4.2

formatted,

type

CTRL/C

return

the

COS

cause

Error Messages

Errors typed in response to the DTA? and DIRECT?
return to DTA? or DIRECT? respectively.
Error messages for response to OK?

questions

(YES OR NO):

Message

Explanation

SETUP?

Indicates an error in the DECtape setup, unit in
WRITE LOCK position, nonselectable unit or switch
not in WRTM position.
After display of this
message, the DTA? question is displayed.

Error messages which may occur during marking and verifying a tape:
xxxx SHOULD BE yyyy BLK ERROR PHASE n
xxxx SHOULD BE yyyy DATA ERROR PHASE n
END TAPE ERROR PHASE n
LAST INT NOT END ZONE
MARK TRACK ERROR PHASE n
PARITY ERROR PHASE n
SELECT ERROR PHASE n
TIMING ERROR PHASE n
Recovery from error messages:
Although an error should cause doubt concerning the entire process,
restarts may be made (except when in phase 0), by typing RETRY and the
CR key.
RETRY causes the program to return to PHASE 1. Type RESTART
and the CR key to return to the DTA? question.
PHASE
PHASE
PHASE
PHASE

0:
1:
2:
3.

Mark track write.
Writing last reverse block number forward.
Writing block numbers and data in reverse.
Reading and checking block numbers and data.

If it is necessary to restart the DTMARK program,
return to the COS Monitor then type RUN DTMARK.

type

CTRL/C

If CTRL/C does not work, rebootstrap the COS system (refer to Appendix
B) and then call DTMARK.

13-11

CHAPTER 14
CREF

CREF (Cross REFerence) is a utility program.
Primarily an aid to
program development,
it provides an alphabetical listing of all
symbols used in a DIBOL program, along with the line number where each
symbol is defined and all the line numbers where each symbol is used.
CREF requires 8K of core and can handle any reasonable 8K program.
If
12K of core is available CREF expands its tables to make use of the
extra space. CREF expects the source program to be free from errors.
Only a very minimal amount of error checking is preformed by CREF and
no attempt should be made to CREF programs with compilation errors.
If CREF finds a line it cannot handle it will print:
xxxx IS BEING IGNORED
where xxxx is a line number.
For an example of CREF operation, consider the listing in Figure 14-1.
It is a short program that computes the square roots of numbers from 1
to 99.
(Output of program is shown in Figure 14-3.) Its CREF table,
Figure 14-2, illustrates the basic functions of CREF.
The columns of the CREF Table are:

14.1

NAME

Name of the symbol.

DEF

Line number
defined.

program

where

symbol

REFERENCES

Line numbers
referenced.

program

where

symbol

OPERATING PROCEDURES

Type:
~RUN

CREF[,filel,file2, .•. ]

and the CR key.

14-1

Where
filel,file2, •.•

are the parts of a DIBOL program, just as
in running COMP (maximum 7). If no files
are specified, the Editor scratch area is
used.

CREF reads the DIBOL program, produces its table on the
and returns control to the COS Monitor.

14-2

line

printer

COS DIBOL
0010
0020
0030
0040
0050
0060
0070
0080
0090
0100
0110
0120

01~MAR-73 THUR
COMPILATION LISTING
V12:28 PAGE 01
DATA DIVISION
S QUA R E ROO T SUB R 0 UTI N E

START

S Q U ARE

A,
S,

RECORD PRINT
AIS
AIS

X,
SQRTX,
TEMPX,
N,

RECORD
DIS
015
DIS
03,001

R 0

o T

Figure 14-1
Square Root Subroutine
(Sheet 1 of 3)

14-3

S U B R 0 U T I N E

COS DIBOL
01-MAR-73 THUR
PROCEDURE DIVISION
0130
0140
0150
0160
0170
0180
0190
0200
0210
0220
0230
0240
0250
0260
0270
0280
0290
0300

COMPILATION LISTING

V12:28

PROC 1
LOOP,

INIT(l,LPT)
X=N*1000000
CALL SQRT
A=N
S = SQRTX,'XXX.XXX'
INCR N'
XMIT(l,PRINT)
IF (N.LE.99) GOTO LOOP
STOP

SQRT,
TEMPX=X/2
iTRIAL VALUE
SQLOOP, SQRTX=(TEMPX+(X/TEMPX))/2
iNEWTON'S METHOD
IF(SQRTX-TEMPX .EQ. 0) RETURN
TEMPX=SQRTX
GOTO SQLOOP
END

Figure 14-1
Square Root Subroutine
(Sheet 2 of 3)

14-4

PAGE 02

01-MAR-73 THUR
COS DIBOL
NAME
TYPE
#
0001
0002
0003
0004
0005
0006
0007
0010
0011
0012
0013
0014
0015
0016
0017
0020

PRINT
A
S
X
SQRTX
TEMPX
N
• .1
LOOP
• .1000
SQRT
• • XXX.
•• 99
• .2
SQLOOP
• •0

RECORD
ALPHA
ALPHA
DECMAL
DECMAL
DECMAL
DECMAL
DECMAL
LABEL
DEC MAL
LABEL
ALPHA
DECMAL
DEC MAL
LABEL
DECMAL

STORAGE MAP LISTING
DIM SIZE
ORIGIN
01
01
01
01
01
01
01
01
00
01
00
01
01
01
00
01

32
15
15
15
15
15
03
01
01
07
01
07
02
01
01
01

V12:28

PAGE 03

20000
20002
20021
20042
20061
20100
20117
20122
10071
20123
10132
20132
20141
20143
10141
20144

0016 SYMBOLS
NO ERRORS DETECTED.

08 K CORE REQUIRED [3940 FREE LOCS

Figure 14-1
Square Root Subroutine
(Sheet 3 of 3)

14-5

14 BUFFERS]

COS300 CREF

Vll.08

SYMBOL

DEF

REFERENCES

A
LOOP
N
PRINT
S
SQLOOP
SQRT
SQRTX
TEMPX
X

40
160
110
30
50
260
250
90
100
80

180
220
160
210
190
290
170
190
250
160

THUR

01-MAR-73

180

200

220

260
260
250

270
260
260

280
270

280

Figure 14-2
CREF of Square Root Subroutine

14-6

PAGE 1

1
2
3
4
5
6
7
8
9
10
11
12
13
14
1-5
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55

1.000
1.414
1.732
2.000
2.236
2.449
2.645
2.828
3.000
3.162
3.316
3.464
3.605
3.741
3.872
4.000
4.123
4.242
4.358
4.472
4.582
4.690
4.795
4.898
5.000
5.099
5.196
5.291
5.385
5.477
5.567
5.656
5.744
5.830
5.916
6.000
6.082
6.164
6.244
6.324
6.403
6.480
6.557
6.633
6.708
6.782
6.855
6.928
7.000
7.071
7.141
7.211
7.280
7.348
7.416
Figure 14-3
Output of Square Root Subroutine
(Sheet 1 of 2)

14-7

56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99

7.483
7.549
7.615
7.681
7.745
7.810
7.874
7.937
8.000
8.062
8.124
8.185
8.246
8.306
8.366
8.426
8.485
8.544
8.602
8.660
8.717
8.774
8.831
8.888
8.944
9.000
9.055
9.110
9.165
9.219
9.273
9.327
9.380
9.433
9.486
9.539
9.591
9.643
9.695
9.746
9.797
9.848
9.899
9.949

Figure 14-3
Output of Square Root Subroutine
(Sheet 2 of 2)

14-8

14.2

ERROR MESSAGES

Although CREF expects the source to be error free,
errors are detected.
xxxx IS BEING IGNORED

CREF has encountered a
interpret.

certain
line

types
it

cannot

In addition to the above message, CREF prints an informative message:
SYMBOLS DEFINED BUT NOT REFERENCED;

Symbols were found that were defined but
not referenced by the program.

14-9

CHAPTER 15
PRINT

PRINT is a utility program that-considerably eases the creation of
report programs.
Using a parameter file which describes the report,
PRINT generates a DIBOL program which will produce that report.
PRINT is actually two programs. The first program reads and validates
the parameters while creating core table entries.
If no parameter
errors are detected, the second phase is entered which produces the
DIBOL report program. The programs which generate the coding for the
report require 12K words of memory.
However, execution of the
resulting program requires 8K words.

15.1

DESCRIPTION OF THE PARAMETER FILE

THE PARAMETER FILE has six sections:
IDENT
HEAD 1
HEAD2
INPUT
COMPUTE PRINT
END

to identify the program and author
to provide page headings for
the report
to describe the input file
to describe any computation to be done
to describe the report column headings
an optional directive at the end of the parameter file

Each of these sections is described below.
Optional clauses are surrounded by parentheses.

15.1.1

IDENT

The syntax is:
IDENT

program

[/u]

15-1

[,author]

where:
program
u

author

is the name of the generated program
1-6 characters
is a logical unit onto which the program is
written
any text from 1 to 24 characters

Example:
0010 IDENT TEST4l/l4, JOHN DOE

15.1.2

HEADI and HEAD2

HEADI is the first heading line on each page.
HEAD2 is the second
line. HEADI and HEAD2 are both optional. The only difference is that
HEADI information will be expanded if space permits by inserting a
space between each character.
The syntax is:
[HEADI 'tex t ' ]
[HEAD2 'tex t']
Where text is a string of any characters for the DIBOL set,
of single quotes up to 132 characters.
There may be more than one HEADI or HEAD2 line.
the individual texts are linked together.

exclusive

If this is the

case,

Example:
0020 HEADI 'ACCUMULATION, COMPUTATION, AND SUMMARY RESULTS'
0030 HEAD2 'TEST RESULTS FOR AUGUST, 1973'

15.1.3

INPUT

The INPUT section consists of the INPUT statement on one line,
followed by separate lines describing the fields of the input record
The syntax is:
INPUT

[file

[/u]

[ , S] ]

where:
file
u
S

is the name of the file to be used as input
is the logical unit on which the file resides
is included if the report is to be summarized,
instead of detailed

If the file name is not included, the generated program
it when the report is run.

15-2

will

request

The syntax of the field description is:
A

[name] ,

[ .m]

[ ,Lr

[P] ]

where:
name is the name of the field
A
is alpha or decimal field type
D

n
m

is the size of the field, expressed in characters
is the number of the decimal places in the field,
and
is
valid only for decimal fields
is used only for the break fields (a break field is used
in
conjuction with accumulate to print totals);
r is a single
digit expressing the relative importance of the field, with
1 being least important and 9 being most important; when
totals are printed for a more important break,
totals for
all lesser breaks will also be printed.
means that the report will start a new page after the totals
for this break are printed.

The maximum number of fields in 20.
Example:
0040 INPUT TESTD/5,S
0050
ABLE A5,Ll
0060
BAKER,D4
0070
CHARLI,D5.2,L2

15.1.4

COMPUTE

This optional section starts with a line containing only:
[COMPUTE]
On succeeding lines appear DIBOL statements of the form:
name

= expression

where:
name

is the name of a result field, and must not duplicate any
input field name, nor any previous computation result
name.

expression
is any valid DIBOL expression.
It may be a single
field or literal, or any decimal expression.

alpha

It is important to note that PRINT, unlike DIBOL, notes the number
decimal places.
It will not allow syntax of the form:
expr 1 + expr 2
unless expr 1 and expr2 have an equal number of decimal places.

15-3

To determine the number of decimal places of any expression,
following rules:
a.

use

the

expr 1 + expr 2
No change
expression
expr 2.

in number of decimal places.
The resultant
will have the same number of places as expr 1 and

expr 1 - expr 2
(same as a)
expr 1 * expr 2

The number of decimal places in the resultant expression
the sum of the decimal places in expr 1 and expr 2.
d.

expr ljexpr 2
The number of decimal places in the resultant expression is
the difference in decimal places between expr 1 and expr 2.

To adjust the number of decimal places of any expression, multiply by
the constant 1.
For example, expr 1 has 2 places, and expr 2 no
places, then
expr 1

* e~pr 2 * 1.00

nas four decimal places.
Note that PRINT allows decimal places in decimal constants.

15.1.5

The PRINT section begins with a line containing only:
PRINT
followed by lines describing up to
syntax of these descriptions is:
name,

ltext'

[,A]

fields

printed.

The

[,picture]

Where:
name

is the name of the field, and must be
field or the result of a computation

text

is a string of characters from the DIBOL set, exculsive of
single quote.
This text is used as the heading of the
individual columns, and in the total lines for break
fields

is present only if the field is to be accumulated, and the
sum to be printed in total lines. The field must not be
alpha

15-4

either

input

pic ture
is a string of text equivalent to the DIBOL format field,
and
is valid only for decimal fields.
If the picture is
not included, PRINT will create one using the description
of the field (if this is an accumulated field, two extra
places will be assumed).
That created picture will be of
the form

xx,xxx.xxwith appropriate number of decimal and integer places.
The text used for titles may be several words separated by asterisks.
This will cause each word to be centered over a column, in separate
lines.
For example:
GPAY,

'GROSS*PAY'

will cause
GROSS
PAY
to be placed over the GPAY column.
then:

If GPAY is

also

break

field,

GROSS PAY TOTAL
will be used on the total line, instead of GROSS*PAY TOTAL.
The field descriptor lines may also be of the form:
,An
Where n is the size of the field.
in th ere po r t .

This will produce n

blank

columns

If any two field descriptors are not separated by a filler
(An)
descriptor,
then two blank columns will separate the fields in the
report.
Example:

0080 PRINT
ABLE, 'ABLE'
0090
0100
CHARLI, 'CHARLIE'
0110
BAKER, 'BAKER' ,A
0120
0130 END

15.1.6
The END directive is optional, and consists of a line containing
the word:
END

15-5

only

15.2

OPERATING INSTRUCTIONS

15.2.1

Parameter File

Using the editor commands, create the parameter file and write it
as a source file.

15.2.2

out

Run PRINT using the command:
RUN

PRINTA

PRINTB,

file/xy

where:
file

is the name of the source (parameter)
file which was
just written out
is a switch which determines whether the parameter file
is to be listed; N means no list, L means list
is a switch which determines whether the DIBOL program
(data file)
created by running PRINT is to be listed;
N means no list, L means list.

x
y

If x and yare both N, the switches and their preceding slash may be
omitted.
The output from this operation is a data file which must be
converted to a source file (see Section 15.2.3).

15.2.3

CONVEX

Run CONVEX (see Chapter 12 of the COS 300 SYSTEM REFERENCE MANUAL)
to
create a source file from the data file written by PRINT.
Printouts
by CONVEX are underlined, operator response is not;
user
input must
be terminated with a carriage return .
. RUN CONVEX
VERSION 3.07
INPUT MODE-

FILE NAME- $RPG/u
OUTPUT MODE-

FILE NAME-

program

where us is the logical unit

Program is any name desired, but is probably the same name as used
the IDENT line of the parameter file, i.e.,

TEST4A
When CONVEX asks for INPUT MODE again, type CTRL/C to get back to
Monitor.

15-6

the

15.2.4

Compilation

Compile the program
.RUN COMP, TEST4A
.SAVE TEST4A

15.2.5

Program Execution

Run the report program:
.RUN TEST4A

15.3

TEST CASES

Assume the following data file for all examples in Section 15.3.
There are three fields and 121 records.
If the file were dumped, it
would contain:

AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI

A0002
A0002
A0003
A0003
A0003
A0003
A0003
A0003
A0003
A0003
A0004
A0004
A0004
A0004
A0004
A0004
A0005
A0005
A0005
A0005
A0005
A0005
A0006
A0007
A0007
A0007
A0007
A0007

0191
0133
0256
0049
0241
0384
0283
0184
0100
0030
0147
0376
0241
0386
0290
0222
0287
0207
0215
0040
0196
0159
0386
0289
0217
0082
0409
0403
0377
0247
0007
0153
0129
0235
0355
0138
0278

12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346
12346

1Z346
12346
12346
12346
12346
12346
12346
12346

15-7

A0007
A0007
A0007

AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
A0002
A0002
A0003
A0004
A0004
A0004

AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
A0002
A0002
A0002
A0002
A0002
A0002
A0003
A0003
A0003
A0003
A0004
A0004
A0004
A0004
A0004
A0004
A0004
A0004
A0005
A0005
A0005
A0005
A0005
A0005
A0005

AOOOI
AOOOI
AOOOI
A0002
A0002
A0002
A0002
A0002
A0002
A0002
A0002
A0002

0159
0382
0269
0038
0188
0117
0173
0304
0292
0231
0332
0023
0149
0332
0107
0207
0215
0256
0063
0211
0325
0396
0339
0055
0130
0239
0375
0238
0368
0203
0146
0321
0374
0229
0387
0296
0252
0028
0139
0281
0176
0060
0258
0058
0289
0213
0242
0391
0318
0370
0212
0238
0218
0268
0112
0149
0335
0034
0170
0027
0133

12346
12346
12346
12347
12347
12347
12347
12348
12348
12348
12348
12348
12348
12348
12348
12348
12348
12348
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12349
12340
12349
12349
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350

15-8

A0003
A0003
A0003
A0003
A0003
A0003
A0004
A0004
A0005
A0005
A0005
A0005
A0005
A0005
A0005
A0005
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
15.3.1

0026
0129
0231
0335
0038
0187
0161
0392
0379
0256
0052
0256
0050
0249
0013
0061
0196
0157
0376
0239
0376
0240
0381

12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12350
12351
12351
12351
12351
12351
12351
12351

Example 1

To list each record of this file without any of the special features
of PRINT other than page headings, the following control program can
be used:
0010 IDENT TEST1/2,JOHN SMITH
0020 HEAD1 'NO ACCUMULATE'
0030 HEAD2 'NO SUMMARY, NO BREAKS'
0040 INPUT TESTD/1
0050
ABLE,A5
0060
BAKER, D4
0070
CHARLI,D5.2
0080
0090 PRINT
0100
ABLE,'ABLE'
0110
BAKER,'BAKER'
0120
CHARLI,'CHARLIE'
Follow the instructions in Section 15.2, then test the DIBOL
The report is shown below:
PAGE
TEST 9/13/73
NO ACCUMULATE
NO SUMMARY,NO BREAKS
ABLE
AOOOI
AOO01
AOOOI
AOO01
AOOOI
AOO01
AOOOI
AOOOI
AOOOI
AOO02

BAKER
101
133
256
49
241
384
283
184
100
30

CHARLIE
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46

15-9

program.

AOO02 147
AOO03 376
AOO03 241
AOO03 386
AOO03 290
AOO03 222
AOO03 287
AOO03 207
AOO03 215
AOO04 40
AOO04 196
AOO04 159
AOO04 386
AOO04 289
AOO04 217
AOO05 82
AOO05 409
AOO05 403
AOO05 377
AOO05 247
AOO05
7
AOO06 153
AOO07 129
AOO07 235
AOO07 355
AOO07 138
AOO07 278
AOO07 159
AOO07 382
AOO07 269
AOOOI 38
AOOOI 188
AOOOI 117
AOOOI 173
AOOOI 304
AOOOI 292
AOOOI 231
AOOOI 332
AOOOI 23
AOO02 149
AOO02 332
AOO03 107
AOO04 207
AOO04 215

123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.46
123.47
123.47
123.47
123.47
123.48
123.48
123.48
123.48
123.48
123.48
123.48
123.48
123.48
123.48

TEST1 9/13/73
PAGE
NO ACCUMULATE
NO SUMMARY, NO BREAKS
ABLE

BAKER

CHARLIE

AOO04
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOOOI
AOO02
AOO02
AOOO2

256
63
311
325
396
339
55
130
239
375

123.48
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49

15-10

AOO02 238
AOO02 368
AOO02 203
AOO03 146
AOO03 321
AOO03 374
AOO03 229
AOO04 387
AOO04 296
AOO04 252
AOO04 28
AOO04 139
AOO04 281
AOO04 176
AOO04 60
AOOOS 258
AOOOS 58
AOOOS 289
AOOOS 213
AOOOS 242
AOOOS 391
AOOOS 318
AOOOI 370
AOOOI 212
AOOOI 238
AOO02 218
AOO02 268
AOO02 112
AOO02 149
AOO02 335
AOO02 34
AOO02 170
AOO02 27
AOO02 133
AOO02 26
AOO03 129
AOO03 231
AOO03 335
AOO03 38
AOO03 187
AOO04 161
AOO04 392
AOOOS 379
AOOOS 256

123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.49
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50
123.50

TEST1 9/13/73 PAGE
3
NO ACCUMULATE
NO SUMMARY, NO BREAKS
ABLE

BAKER

CHARLIE

AOOOS
AOOOS
AOOOS
AOOOS
AOOOS
AOOOS
AOOOI
AOOOI
AOOOI
AOOOI

52
256
50
249
13
61
196
157
376
239

123.50
123.50
123.50
123.50
123.50
123.50
123.51
123.51
123.51
123.51

15-11

AOOOI 376
AOOOI 240
AOOOI 381

15.3.2

123.51
123.51
123.51

Example 2

Assume that the following listing is desired:
first,
third,
Print the fields
in the following order:
second
Put addi tional spacing between fields.
total
Accumulate totals for the second field and print the
whenever the first or third field changes.
When the third field changes, skip to a new page.

0
0
0
0

These features may be obtained by the following statements:
0100
0110
0120
0125
0130
0140
0150
0160
0170
0180
0190
0200
0210
0220
0230

IDENT
HEADI
HEAD2
HEAD2
INPUT

TEST2,JOHN DOE
'ACCUMULATE'
'NO SUMMARY,'
'NO BREAKS'
TESTD
ABLE,AS, Ll
BAKER, D4
CHARLI,DS.2,L2P

PRINT
CHARLI, 'CHARLIE'
,AS
ABLE, 'ABLE, 'ABLE*TWO-LINE'
,AS
BAKER, 'BAKER',A

TEST2
9/13/73 PAGE
A C CUM U L ATE
NO SUMMARY,NO BREAKS

CHARLIE

ABLE
TWO-LINE

BAKER

123.46

AOOOI

191
133
256
49
241
384
283
184
100

ABLE TWO-LINE TOTAL
123.46

AOO02

ABLE TWO-LINE TOTAL
123.46

AOO03

1,821
30
147
177
376
241

15-12

286
290
222
287
207
215
ABLE TWO-LINE TOTAL
123.46

AOO04

ABLE TWO-LINE TOTAL
123.46

AOO05

ABLE TWO-LINE TOTAL
123.46

AOO06

ABLE TWO-LINE TOTAL
123.46

AOO07

2,224
40
196
159
286
289
217
1,287
82
409
403
377
247
7
1,525
153
153
129
235
355

TEST2
9/13/73
A C CUM U L ATE
NO SUMMARY,NO BREAKS

PAGE

CHARLIE

ABLE
TWO-LINE

BAKER

123.46

AOO07

138
278
159
382
269

ABLE TWO-LINE TOTAL
CHARLIE TOTAL
TEST2

9/13/73
A C CUM U L ATE
NO SUMMARY,NO BREAKS

1,945
9,132
PAGE

CHARLIE

ABLE
TWO-LINE

BAKER

123.47

AOOOI

38
188
117

15-13

173
ABLE TWO-LINE TOTAL
CHARLIE TOTAL
TEST2

0/13/73
A C CUM U L ATE
NO SUMMARY,NO BREAKS

123.48

AOOOI

ABLE TWO-LINE
123.48

AOO02

ABLE TWO-LINE
123.48

AOO03

ABLE TWO-LINE TOTAL
123.48

AOO04

ABLE TWO-LINE TOTAL
CHARLIE TOTAL
TEST2

9/13/73
A C CUM U L ATE
NO SUMMARY,NO BREAKS

CHARLIE

ABLE
TWO-LINE

516
516
PAGE

304
292
231
332
23
1,182
149
332
481
107
107
207
215
256
678
2,448
PAGE

BAKER

123.49
AOOOI
311
325
396
339
55
ABLE TWO-LINE TOTAL
123.49
AOO02

1,489
130
239
375
238
368
203

ABLE TWO-LINE TOTAL

1,553

123.49

AOO03

146
321
374
229

15-14

ABLE TWO-LINE TOTAL
123.49

AOO04

ABLE TWO-LINE TOTAL
123.49

AOO05

ABLE TOW-LINE TOTAL
CHARLIE TOTAL
TEST2

9/13/73
A C CUM U L ATE
NO SUMMARY,NO BREAKS

1,070
387
296
252
28
139
281
176
60
1,619
258
58
289
213
242
391
318
1,769
7,500
PAGE

CHARLIE

ABLE
TWO-LINE

BAKER

123.50

AOOOI

370
212
238

ABLE TWO-LINE TOTAL
123.50

AOO02

ABLE TWO-LINE TOTAL
123.50

AOO03

ABLE TWO-LINE TOTAL
123.50

AOO04

820
218
268
112
149
335
34
170
27
133
1,446
26
129
231
335
38
187
946
161
392

15-15

ABLE TWO-LINE TOTAL
AOO05

123.50

ABLE TWO-LINE TOTAL
CHARLIE TOTAL
TEST2

553
379
256
52
256
50
249
13
61
1,316
5,081

9/13/72
PAGE
A C CUM U L ATE
NO SUMMARY, NO BREAKS

CHARLIE

ABLE
TWO-LINE

BAKER

123.51

AOOOI

196
157
376
239
240
381

ABLE TWO-LINE TOTAL
CHARLIE TOTAL
REPORT TOTAL

1,965
1,965
26,642

The difference between summarizing or
shown in Section 15.3.3.

15.3.3

not

summarizing

report

Example 3

The next control program will print summary information whenever
the
ABLE field changes.
For each field to be printed, specify A for
accumulate totals. Computational fields may also be printed if the A
option is selected.
0100
0110
0120
0130
0140
0150
0160
0170
0180
0190
0200
0210
0220
0230
0240
0250
0260

IDENT
HEADI
HEAD2
HEAD2
INPUT

TEST3,PROGRAMMER NAME
'ACCUMULATE'
'SUMMARY.'
',COMPUTES'
TESTD/l,S
ABLE,A5,Ll
BAKER, D4
CHARLI,D5.2
COMPUTE
VARNAM=BAKER*l.OO+CHARLI*BAKER+(CHARLI*CHARLI)#2
SALARY+CHARLI*BAKER
PRINT
ABLE, "ABLE*AA'
CHARLI,'CHARLIE',A
BAKER,'BAKER',A
VARNAM,'COMPUTATIONAL*FIELD',A
SALARY, 'SALARY' ,A

15-16

0270 END

SALARY

ABLE
AA

CHARLIE

BAKER

COMPUTATIONAL
FIELD

AOOOI
AOO02
AOO03
AOO04
AOO05
AOO06
AOO07
AOOOI
AOO02
AOO03
AOO04
AOOOI
AOO02
AOO03
AOO04
AOO05
AOOOI
AOO02
AOO03
AOO'04
AOO05
AOOOI

1,111.14
246.02
987.68
740.76
740.76
123.46
987.68
1,111.28
246.96
123.48
370.44
740.94
740.94
493.96
987.92
864.43
370.50
1,111.50
741.00
247.00
988.00
864.57

1,821
177
2,224
1,287
1,525
153
1,945
1,698
481
107
678
1,489
1,553
1,070
1,619
1,769
820
1,446
946
553
1,316
1,965

363,822.99
52,514.16
398,738.24
251,634.24
281,255.72
34,284.75
364,013.66
348,577.79
90,369.50
28,566.67
130,139.37
276,864.65
284,831.65
194,203.42
323,547.555
326,971.27
147,846.75
317,297.25
209,290.50
99,353.00
285,860.00
351,445.19

224,820.66
21,852.42
274,575.04
158,893.02
188,276.50
18,889.38
240,129.70
209,663.88
59,393.88
83,719.44
83,719.44
183,876.61
191,779.97
132,134.30
199,930.31
218,452.81
101,270.00
178,581.00
116,831.00
68,295.50
162,526.00
242,697.15

14,941.32

26,642

5,161,427.72

3,289,801.93

REPORT TOTAL

15.4

PAGE 1

10/05/73
A C CUM U L A T E
SUMMARY.,COMPUTES

TEST3

ERROR MESSAGES

PRINT evaluates each stacement for correctness.
Whenever errors
occur,
the line in error and a message are printed. The number
printed near the error message indicates the character position where
the error occurred.
The following error messages are printed by
PRINT.
Meaning

Message
ALPHA LITERAL REQUIRED

Expected
missing.

ALREADY DEFINED

An attempt was made to name a
field
in the INPUT or COMPUTE
section with a name that was
previously used.

HEADER IS TOO LONG

The header line
characters.

IMPROPER DEFINITION

Filler
item in
the
PRINT
section is used incorrectly.

IMPROPER LITERAL

Literal too large.

15-17

alpha

1 i teral

exceeds

132

Message

Meaning

IMPROPER USE OF DECIMAL PLACES

The number of decimal places
exceeds the size of the field
being defined.

INTEGER FROM 1-15 REQUIRED

The size of a decimal field
specified in the INPUT section
must be between 1-15.

INTEGER FROM 1-132 REQUIRED

The expected
out of range.

INTEGER REQUIRED

Integer
expected.

LITERAL TOO LONG

Field description
characters.

MUST BE IDENT

The first section
in
the
parameter file must be IDENT.

MUST BE NUMERIC ITEM

An item expected to be numeric
is defined incorrectly.

MUST BE S

S is the only legal option in
the
INPUT
statement
that
follows the comma.

NEED FILE NAME

File name missing
statement.

NO ENDING QUOTE

No closing quote for a HEADl,
HEAD2 or PRINT statement.

NO INPUT DIRECTIVE

INPUT
The
missing.

NO PRINT ITEMS

No
fields
are
specified
following the PRINT statement.

NOT DEFINED

An attempt was made to print a
field
that
has
not been
defined.

NOT ENOUGH RIGHT PARENTHESES

A statement in the COMPUTE
section
has too few right
parentheses.

PICTURE TOO LONG

The picture or edit mask
printing
exceeds
characters.

SYNTAX ERROR

Statement
contains
illegal
characters or options.

TOO MANY COLUMNS IN REPORT

Mor.e than 132 columns
HEADl,
HEAD2,
or
sections.

TOO MANY COMPUTE STATEMENTS

More
than
eight
COMPUTE
statements were specified.

15-18

decimal

fields

missing

where
exceeds

from

statement

IDENT

for
22

under
PRINT

Meaning

Message
TOO MANY DATA ITEMS

More than 20 data items in the
INPUT section.

TOO MANY LEFT PARENTHESES

A statement in the COMPUTE
section is too complicated to
be deciphered by PRINT.

TOO MANY LIST ITEMS

More than 20 data items in the
INPUT section.

TOO MANY RIGHT PARENTHESES

A statement
section has
parentheses.

UNKNOWN DIRECTIVE

An invalid section statement.
Only
IDENT,
HEAD1, HEAD2,
INPUT, COMPUTE, PRINT and END
are legal.

15.5

in the COMPUTE
too many right

COMPILING PRINT

PRINT consists of several source files:
0
0
0

PRINT1, PRINT2 are the data sections.
PRINT3, PRINT4, PRINT5, PRINT6 are the procedure sections
the parsing phase.
PRINT7, PRINT8, PRINT9, PRINTO are the procedure sections
the generation phase.

To compile:
.RUN COMP, PRINT1, PRINT2, PRINT3,PRINT4,PRINT5.PRINT6
. SAVE

PRINTA

.RUN COMP, PRINT1, PRINT2, PRINT7, PRINT8, PRINT9, PRINTO
. SAVE

PRINTB

15-19

of
of

CHAPTER 16
LINCHG

The regular page of computer printout is large enough to contain 66
lines of print. Although 66 lines per page is not always required by
the programmer, its use is sufficient to make it a standard.
Therefore, more often than not, when SYSGEN builds a system, the
lines-per-page default is 66.
This number may be changed at any time by reconfiguring SYSGEN. It is
normally not convenient, to change page size by running SYSGEN. On
printers, such as the LS8E, a paper tape drive ribbon is used to
further control the number of lines per page. When a different page
size is necessary, a specific size paper tape drive is installed.
However, in place of the mechanical forms control, it is possible to
use the software program LINCHG. LINCHG is a utility program which
allows the temporary use of any number of lines per page (1-99) with
programs (run singly or in batch), without affecting the SYSGEN
default number. To use LINCHG the programmer types:
.R LINCHG
The computer responds
HOW MANY LINES PER PAGE?
The programmer types the required number
10./
LINCHG now installs 10 as the number of lines per page.
(If <CR> is
typed instead of the number the system defaults to 66 lines-per-page.)
The same effect can be produced by typing
.R LINCHG/IO
which eliminates the HOW MANY LINES PER PAGE question (there
default in this case. The action must be explicitly stated.)

The LINCHG number will remain in force for all programs which follow
unless a further call to LINCHG is made, the system is rebooted, or
the system is closed down.

16-1

Example of LINCHG when the SYSGEN default is 66 lines per page:
-BATCH PRONAM
.0090 Run Job 1 - 66 lines-per-page (default)
.R LINCHG/iO
.000 Run Job 2 - 10 lines-per-page
.R LNCHG/33
.0110 Run Job 3 - 33 lines-per-page
.R LINCHG/99
.0120 Run Job 4 - 99 lines-per-page
.R LINCHG/66
.0130 Run Job 5 - 66 lines-per-page
The above example shows how a batch stream may be run starting with
the normal default for the first program, then incorporating a variety
of changes for the various programs which follow, and finishing with
the SYSGEN default number. The 66 lines per page may also have been
reestablished by rebooting the system.

16-2

APPENDIX A
COS CODES

In both source and data files, characters (alphanumeric and decimal)
are stored 2 characters per word in 6-bit binary form (combinations of
1 and 0). Negative numbers are stored with the high order bit of the
low-order digit set to 1.
For example, the number 1234- is stored as
two words in the form
22

WORD 1

65
4

WORD 2
(with high order bit on)

24
3

and would be recognized as 123T. This means that any program where
the decimal
to alpha conversion is neglected might produce negative
numbers with letters. Refer to Table A-I for
a list of equivalent
characters for negative numbers.
The sequence of the COS codes is the same as the output from the
utility.

A-l

SORT

Table A-I
COS Codes

Character

Equivalent
Negative
Numbers

0
1
2
3
4
5
6
7
8
9
A

C
D
E

F
G
H
I
J
K

L
M

p
Q
R

T
U
V

X
y
Z

null

-0
-1
-2
-3
-4
-5
-6
-7
-8
-9

COS Code
21
22
23
24
25
26
27
30
31
32
42
43
44
45
46
47
50
51
52
53
54
55
56
57
60
61
62
63
64
65
66
67
70
71
72
73
00

Character
!

#
$
%
&

(
)

:
;

>
?
@

[

]

Leader/Trailer
LINE FEED
Carriage Return
Space
RUB OUT
Blank
Bell
TAB
FORM

A-2

COS Code
02
03
04
05
06
07
10
11
12
13
14
15
16
17
20
33
34
35
36
37
40
41
74

-76
77

---01

---

-75
--

APPENDIX B
LOADING COS

The COS system is supplied on DECtape
(LINCtape for PDP-12 users)
diskettes and RK05s.
It is suggested that the user copy the COS tape
on a certified DECtape (or LINCtape) before using the system (Refer to
Chapter 5.) The copy can be used and the original stored in a safe
place in case part of the system should be inadvertently destroyed
during subsequent use.
The Monitor can be put into core memory in one of three ways.
If the
computer is equipped with the automatic load feature, the Monitor is
loaded by pressing the INITIALIZE switch under the table or by raising
the sw switch on the computer console.
Otherwise the Monitor must be loaded into memory using the appropriate
Bootstrap routine.
The editing scratch area is not
start-up.

automatically

erased

during

system

If the system is equipped with an Analex printer, the printer must
turned off until the system is bootstrapped and re-configured.

B.l

AUTOMATIC LOAD

If one of the optional automatic bootstraps
MI8-EC
MI8-ED
MI8-EH
MI8-E

for
for
for
for

the
the
the
the

TC08
RK8
TD8E
RF08

was purchased, push the SW switch from down position to
Then return it to the down position.

B.2

position.

the

following

TCOS DECTAPE AS THE SYSTEM DEVICE

If your system uses DECtape as the system device,
procedures:

B-1

use

Mount the COS DECtape (DEC-08-0COSA-D-UO) on DECtape unit O.

Set the LOCAL/OFF/REMOTE switch to REMOTE and the WRITE
ENABLE/WRITE LOCK switch to WRITE LOCK. Set SWITCH REGISTER
to zero (all switches down) and press EXTD ADDR LOAD.

Enter the binary equivalent of the following locations and
instructions
(refer
to
Chapter 1 of Introduction to
Programming for information on binary numbers). To bootstrap
a TC08 system:
LOCATION

INSTRUCTION

7613
7614
7615
7616
7617
7620
7621
7622
7623

6774
1222
6766
6771
5216
1223
5215
0600
0220

7754
7755

7577
7577

Set the SWITCH REGISTER (SR)
to enter the first location
(7613)
and press, ADDR LOAD.
Set the switches to enter the
first instruction and push up on DEPosit switch.
Continue
entering instructions as described up to and including
instruction 0220.
Then set 7754 in SR and press ADDR LOAD.
Set SR to 7577 and push up on DEP switch twice.
4.

Set terminal REM/LOC switch to REM.

Set SR to binary equivalent of 7613; press ADDR LOAD, CLEAR
and CONT keys in that order. The Monitor should respond with
the message
COS MONITOR 5.05 CURRENT VERSION
NUMBER)

Enter the DATE command as described in Chapter 2.
At this
point the SYSGEN program can be used to change the system
unit number assignment (refer to SYSGEN, Chapter 3).
Set the switch to WRITE ENABLE.
6.

B.2.1

If the Monitor does not 'respond correctly, start
and reload the bootstrap routine.

step

System Restart on DECtape

If the system halts or CTRL/C is inoperative, use the ,automatic
bootstrap switch to restart the system.
Each time the Monitor is
reloaded, the old date is erased and the Monitor requests a new date.
Use the DATE command to enter the new date.

B-2

For users without a bootstrap switch, set SR to 7600 and press LOAD
ADDR, CLEAR and CONT, assuming the core resident monitor has not been
destroyed.

B.3

DISK AS THE SYSTEM DEVICE

If your system uses an RK08, RK8E, or RF08 disk as the system device,
the programs must be loaded on the disk from the DECtape supplied.
1.

Follow steps 1-6 described in section 2.0
system device.

Enter the DATE command in response
message.
Then enter the command:

the

DECtape

Monitor's

DATE?

.RU SYSGEN/C
in response to the
following questions:

Monitor's

period.

SYSGEN

asks

the

DO YOU DESIRE TO CHANGE THE SYSTEM DEVICE?
Type YES and a CR key.
WHAT IS THE NEW SYSTEM DEVICE
Type the 3 character designation fo·r the disk
(DKn) and the CR key.

system

device

DO YOU WANT TO TRANSFE-R YOUR FILES?
Type YES and the CR key.
DO YOU HAVE AN ANELEX PRINTER?
Type YES and
the CR key and SYSGEN skips to the last
question.
Type .NO and the CR key and SYSGEN assumes an LP08
or LS8E printer and asks:
132 COLUMN PRINTER?
Answer YES and the CR key if there is a 132-column line
printer.
If NO is the answer, SYSGEN assumes an 80-column
printer.
IS EVERYTHING CORRECT?
Type YES or NO and the
questions are reasked.

key.

entered

the

The files are transferred to the disk specified. To operate from the
disk as the system device, refer to BOOT, Chapter 9, or restart the
system with one of the following bootstraps.

B-3

RK08 AND RK8E BOOTSTRAP ROUTINE
Location

RK08 Instruction

RK8E Instruction

0030
0031

6733
5031

6743
5031

Set the SWITCH REGISTER to 0030, press the ADDR LOAD, CLEAR, and
keys in that order.

CONT

NOTE
If a PDP-12 is being used, execute an I/O
PRESET in 8-mode before starting at location
30.
The Monitor should respond with the version number and the
message. Enter the DATE command as described in Chapter 2.

DATE?

RF08 BOOTSTRAP ROUTINE
Location

Instruction

7750
7751
7752
7753
7754

7600
6603
6622
5352
5752

Set SR to 7750; press ADDR LOAD, CLEAR, and CONT in that order.
The
Monitor should respond with the version number and the DATE? message.
Enter the DATE command as described in Chapter 2.

B.3.1

System Restart on Disk

If the system halts and CTRL/C is inoperative, use one of the
bootstrap routines to restart the system.
Each time the Monitor is
restarted the Monitor requests a new date. Use the DATE command to
enter the new date.
For users without the bootstrap switch if the core resident portion of
the Monitor
is intact, set SR to 7600 and press LOAD ADDR, CLEAR and
CONT.

B.4

TD8E BOOTSTRAP

If your system uses TD8E's use the following bootstrap routine to load
the COS software.

Location

Instr uc tion

7300
7301
7302
7303

1312
4312
4312
6773

B-4

7304
7305
7306
7307
7310
7311
7312
7313
7314
7315
7316
7317
7320
7321
7322
7323
7324
7325
7326
7327
7330
7331
7332

5303
6777
3726
2326
5303
5731
2000
1300
6774
6771
5315
6776
0332
1327
7640
5315
2321
5712
7354
7756
7747
7400
0077

Mount the COS 300 DECtape (DEC-8E-OCOSA-D-UO) on DECtape unit
o. Set the LOCK/OFF/REMOTE switch to REMOTE and the WRITE
ENABLE/WRITE LOCK switch to WRITE LOCK.
Set the switch
register (SR) to zero (all switches down) and press EXTD ADDR
LOAD.
Set the SR to the binary equivalent of the first location and
press ADDR LOAD. Set the switches to the binary equivalent
of the first instruction and push up on the DEP switch.
Continue entering the instructions as described, the location
is incremented automatically and need not be entered again.
2.

Set terminal REM/LaC switch to REM.

Set SR to binary equivalent of the first location;
press
ADDR LOAD, CLEAR and CaNT keys in that order. The Monitor
should respond with the message
COS MONITOR 5.05

(or current
version)

DATE?

WRITE ENABLE unit 0 and enter the DATE command
in Chapter 2.

If Monitor does not respond correctly start
reload the bootstrap.

step

B-5

described
1

and

B.5

PDP-12 USERS

For a PDP-12 with LINCtape,
1.

Set left switches to 0700.

Set right switches to 0000.

Press I/O PRESET with mode switch set to LINCmode.

Mount system tape
LOCKED, REMOTE.

Press DO.

When LINCtape stops moving, press the START 20 switch.

(DEC-12-0COSA-D-UO)

drive

WRITE

Monitor displays
COS MONITOR 5.05
DATE?

(or current version)

WRITE ENABLE unit 0 and enter the DATE command
in Chapter 2.

If Monitor does not respond correctly, start at
reload the system.

step,

Any LINCtapes to be used with the COS system
MARK12 (DEC-12-SE2E-UO) (P option, 129 words).

B-6 '

must

described

marked

and
with

B.6

RX8E USERS

If your system uses the RX8E as the system device, do the following:
MOUNTING THE ~X8E
1.

Push the top of red power button on front of machine so that
it stays In. Very shortly the cursor (flashing light) will
appear in the screen.

Remove "COS-3l0 OPERATING SYSTEM" diskette from its envelope.
(do not touch the brown surface which appears through holes
in cover) •

After lifting the left-hand door (drive 0) on the front of
the DS3l0, slide the COS-3l0 OPERATING SYSTEM diskette into
the slot.
(Close door over diskette - it will "click" when
closed properly).

Remove data diskette from its envelope and repeat
procedure with the right-hand slot (drive 1).

the

Push the top of the white START button on
let it go allowing it to rock out.

The screen now displays a number.
Respond by typing any
character.
This causes the DS310 to perform an automatic
self-test, following this test the operating system is loaded
into memory.
The screen now displays

front

above
machine

COS MONITOR 5.05
DATE?
In response, type the data in the following format
.DA MM/DD/YY
The total display on the screen is now:

o1 2 3 Z
COS MONITOR 5.05
DATE?
DA 12/31/75

The DS3l0 is now ready to run the program.
DISMOUNTING RX8E AND TURNING DS3l0 OFF.
When you have finished using the DS310, close all computer's files
according to the instructions furnished with the application program.
Then open the doors over the diskette and gently slide the diskettes
out from the drives.
Place the diskettes back in their envelopes so
that the labels can be seen.
Hold the diskettes to avoid contact with
the recording surface.
After the diskettes have been removed, push the bottom of
POWER button. The display on your screen will disappear.

B-7

the

red

APPENDIX C
SIZE OF CODE PRODUCED BY DIBOL COMPILER

DIBOL source statements are compiled by the DIBOL compiler and produce
code which can later be executed by the run time system (RSYS).
Memory is also set aside for the constants and variables that are used
by the program.
.
Each variable (field name) takes up as much memory (in bytes) as
specified in its data specification statement.
For example, a
variable defined by a 6D3 specification requires 18 (decimal) bytes of
storage. This is 9 words since a computer word consists of 2 bytes.
Variables defined in an overlay record share memory with the variables
in the record being overlaid.
Each RECORD statement requires one additional word of storage.
This
word is reserved at the beginning of the record for storing the COS
word count during I/O operations.
Each record begins on a word boundary (even byte address).
If
record has an odd length (in bytes), one byte of storage is wasted.

Each constant used in the procedure section of a DIBOL program
requires storage (in bytes) equal to the length of the constant. The
length of a decimal constant is equal to the number of digits in the
number, including leading zeros.
There are no negative constants.
The length of a character string is the number of characters in it.
Each distinct constant with a length of four or less characters or
digits appears only once in the space reserved for constants. Thus,
if the constant 32 appears three times in the program it will appear
only once in the reserved data area. However, constants larger than
four characters or digits require space each time they appear in the
program.
The code produced by statements in the procedure section must start on
a word boundary.
Each statement requires an overhead of one word.
Each statement label used requires one word.
Unlabeled statements
with line numbers 1000 more than the previous line number require one
additional word each.
The number of words of code generated
determined by the following rule:

C-l

expression

can

Add together the number of variables and constants used.

Add in the number of operators which appear
(do not
unary + signs).
The operators include +, -, I, *, #.

count

Add one for each subscript reference (be it of the form
or V(A,B».

V(A)

The following table shows how much code is used by various DIBOL
statements.
Jldenotes the number of words required by the expression
E.
No. of Words of
Code Generated

Statement·

y+x+l

ACCEPT (y,x)
CALL label
CHAIN chnum
DISPLAY (line,column,expr)
END [/list control]
FINI (channel)
FORMS (channel,skipcode)
GOTO label
GOTO (labell,. ~ . ,labeln) ,key
IF (exprl.rel.expr2)stmnt
INCR var
INIT (channel, dev)
IN IT (channel, dev,filnam[,unit])
ON ERROR label
PROC en] [II ist control]
READ (channel,record,number)
RETURN
START [/list control]
STOP
[NO] TRACE
TRAP
var=
var=expr
var=exprl,expr2
WRITE (channel,record,number)
XMIT (channel,record[,label])

chnum+l
line+column+expr+l
1

channel+l
channel+skipcode+l
1

key+n+2
exprl+expr2+3
var+l· - Channel+2
chann~1+3+filnam+unit

channel+number+record+l
1

1
1

2
var+l
var+expr+l*
var+exprl+expr2+1
charrnel+number+recotd+l
channel+record+2

For the statement marked with an asterisk (*) in the previous table,
subtract 1 if the principal operator of expr is binary + or - and if
both types are decimal.
For example,
D = 3+5 takes 4 words of storage, while
D = 3*5 takes 5 words.
Similarly,
D=3+4+5 takes 6 words while
D=3*(4+5) takes 7 words.
Additional space is also required by the symbol table.
This consists
of two words for each distinct variable, statement label, or constant
used.

C-2

APPENDIX D
DESIGNING A RECORD

The file design begins with the set of related information shown in
blank form.
It is a simplified form which could be processed to print
out customer invoices.
DATE
PART#
PART NAME
WHOLESALE PRICE
VENDOR#
This information is used in a BUILD control program such as described
in Chapter 6.
First, the file designer notes the type and size of
each field in this record. Numerical fields on which calculations
will be performed later by a program are defined as decimal fieldsi
other fields are defined as alphanumeric fields.
0010
0020
0030
0040
0050
0060

DEFINE
Fl, D6 iPART #
F2, A30 iPART NAME
F3, D7 iVENDOR #
F4, D6 iDATE
F5, D6 iWHOLESALE PRICE

The designer now assigns BUILD input line keywords. Here, he
group related items such as part #, name and wholesale price.
0100
0110
0120
0130
0140
0150
0160
0170
0180

should

INPUT
PART
Fl
F2
F5
DATE
F4
VENDOR 1
F3

A sample data entry form is now completed, showing the line names and
the size and type of each field.
This can be used for reference when
data is entered. When this record design has been proven, it will be
more efficient to print these line names on the forms which are used
to collect the original data.

D-1

DATE (D6) _ _ _ _.,........._
PART# (D6)
. NAME (A30) _ _ _ _ WHOLESALE PRICE
VENDOR# (D7) __
_________________
~

D-~

(D6) _ _ __

APPENDIX E
BUILD CHECKDIGIT FORMULA

In most applications involving identification numbers, each number may
be verified for accuracy by a checkdigit, which is essentially a
redundant digit added to the normal number.
The checkdigit is
determined by performing an arithmetic operation on the number in such
a way that the usual errors encountered in transcribing a number are
detected. The checkdigit is determined as follows:
1.

Start with a number without the checkdigit ... 5764.

Multiply the first digit and every other digit by 2 (left
right) •
10 12

Sum the digits in the resulting numbers and the digits
multiplied.
1+0+7+1+2+4 =15

not

Subtract sum from next higher number ending in zero.

Add checkdigit to the end of the number. 57645
(This is the correct checkdigit if the number is entered in a
04 field.)

20-15=5

Note that a checkdigit procedure is not completely error proof.
In
the example given above, 5764 or 5673 give the same checkdigit. It is
unlikely, however, that transpositions of this sort will ~ occur.
The
checkdigit does not guard against the possible assignment of an
incorrect but valid code, such as the assignment of a wrong valid
identification code to a customer.
If the number entered for a checkdigit calculation is shorter than the
field, the rightmost digit is used as the checkdigit and the remainder
of the number is right-justified and padded with zeroes on the left.
The zeroes are considered when the checkdigit formula is calculated.

E-l

APPENDIX F
HARDWARE DESCRIPTIONS

Hardware descriptions for both COS-300 and
follows:

F.l

COS -310

HARDWARE COMPONENTS FOR COS-300 ARE:
PDP-8/E

Computer with at least l6KB of core

TD8E

DECtape drives, (or TC08 drives)
RK08 (or RF08/RS08 disk)

VT05

Terminal (or LA30 DECwriter)

PC8-E

Paper tape Reader/Punch (optional)

CR8-E
CM8-E

Card reader (optional) or
Card reader (optional)

LP08

Line printer or

LS8-E

tine printer

LP05

300 lines per minute printer

F-l

systems

are

Figure F-l
DS 300 Hardware Configuration

F.2

HARDWARE COMPONENTS FOR COS-310 ARE:
PDP8/E
8M
8F
8A

COMPUTER WITH AT LEAST 16KB CORE.

RKOI
RK05

Disk

TD8E
TC08
RX8E

DECtape

VTOS
VTSO
LA30
LA36

Terminals

(mutually exclusive)

Printer

(mutually exclusive)

LS8-E
LA36
LPOS

(mutually exclusive)
(mutually exclusive)

NOTE
The central processor unit (CPU) of a
COS system is located under the disk
area of the system.
There are two
switches necessary for the operation of
COS system.
OFF/POWER Switch

In the OFF position the
switch
disconnects all
primary power
to
the
machine.
In
the
ON
position
all'
manual
controls are ~nabled and
primary computer power is
applied.

INITIALIZE Switch
(COS-300 only)

Push
this
switch
to
initialize
the
system Bootstrap Loader.

START swi tch

As above.

F-3

Figure F-2
DS 310 Hardware Configuration

F-4

TERMINALS

F.3

VT05 TERMINAL

Basically, the VT05 terminal (Figure F-3) consists of a
CRT display.

keyboard

and

The basic function of the keyboard is to provide a convenient, on-line
means of transmitting ASCII characters to the computer to be processed
and displayed. An ASCII code is transmitted to the computer each time
a key is depressed.
The VT05 display allows direct viewing of all displayable characters
contained in the VT05 character set that are transmitted from and/or
received by the VT05.
The VT05 displays up to 20 lines, 72 characters per line, or a maximum
of 1440 characters full screen.
A blinking cursor indicates the
position of the next character to be generated and moves to the right
automatically as each character is displayed.
The VT05 can be
operated in LOCAL or REMote mode.
In LOCAL mode the terminal
is off
line and there is no communication with the computer.
In REMote mode
(the normal operating mode) data is transmitted between the terminal
and the computer.
VT05 controls and keys and their respective functions are listed in
Table F-l.
Keys which are operative off-line only are not covered
here~
refer to the hardware manual for the VT05 for a description of
their function.

F-5

Figure F-3
VT05 Terminal

F-6

Table F-l
VT05 Controls
Control

Function

Power ON/OFF S

This switch is located on the right front of
the VT05.
When the ON/OFF switch is turned
ON, power is applied to the complete system
and
the VT05 display refresh memory is
cleared. After approximately one minute has
elapsed, the cursor
(bright dash)
should
appear in the upper left-hand corner
(HOME
position) of the screen.

REM/LOC Switch

This switch is located on the right front of
the VT05.
In LOCal mode, the terminal is
off-line and data can be typed at
the
keyboard but not entered into memory.
In
REMote mode, keyboard data is transmitted
from the VT05 to the processor.
NOTE
The CONTRAST, BRIGHTNESS, VERTICAL
and HORIZONTAL controls are adjusted
in the same way as those of
a
commercial television receiver.

CONTRAST Control

The CONTRAST control is located on
the
right-hand side of the VT05 and is used to
adjust for display contrast (clarity).

BRIGHTNESS Control

The BRIGHTNESS control is located on the
right-hand side of the VT05 and is used to
adjust for
CRT
display
brightness
or
intensity.

VERTICAL Sync Control

The VERTICAL sync control is located on the
right-hand side of the VT05 and is used to
properly synchronize the screen
in
the
vertical direction.

HORIZONTAL Sync Control

The HORIZONTAL sync control is located on the
right-hand side of the VT05 and is used to
adjust the
screen
picture
for
proper
synchronization in the horizontal direction.

Returns cursor to the left margin of screen.

Moves cursor down one line in same location.
If pressed when cursor is on the bottom line
(line 20), the display scrolls
(top line
disappears and rest of lines move up one line
position) •

F-7

Table F-l (Cant.)
VT05 Controls
Control
RUB OUT

Function
Works in REMote mode in conjunction with the
Monitor to erase characters.
Erases the
character above the cursor from memory and
echoes the deleted character on the screen,
i.e.,
KAYYAEY
rubouts

echo

TAB

Moves the cursor to the right to the next TAB
stop each time TAB is pressed.

ALT

Has no effect on the display and
is ignored
by the terminal.
It is included to provide
an alternate escape
character
(replaces
Carriage Return or Line Feed) for use in the
user program.

CTRL

Used simultaneously with character
keys to
perform special functions.
(Chapter 2, Table
2-1)

SHIFT

Produces the character in upper case portion
of the key pressed simultaneously with SHIFT,
e.g.,!
rather than 1.
Control characters
are "SHIFT inhibited"
and alpha characters
are permanently set to upper case.

SHIFT LOCK

Enables and holds the shift function or
releases it if already set. When enabled the
SHIFT LOCK key is lit.

Space

Produces a blank character position (wherever
the cursor
is located)
each time it is
pressed. Cursor moves one position to
the
right.

F.3.l

VTOS Start Up Procedures
1.

Press the OFF portion of the ON/OFF power switch.

For local
alignment
switch.

operation
purposes)

(used
press

For on-line operation,
REM/LaC switch.

for
training, maintenance and
the LOCal portion of the REM/LaC

press

the

REMote

portion

the

Press the ON portion of the ON/OFF power switch.
The
blinking cursor
(bright dash) appears in the HOME position
(the first character position of the top line), and a beep is
emitted.

Allow approximately one minute for the CRT filament warmup.

If the cursor does not appear as specified,
key.

F-B

press

the

HOME

If the cursor still does not appear, check the BRIGHTNESS
control to be sure it is not set too low. Contact the local
DEC Field Service Office if these steps do not make the
cursor appear.
6.

F.3.2

In REMote mode all keyboard and control functions are
operational and the VT05 is ready for on-line operation. To
verify correct system operation, press the various controls
and displayable' character keys individually and observe the
display.
In LOCal mode
all
keyboard
functions
are
operational. To verify this, press the keys individually and
observe the screen.

Additional Features:

The VT05 sounds a warning beep when the line reaches character
position 63.
This alerts the user that there are 8 positions before
the end of the line. The cursor moves to the right automatically as
each character is displayed until it reaches position 72.

F-9

F.4

VT50 TERMINAL

The VT50 terminal (figure F-4) is a low cost, high speed, alphanumeric
CRT display with a keyboard for data entry. The VT50 is quiet because
it needs no cooling fans.
The screen is designed for optimum
character definition:
it is recessed and tilted to avoid glare from
overhead lighting. Since no filter covers the screen, it can be
cleaned easily.
Operator interaction with the VT50 is based on the cursor: a flashing
underline that shows where the next character will appear. The
screen's 80 character per line format allows the exact display of
full-card images. Thus, the VT50 can take the place of a keypunch and
card reader for program entry.
The VT50 transmits data at a rate of 9600 baud. Twelve (12) lines of
information may be displayed at anyone time depending on the device
involved. When the bottom line on the screen is displayed and the
CURSOR
is
directed to move to the next line, the top line
automatically "scrolls" off the screen to allow space for the line at
the bottom.

., F-IO

Figure F-4
VT50 Terminal

F-ll

Table F-2
VT50 Controls
Control

Function

Adjust the controls on
the underside of the Unit

Access
to the controls
on the
underside is gained by tipping the
unit up on its rear.
The unit
should not be tilted forward since
this
will apply stress to the
controls which may damage the unit.
A label on the underside of the
unit contains a condensed form of
this information, and a diagram
which shows the relative location
of each of the controls. The two
rotary switches govern the mode and
rate of data transmission.

Plug the Unit in
and Apply Power

There are two cords corning from the
underside
of
the
unit:
an
input-output cable which is not
used in the VT50 test, and a power
cord, which should be plugged into
an outlet at this time. The ON/OFF
Switch is on the right side of the
unit. Turn the unit on by sliding
this switch toward the keyboard.

wait for the Cursor

After approximately one minute, you
should see a flashing underline,
called the cursor, at the upper
left corner of the screen. The
terminal uses the cursor to tell
you where on the screen you are
typing.

INTENSITY CONTROL

Adjust the intensity control for
the best image on the screen. This
control works like the brightness
control on a television set.

Returns cursor to the
of screen.

Moves cursor down one line in same
location.
If pressed when cursor
is on the bottom line
(line 12),
the
display
scrolls
(top line
disappears and rest of lines move
up one line position.).

DELETE

Works in REMote mode in conjunction
with
the
Monitor
to
erase
characters. Erases the character
above the cursor from memory and
echoes the deleted character on the
screen, i.e.,
KAYYAEY

left

ft~
~echo

rubouts

F-12

margin

Table F-2 (Cont.)
VT50 Controls
Control

F.4.1

Function

TAB

Moves the cursor to the right to
the next TAB stop each time TAB is
pressed.

BACKSPACE

moves Cursor toward left margin one
character at a time.

ESC (ESCAPE KEY)

See Table 2-1

BREAK

Not used with DS 310.
Used only
when
VT50
is
installed as a
stand-alone remote terminal.

CTRL

Used simultaneously with character
keys to perform special functions,
(Chapter 2, Table 2-1)

SHIFT

Produces the character
in upper
case portion of the key pressed
simultaneously with SHIFT, e.g.
!
rather
than 1. Control characters
are "SHIFT inhibited" and alpha
characters are permanently set to
upper case.

SHIFT LOCK

Enables
and
holds
the
shift
function or releases it if already
set. When enabled the SHIFT LOCK
key is lit.

Space

Produces a blank character position
(wherever the cursor
is located)
each time it is pressed.
Cursor
moves one position to the right.

Additional Features

The VT50 sounds a warning beep when the line reaches character
position 63.
This alerts the user that there are 8 positions before
the end of the line. The cursor moves to the right automatically as
each character is displayed until it reaches position 72.
The operator invokes the cursor control commands by transmitting from
the
keyboard
(to. the computer)
the editing commands that are
recognized by the host's system software.
It is the responsibility of
the system software to translate commands coming from the operator
into the proper cursor control commands recognized by the terminal
System software for host computers generally includes editing routines
that expect to receive their own unique editing commands from the
operator.
In such systems Escape Sequences received from the keyboard
might not be recognized and therefore could cause a completely
different action than desired.
For further information on the VT50 Terminal
TERMINAL USERS MANUAL (DEC-OO-HVTUA-A-D).

F-13

refer

VT50

VIDEO

CARD READER

F.S

CARD READER

There are two types of card readers, the CR8-E which reads l2-row, 80
column punched cards and khe CM8-E which reads l2-row, 40 column mark
sense cards or l2-row, 40 column punched data cards with timing marks.
The CR8-3 card reader
F-3.

switch~s

and indicators are described

Figure F-5
Card Reader

F-14

Table

Table F-3
Card Reader Switches and Indicators
switch/Indicator

Function

POWER (toggle circuit breaker and indicator)
All power
to the
card
controlled by this switch.

reader

STOP (momentary-action pushbutton/indicator switch)
Actuation of the STOP switch immediately
overrides the PICK COMMAND and negates
the READY status.
The card reader will
stop operation after the card currently
in the track is read completely.
Power
is not removed from the reader by this
action.
The red STOP indicator
is
illuminated as soon as the switch is
depressed.
RESET (momentary-action pushbuttom/indicator switch)
Actuation of the RESET switch clears all
error
logic
and
initializes
all
counters.
READ CHECK (indicator)

The READ CHECK alarm indicator denotes
that the card just read may be torn on
the leading or trailing edges or have
punches in the 0 or 8lst columns. A
READ CHECK error will cause the reader
to stop.

STACK CHECK (indicator)

The STACK CHECK alarm indicator signals
that a card has failed to read the read
station after a PICK COMMAND has been
received.

HOPPER CHECK (indicator)

The HOPPER CHECK alarm indicator denotes
that either the input hopper is empty or
the stacker is full.

F-15

LINE PRINTERS
F.6

LP08 LINE PRINTER

The switches and lights described in Table F-4 are part
line printer, Fiqure F-6.

Figure F-6
LP08 80-Column Line Printer

F-16

the

LP08

Table F-4
and Indicator Lights

Switch~s

Functions

Switch/Indicator
ON LINE/OFF LINE (switch)

Push up to put printer on line with
computer.

PAPER STEP (switch)

Push up to advance paper
disabled in on-line mode.

TOP OF FORM (switch)

Push up to advance paper to top-of-form;
disable in on-line mode.

ON LINE (indicator)

Lights when ON LINE/OFF LINE
set to ON LINE.

READY (indicator)

Lights when printer is ready to
computer output.

one

80- and l32-column versions.

F-17

line;

switch

The ON/OFF switch controls the power to the printer.
There are two LP08 line printers:

the

receive

F.7

LSB-E LINE PRINTER

The basic character set for the LSB-E (Figure F-7) contains the 10
n urn e ric dig its, 0 through 9; 26 upper case letters, A through Z; and
the 2B special char ac ter s shown below:

[
]

@
(
)

/
>
)

(space)
The LSB-E controls and their functions are shown in Table F-S.

Figure F-7
LSBE Printer

F-18

Table F-5
LS8-E Controls
Control

Function

ON/OFF (Switch)

Provides power to the printer and lights
when in the ON position.

SELECT (switch)

Enables and halts printer and lights
the ON position.

TOP OF FORM (switch)

Moves line printer paper to top of sheet
of proper location of the vertical forms
control tape
(Refer
to
Centronics
Manual) when installing forms.

FORMS OVERRIDE (switch)

Allows completion of form being printed
even
though
paper
out switch was
activated.

HARDWARE ALARM (light)

Signals when head carrier has exceeded
the right margin stop. An audio alarm
also sounds.
If this should happen,
reset the printer and restart the job.

PAPER OUT (light)

Indicates printer is out of paper or
paper handling malfunction. Correct the
problem and restart the job.

MANUAL CONTROLS
Paper Advance Knob

of
carriage;
Located on left side
provides
manual forward and reverse
paper positioning.

Paper Feed Level
(older models only)

Located on rear, right top of printer;
provides initial paper
feed
around
platen to allow positioning of paper in
the tractor feed mechanism.

Forms Thickness Control

Located on inside of printer, this dial
with
graduated
markings
provides
adjustment of clearance between platen
and face of printer head. Adjustment
must be made according to the thickness
of the forms being used.

F-19

F.8

LPOS LINE PRINTER

The LPOS line printer is a high quality drum printer.
The basic
printer contains the following: 132 printing columns; 64 character
drum; paper advance mechanism; top of form control;
self testing
capability;
variable form length switch (11 positions); a static
eliminator; a paper receptacle, and the necessary interface and
control electronics.

Figure F-8
LPOS

F-20

F.B.l

Operator Controls

The Operator Controls are Separated into three groups.
Those most
frequently used are located on the operators panel. The less used
controls are located internally.
Those directly related to forms
positioning are on the printer mechanism.
Table F-6
LP05 Controls
Controls

Function

Power On

An indicator that is illuminated when
primary power and DC voltages are ON in
the printer.

Ready

An indicator that illuminates
interlocks
are closed and
conditions exist.

On Line

A switch/indicator that is illuminated
when
the
printer is in the Ready
condition and the ON LINE switch has
been actuated. The interface is active
only
in
an
ON
LINE
condition.
Depressing that switch will alternately
place the printer ON LINE and Off line.

Paper Step

A momentary switch that can be used to
advance the form one line if the printer
is in an OFF LINE condition.

variable Forms

A switch with 11 positions for
form length.

Top of Form

A momentary switch that can be used to
advance the form to the next page. This
switch is disabled if the printer is ON
LINE.

Alarm/Clear

A switch/Indicator that illuminates on
the occurrence of a fault condition or
when the print inhibit switch is in the
ON
position.
Depressing the Alarm
switch Master clears the printer logic.

Drum Gate

An Indicator that illuminates
drum gate is unlatched.

Paper Fault

An
Indicator
that
illuminates
to
indicate an out of paper, torn paper,
and runaway condition.

Ribbon Fault

An
Indicator
that
indicate
a
ribbon
failure.

Hammer Fault

An
Indicator
that
illuminates
to
indicate a hammer fault has occurred.

F-21

when all
no fault

variable

when

the

illuminates
to
jam or reversal

Table F-6 (Cant.)
LPOS Controls
Controls

Function

Format Fault

Illuminates in the event of a loss of
synchronization in the basic printer
paper feed system or a VFU parity error
has been detected.

Tape Faul t

An Indicator that illuminates when DAVFU
is not loaded.

Copies Control

A knob that varies the distance from
drum tb hammer bank. This control is
used to compensate for various forms
thickness.

Drum Gate Latch

A lever that allows the drum gate to be
opened for access to the print station.
An unlatched condition stops the drum
motor, clears the printer logic and
inhibits
the
READY
and
ON
LINE
cond i tions.

vertical Positioning

Two controls; one to move the form up
or down at the print station for primary
forms alignment, the second to provide
file adjustment while printing.

Horizontal Positioning

File forms alignment is provided by a
control knob that horizontally positions
the forms by moving both tractors in
synchronism.

F-22

DECWRITERS
F.9

LA30 DECWRITER DATA TERMINAL

The LA30 DECwriter (Figure F-9) is a dot matrix impact printer and
keyboard for use as a hard copy I/O terminal. The unit can be
interfaced with all DEC PDP product lines.
It is capable of printing
a set of 64 ASCII characters at a speed of up to 30 characters per
second on a sprocket-fed 9-7/8 in. wide continuous form.
Data entry
is made via a keyboard capable of generating either 97 or 128
characters.

Figure F-9
LA30 DECwriter
F-23

LA30 CONTROLS AND INDICATORS
Table F-7
LA30 Controls and Indicators
Function

Control/Indicator
READY

Lamp - Indicates power up on printer
electronics and that the DECwriter is
READY for use.
Indicates input
is
enabled by keyboard electronics, if. INT
bit is set by software.

LOCAL LINE FEED

Switch - When depressed, causes a local
line feed to be applied to the printer
without a code being sent out to the
computer.
This
control
will also
disrupt printing, but no characters will
be lost.

MODE LOCAL LINE

2-Position Switch - Selects either local
or on-line operation.

BAUD RATE 110, 150, 300

3-Position Switch - Selects the baud
rate clock frequencies for 110, 150, and
300 baud.

MOTOR POWER

Breaker (CB2) - Applies power to
stepping motor electronics.

printer

AC POWER

Breaker (CBl) - Applies ac power
unit power supply.

the

time,

F.9.l

Operating Procedures

F.9.l.l Loading Paper - When loading paper for
replenishing the supply, proceed as follows:

the

first

Remove main power at CBl.

Open the cover of the DECwriter by pulling up on
of the front.

both

sides

NOTE
When opening the top cover, always release both ball
studs from their spring retainers at approximately
the same time to avoid skewing the top cover.
3.

Raise the cam lever located on the left-hand side of the
print bar assembly until it disengages and slide the assembly
back.

F-24

Feed the paper from its box on the floor under the DECwriter
up through the opening in the bottom of the base casing.
Pull the fresh supply into the machine and discard any
remnants of the old supply.

Make certain that the paper is feeding straight and re-engage
the paper with its sprocket tractor pins.

Pull the assembly forward and re-engage it by pressing the
cam lever down until it locks in place on its retainer stud.
Adjust tractor width, using vernier knob concentric with
right-hand drive sprocket. Tractor pins should not elongate
holes in paper.

Advance the paper by rolling the knobs.
Make certain that
the paper indexes crisply from one line position to the next.

Feed end of paper through the smaller hole in the cover and
close the cover, latching both ball studs to the base
assembly.

Re-apply main power at CBI.

F-25

Table F-8
LA30 DECwriter Specifications
Description

Specification

asynchronous.
30
300 ms carriage

Printing Speed

30 character/second,
line
feeds/second,
return

Line Length

80 character positions

Character Spacing

10 characters/in.

Line Spacing

6 lines/in.

Paper

9-7/8 in. wide continuous form, tractor
dr i ven.
(1/2 in.
pi tch X 9-3/8 in.
wide X 0.150 in. diameter feed holes)

Copies

One part:
Two par t:

Ribbon

4 mil Nylon, 1/2-in.
inking

Typeface

5 X 7 dot matrix

12 to 20 lb paper
12 to 13 lb paper
7 to 7-1/2 lb carbon

X 120

ft,

Printing Character 64 upper case ASCII subset (lower case codes
in upper case)

medium

Keyboard Characters

97 or 128 (switch selectable)

Code

USASCII - 1968

Interface

Available
for
all
DEC
Computers.
TTL-compatible, parallel (17 pairs plus
1 spare pair in cable).
Interface is
ASCII encoded. The Serial Interface is
an extra cost option. Switch selectable
110, 150, or 300 baud. Four-wire 20 rnA
current loop interface compatible with
all DEC computers or EIA RS232C.

Temperature

50 F (10 C) - 122 F (50 C)

Humidity

5 - 90% (non-condensing)

Dimensions
Depth
Width
Height

24 in.
(o.6lm)
20-1/2 i.
(o.5m)
31 in.
(0. 79m)

Weight

110 lb (50 kg)

DC Power Supply

Self-contained (DEC Type H735)

Power Input

115/230 Vac + 10%
50/60 Hz
300W, maximum

F-26

F.IO

LA36 DECWRITER II DATA TERMINAL

The DECwriter II (LA36) is a fast, small, quiet, 300-baud keyboard
printer for use as a hard-copy terminal in remote or local I/O
applications. The LA36 is a serial transfer, asynchronous device
offering improved throughput reliability, good print quality and form
handling versatility in local operation.
The LA36 functions as a
typewriter allowing operators to add information to the print out.

Figure F-IO
LA36 DECwriter

F-27

Table F-9
LA36 Controls and Indicators
Control/Indicator

Function

Power On-Off

Applies and removes AC power
machine

Line/Local

Selects on-line or local operation

Baud Rate-llO,150,300

3-position switch selects the baud rate
clock frequency for communications line
operation

Forms thickness adjustment

Located on right side of print head
carriage.
Selects
proper
gap for
I-through 6-part form. Approximately 1
click for each part.

Right tractor adjustment

Thumb screw may be
movement of right
forms widths

Fine vertical tractor releasE

Line feed knob may be pressed inward and
rotated'in the appropriate direction for
precise location
of
printing
with
respect to vertical zones

entire

loosened to allow
tractor for various

Table F-IO
Specifications
Printing

True 30-cps throughput
with
catch-up mode 132-column format
Spacing
10-inch horizontal,
vertical

50-cps
6-inch

Characters

Full 96 character upper/lower case ASCII
7X7 dot (wire) matrix (0.070 X 0.100
inches)

Paper

variable width, 3 to 14 7/8 inches
Up to 6-part forms, 20 mils maximum pack
thickness
Tractor drive, pin feed

Keyboard

Standard ANSI
roll-over

Interface

Integrated 20mA current loop

Ribbon

DIGITAL-specified Nylon fabric, spool
assembly, 0.5 inches wide C 40 yards

Power Requirements

90-132 Vac or 180-264 Vac, 47-63 Hz
300 W maximum
(printing) ,
160
non-printing

Operating 10 C (50 F) to 40 C (104 F)
Non-operating: -40 C (-40 F) to 66
(151 F)

Temperature

F-28

keyboard

layout,

N-key

Table F-IO (Cant.)
Specifications
Humidity

Operating:
10% to 90% with max.
wet
bulb 28 C (82 F) and minimum dew point
2 C (36 F) non-condensing
Non-operating:
5% to
95%
relative
humidity non-condensing

Dimensions

27.5 inches wide, 33.2 inches deep

Weight

102 pounds

F-29

READER PUNCH

F.ll

HIGH-SPEED PAPER TAPE .READER AND PUNCH UNIT

The high-speed paper tape reader and punch unit shown in Figure F-ll
performs paper tape inp~t and output.
The high-speed paper tape
reader is used to input data into core memory from eight-channel,
fan-folded
(non-oiled), perforated paper tape.
The reader inputs
information photoelectrically at a rate of 300 characters per second.

Figure F-ll
High-Speed Paper Tape Unit

F-30

The reader controls are shown in Figure F-ll and
F-ll.

described

Table

Table F-ll
High-Speed Reader Controls
Control

Function

ON/OFF

Press down on the ON portion of this
switch to supply power to the reader
unit.
The
OFF
position
setting
disconnects the reader power supply.

FEED

Press the right side of this switch to
advance tape through the reader without
recording by the photoelectric sensors.

Control Knob

Turn this knob to the left to raise the
tape retainer lever so paper tape can be
inserted or removed. Turn the knob to
the right to lower the lever to hold the
paper tape over the sprocket wheel.

The reader is controlled by the computer, although the operator may
indirectly contol the reader from the keyboard through the computer.

F.ll.1

Papertape Positioning

Paper tapes are manually positioned in the high-speed reader with
following steps.

the

Turn the control knob to raise the tape retainer lever.

Place the paper tape in the right-hand bin so
of the tape passes over the sensors first.

Place several folds of leader tape in the left-hand bin
the tape passing over the sprocket wheel.

Turn the control knob to lower the retainer lever over the
tape until the feed holes are engaged in the teeth of the
sprocket wheel.

Press the tape FEED button until
reader head.

Tape is
advanced
instructions.

and

read

the

leader

tape

programmed

beginning

over

with

the

computer

Once the paper tape has been properly placed in the reader and the
leader/trailer has been positioned as outlined in the preceding steps,
the tape is normally read under control of system software.
The high-speed paper tape punch is used to record computer output on
eight-channel, fan-folded paper tape at 50 characters per second. All
characters are punched under program control from the computer.
The punch controls are shown in Figure F-Il and are described in Table
F-12.

F-3l

Table F-l2
High-Speed Punch Controls
Function

Control
ON/OFF

Press down on the ON portion of this
switch to supply power to the punch
position
setting
unit.
The
OFF
disconnects the punch power supply.

FEED

Press the right side of this switch to
advance feed-hole-only punched tape for
leader/trailer purposes.

F-32

DECTAPE

F.12

DECTAPE TRANSPORT UNIT

The DECtape transport unit shown in Figure F-12 applies to both the
TD8E and the TC08.
This unit is a bidirectional magnetic tape
transport which reads and writes the lO-channel magnetic tape.
Tape
movement can be controlled by programmed instructions from the
computer or by the manual operation of switches located on the front
panel of the transport.
Data is transferred only under program
control.

Figure F-12
DECtape Transport Unit

F-33

The transport controls are described in Table F-13.
Table F-13
DECtape Controls
Function

Control
REMOTE

This switch position
energizes
the
DECtape transport and places it under
program control.
The REMOTE
SELECT
light comes on whenever the unit is
selected for use.

OFF

This switch
position
DECtape transport.

LOCAL

This switch position
energizes
the
DECtape transport and places it under
operator control from external transport
switches.

WRITE ENABLE

This switch position enables the DECtape
for search, read, and write activities.
The WRITE light comes on when the switch
is put in the WRITE ENABLE position.

WRITE LOCK

This. switch position limits the DECtape
transport to search and read activities
only.
(This
prevents
accidental
destruction of permanent data.)

unit Selector

The
value
specified
by
this
eight-position dial (0-7) identifies the
transport to the control unit.

disables

the

On TD8E devices, the first controlier is
hardwired to accept only 0 or 1 and the
second controller only 2 or 3.
with the transport in
LOCAL
mode,
depressing this switch causes the tape
to feed onto the right-hand spool.
with the transport in
LOCAL
mode,
depressing this switch causes the tape
to feed on the left-hand spool.

F.12.1

DECtapes are mounted on the transport unit as follows:

Set REMOTE/OFF/LOCAL switch to OFF.

Place DECtape on left spindle with DECtape label out.

Wind four turns on empty DECtape reel on right spindle.

Set switch to LOCAL.

Depress the switch for a few seconds to
properly mounted.

F-34

make

sure

tape

Dial correct unit number on unit selector.

Set switch to REMOTE.
LOCK setting.

Select either WRITE

F-35

ENABLE

WRITE

DISK

F.13

RK08/RKOl DISK CARTRIDGE SYSTEM

The RKOl Disk Drive and Control contains the drive electronics and
mechanism for accepting and releasing the disk, positioning the
read/write heads, and controlling reading and writing of data from
RK08 commands.
The RK08 interfaces up to four RKOl Disk Drive and
control units.
Each RKOl Disk Cartridge stores 831,488 12-bit words
(including three spare tracks).
Four RKOl drives controlled by one
RK08 have a total of 3,325,952 words.

Figure F-13
RK08 Disk
The switches, buttons and lights described in Table F-14 control the
operation of the RK08 Disk.
(See Figure F-13.) An RK8E disk is also
available.

F-36

Table F-14
RK08 Controls and Indicators
Function

Control

ON (pushbutton and indicator) Push ON button to supply power to the
disk.
ON button lights when power is
on.
OFF (pushbutton and indicator)
Push OFF button to halt the power supply
to the disk.
OFF button lights when
power is off.
ATTN (pushbutton and indicator)
Indicates that the disk power supply has
a malfunction. Push the STOP switch to
reset the power supply.
NOTE
This button also lights when
system power is turned off with
the computer
ON/OFF
switch.
Push this button to reset the
power
supply.
This
inter-locking
action protects
the data on the
disk
from
spurious
write signals while
computer and disk controls are
powered down.
DISK LOCKOUT 0-3 (pushbutton and indicator)
Press the appropriate buttons
(0-3)
to
lock out the corresponding disk so data
cannot be written onto the disk platter.
The button lights when pressed. The
panel controls up to four disks.
SECT PROT

Prevents writing on protected sectors,
as specified by bit 0 of the second
header word of the disk sectors.
Press
to
protect
sectors
of
the
disk
previously coded for sector protect by
setting bit 0 of the sectors 2nd header
word to 1.
(not implemented in COS)

START/STOP (switch)

Place in the START or up position to
initiate the disk starting cycle (only
if the disk cartridge is in place and
the disk cartridge receive handle is in
the raised position, to initiate the
disk STOP sequence. Switch must be in
the STOP position to mount or dismount
the disk cartridge.

SAFE (light)

The SAFE light comes on approximately 35
seconds after a STOP sequence has been
initiated. When SAFE is lit, the disk

F-37

RK08 Controls and Indicators
Function

Control

cartridge access door interlock is off,
and it is safe to lower the
disk
cartridge
receiver
handle
for the
removal or insertion
of
the
disk
car tr idge.
0,1,2 or 3
(unit #) (light)

F.13.l

F.13.l.l

The READY light (left-hand light) comes
on approximately 90 seconds after the
START sequence has been initiated. When
READY is lit the disk is ready to begin
comm~nicating with the computer.

Mounting and Dismounting the RK08 Disk Cartridge

Mounting

Follow the steps below to initiate operation of the RK08 disk.
The DC power must always be ON before inserting a disk cartridge into
the disk. If the disk cartridge is inserted without first turning the
DC power on, the READ/WRITE head carriage assembly may be in the way.
1.

Press the disk power supply DC ON switch.
on.

The ON light comes

When the SAFE lamp, on the front control panel of the disk,
comes on and the handler lock unlatches, the disk is ready to
receive the disk cartridge.

Pull open the disk cartridge access door panel.
The
cartridge re~eiver is now raised into position for insertion
of the disk cartridge. The handle interlock will keep the
cartridge receiver handle locked in its close position
whenever the SAFE lamp is turned off.

Slowly slide the disk cartridge into the receiver assembly
and make certain there is no internal resistance to its
insertion.
NOTE

If there is the slightest resistance to the insertion
of the disk cartridge into the disk receiver, the
cartridge should be removed and the interior of the
disk receiver examined to determine the cause of
resistance.
5.

Raise the disk cartridge handle to the up or loaded position.
Do not force the handle.

Close the disk cartridge access door.

F-38

NOTE
An interlock switch is activated when the disk
cartridge is properly inserted.
However, if the
switch is not activated, the drive is not put in the
ready condition.
If the switch should transfer to
its inoperative position during an otherwise normal
operation, the disk will immediately stop and the
SAFE light will light.
7.

Set the START/STOP switch to the START or up position to put
the
disk drive system into operation.
A disk handle
interlock activates to ensure that the handle cannot be
raised during normal operation.

After a 90 second delay the READY lamp comes
that it is ready for actual operation.

F.13.l.2

on,

indicating

Dismounting

Follow these steps to remove the disk cartridge.
1.

Make certain the DC power is ON.

Set the START/STOP switch to the STOP or down position.
disk decelerates and the SAFE light comes on.

Open the disk cartridge access door and
receiver handle down.

pull

the

cartridge

Remove the
assembly.

cartridge

receiver

Close the disk cartridge access door.

disk

cartridge

F-39

from

the

The

F.14

RK8E/RK05 DISK CARTRIDGE SYSTEM

The RK8-E consists of a single-cycle data break control and one RK05
DECpack drive.
The Modular design permits computing power to be
easily expanded since one RK8-E control can support four
RK05
DECdrives. Each RK05 contains its own head position control and power
supply, so that expansion doesn't require the purchase of additional
equipment.
The RK05 DECpack Drive is a random-access, high density, rotating
memory with movable read/write heads and a removable disk cartridge.
Each RK05 provides 1.6 million words of storage, giving a fully
expanded RK8-E disk system over 6 million words of information and
offering the PDP-8 user an economical alternative to magnetic tape.

Figure F-14
RK8E/RK05

F-40

F.14.l

F.14.l.l

Mounting and Dismounting the RK8E Disk Cartridge

Mounting

The following is the correct cartridge mounting procedure for the RK05
disk drive.
Any deviation encountered during this procedure will be
considered an error condition.
1.

Set switch labeled RUN/LOAD to the LOAD position.

Verify that light labeled PWR is on.

Wait for light labeled LOAD to come on.

Verify that lights labeled RDY, ON CYL, FAULT, WT, and RD are
off.

Open access door.

Insert cartridge.

Close access door.

Set switch labeled RUN/LOAD to the RUN position.

Wait for lights labeled RDY and ON CYL to come on.

10.

Press switch labeled WR PROT and
labeled WT PROT goes on and off.

11.

Press switch labeled WT PROT until light labeled WT PROT goes
off.

12.

Verify that lights labeled FAULT, WT, and LOAD are off.

F.14.l.2

verify

that

the

light

Dismounting

Press switch labeled WT PROT and wait unti) light labeled
PROT goes on.

Set switch labeled RUN/LOAD to the LOAD position.

Wait for light labeled LOAD to come on.

Open access door and remove cartridge.

Close access door.

F-41

F.lS

RX8E/RXOI DISKETTE

The RXOI Diskette is a random access, mass storage device designed to
permit high-speed data transfer with a minimum of hardware interrupts
and a low software overhead. A sector buffer within the RXOI allows
asynchronous transmissions between the drive(s) and computer.
Data are recorded on one-side of the floppy disk only.
This surface
is divided into 77 tracks. ~ach track has 26 sectors each of which
contains a header field.

Figure F-15
RX8E/RXOI

F-42

F.lS.l

F.lS.l.l

Mounting and Dismounting the RXOI Diskette

Mounting RXOI

Lift the door of one of the drives on the front of the DS310 and slide
the COS 310 floppy diskette into the slot. Close the door over the
disk - it will click when closed properly - the disk is now mounted.

F.lS.l.2

Dismounting RXOI

The RXOI Diskette is dismounted
withdrawing the diskette.

F-43

opening

the

drive

door

and

APPENDIX G
COS FILES

There are four types of files in the
binary,
system program, and data.
files have similar structure.
The
standard OS/8 SAVE format.

G.l

COS system:
source, compiler
Source, compiler binary and data
fourth,
system programs, uses

COS SOURCE FILES

All source files, whether control programs (e.g., for BUILD) or DIBOL
program sources must be input with the LN command or number commands.
This makes all source files look the same and makes them compatible
with COS.
The source file format is:
line
number

word
count n

n-l words, two -237 ASCII
characters per word

The first word contains the word count for that line.
as
n

It is

computed

= (# of characters on line+l)
--------------------------- +1
2

The second word is the text line number, 0000-7777, base 8 (0000-4095,
decimal) .
The third and successive words contain stripped -237 ASCII
packed two per word (e.g., AB is 4243).

G.2

characters

COS DATA FILES

The format of data files is exactly the same as
there are no line numbers.
A data file looks like:
word
count n

n words, two characters
per word

G-l

source

files

except

Word I of a data record contains its length in words.
The next and
successive words contain the text of the record, two stripped ASCII
characters per word.

G.3

COS DIBOL COMPILER BINARY

Although the interpretation of the contents is entirely different,
externally a file of compiler binary is structured exactly like a data
file.
That is, for each line of DIBOL source
program,
the
interpretive code for that line is stored as a word count followed by
the interpretive code to be used by RSYS.

G.4

COS SYSTEM PROGRAMS

All system programs (e.g., PIP, CaMP, BUILD, etc.) are stored in OS/8
SAVE format.
The first block of the file is a core control block
indicating where the rest of the blocks of the file are to be loaded.
Each successive block is a 256 word core image. See the OS/8 Software
Support Manual for details.

G.5

SYSTEM AND DATA TAPE FORMATS

COS puts a label on all tapes and disks.
This label occupies the
first 256 words of each device, of which four are the actual label,
one is the date, and the rest may be a bootstrap.
Files on DECtape do
not use the last two blocks of storage.
Data files are completely devoted to the storage of data, beginning at
block 1.
Each logical unit holds one data file only.
Labels on data
files are put on by the Monitor in conjunction with DIBOL programs or
by system programs such as SORT and BUILD.
System files have the COS Monitor, the system programs such as CaMP,
and a directory in OS/8 format.
(There is a discrepancy of two years
between directories printed by OS/8 and those printed by COS.)
System
files can also be used to save the sources of DIBOL programs, BUILD or
SORT control specifications, or compiled DIBOL programs, but no data
files.
Their bootstrap starts up the COS Monitor.
Figure G-I illustrates the layout of the Monitor portion of the system
device.
As noted in the figure, CaMP shoul9 be the first file in the
file area. The location of CaMP is particularly relevant when the
binary scratch area is to be expanded.

G-2

BLOCK ..

(OCTAL)

BOOTSTRAP

0
1