This is the RSX/IAS SIG Symposium Tape from the 1981
Spring DECUS meeting in Miami, Florida. The tape contains material submitted
by the user community to the SIG at the Spring DECUS meeting.
A very important submission was made to this tape. Ralph Stamerjohn
submitted a document that has an abstract for each program on all past
RSX/IAS sig tapes. In addition he supplied an index at the end so that
you can pick a subject and look to see which sig tape the program is on.
This is a very valuable tool and should be printed right away.
The programs on this tape are from user submissions. The DECUS
staff, the RSX/IAS SIG staff, and DEC are all in relative ignorance of the
contents of the tapes. No warranty of any kind is implied in the
distribution of these tapes. The programs may or may not be well documented,
they may or may not work, they may even crash your system. If you have a
problem with the contents of the tape,
contact the author of the program. Do not contact DECUS, DEC, or
the RSX/IAS SIG.
The tape contains approximately 48,690 blocks of software in 2891 files.
Since 48,690 blocks would require four 2400-foot reels of tape in FLX DOS format
at 800 bpi, we elected to distribute the tape in BRU format.
The distribution fits on one 2400-foot tape in BRU format at 1600 bpi.
The distribution at 800 bpi will be two 2400 foot reels. I intend to
use BRU with the length switch so that the first reel is only 2000 feet
long. This will leave librarians who have two tape drives and do reel to
reel copies plenty of spare room at the end of the first reel.
At the time of this writing I have not actually used the length switch,
I think it will look something like :
BRU /REW/DENS:800/VERIFY/LENGTH:2000 MM: DB0:
I think BIGTPC will correctly treat the two 800 bpi tapes as
separate entities and faithly reproduce them. However, I have not
tried it.
One other note on the distribution, I think the collection will
fit on an RK07 but one might have to use the /HEADERS switch on BRU.
If you are a VAX site and, for whatever reason, cannot read theh
contents of this tape (so graciously submitted by the hard-working
RSX/IAS SIG members), please contact Philip Cannon.
UIC account [300,1] contains several files of interest. The
file RSXSPR81.DIR contains a directory of the contents of the tape
(BRU does not produce nice directories). The file RSXSPR81.DOC contains
an abstract of the contents of the tape by UIC. The file
READ1ST.ALL contains a concatenated list of all the README.1ST files on the
Collection. The file SUBMIT.DOC contains the guidelines for submissions to
the RSX/IAS sig tape collection. This is must reading for everyone who
desires to submit a program to the SIG tape.
A copy of this letter appears in the file BEGINS81.TXT in
the same account. The file UICSETS81.CMD contains the UFD commands
to create all the needed UIC's on device XX:. Edit it to match
your needs before using BRU to extract the tape contents. Note that
a partial extraction can be achieved by only creating the desired
UIC's.
The UIC account [300,2] contains the program that is used to
copy this and other tapes, BIGTPC. This is a new version of TPC that
does not choke at the end of the copy (did not effect data). To use
TPC with the 1600 bpi distribution one needs a disk with at least
48,690 blocks of free space, not necessarily contiguous.
To read in a copy of the master tape onto your disk,
one enters the following command:
RUN BIGTPC
TPC> DN:RSXSPR81.BRU=MT:
To make a copy of the tape for someone, one enters the following
command line:
RUN BIGTPC
TPC> MT:=DN:RSXSPR81.BRU
The distribution is again being made through the SIG tree-structured
distribution system. A geographically oriented distribution scheme
is used, where each LUG makes a few copies of the tapes and sends
them on to other LUGS, who in turn make a few copies of the tape and send
them on to other LUGS, and so on. The SIG is relying on each node to
contact the next higher node for necessary tape-transfer arrangements.
Since this is a volunteer operation, sincere interest on the part of
soon-to-be participants is very helpful in assuring a timely distribution.
Remember since it is a volunteer operation, DECUS is NOT paying for
postage or free copies of magnetic tapes. Make arrangements so that this
does not cost anybody a lot of money or magnetic tapes.
The Tape Copy tree will be mailed in advance to all the nodes with
two post cards included. The first post card is to be mailed on receipt
of the tree, to acknowledge receipt. If there is some reason why
a node cannot participate in the distribution, please contact Philip
Cannon IMMEDIATELY. The second post card is to be returned when
the Tape Collection copy is received.
Now for a new project, we are in the beginning stages of
compiling a tape which contains the most popular programs on all
past sig tapes. This tape "best of the RSX/IAS sig tape" will be
kept updated. The problem is how do we determine the most popular
programs. If you have taken a program off the SIG tape and
used it, write me (Philip Cannon) a letter and let me know. This
be a start anyway. Let me know about things that did not work also.
Some programs to start with would be TECO,SRD (which one ???),
RUNOFF (whose mods ? or no more mods ?),Ralph's INDEX, and
Jim Downward's KMSKIT (includes CCL). Write and let me know what
other programs are being used and WHERE YOU FOUND THEM. Tell me
which tape and what uic.
Speaking of writing letters, if you use a program off the
SIG tape, take the time to write to the author and let him know.
If you find a bug or make an enhancement, write and let him know.
Don't call (we all get too many phone calls) and don't expect a bug
fix, but a little positive feedback can go a long way. Just think
we have over 200,000 blocks and 12,000 files. With a little
positive feedback we will all have to get 6250 bpi tape drives.
Anyway my thank you to all the people who have submitted to the sig
tape. You are the ones who have made it the huge success it is.
The following people burned the midnight oil to create the
1981 spring RSX/IAS sig tape:
Angel Li (generously provided the Collection assembly site)
James Brown,Joe Sventek,Wayne Graves,George Hamma,Philip Cannon
Glen Everhart,Jim Neeland,Tony Scandora, and Mark Weston
Respectfully submitted,
Philip Cannon
RSX/IAS SIG Librarian
Science Applications, Inc.
1211 West 22 nd street
suite 901
Oakbrook, Il.
60521
(312) 655-5960
THE CONTENTS OF THE SPRING 1981 RSX/IAS SIG TAPE
This document contains an annotated description of the contents
of the 1981 spring RSX/IAS SIG tape. This annotated directory is considerably
longer than past annotated directories, but hopefully it contains more
useful information. I made the annotated directory by concatinating all
the readme.1st files on the collection. I then did a massive editing job
to condense the material to a reasonable size. Well I failed, the size
is not reasonable, but I am out of time. I hope you find this document
informative, but not too cluttered. I recomend taking a weekend and reading
it front to back, making notes in the margin as you go. Well as always,
I have to try something new. If you have any comments or critisims
please write to me (don't call) an express your views. I do need some feedback
to indicate whether I should continue with this long format or go back to
a shorter format.
I would like to thank my wife Cathy, without her patience and
understanding this would not be possible.
respectively submitted,
Philip H. Cannon
RSX/IAS SIG librarian
Science Applications, Inc.
1211 west 22 nd street
Oakbrook, Il.
60521
**************************************************
NOW FOR THE USUAL DISCLAIMER
The programs on this tape are from user submissions. The DECUS
staff, the RSX/IAS SIG staff, and DEC are all in relative ignorance of the
contents of the tapes. No warranty of any kind is implied in the
distribution of these tapes. The programs may or may not be well documented,
they may or may not work, they may even crash your system. If you have a
problem with the contents of the tape, contact the
author of the program. Do not contact DECUS, DEC, or the RSX/IAS SIG.
**************************************************
A quick review of some of the material on the tape. Its just
a tiny glimpse of what is there.
1. TAPESB.MEM - an INDEX to all previous RSX/IAS sig tapes.
The document lists all uics on all past tapes and
has a brief abstract of what the programs are about.
At the end of the document is a index section. Here
you can look up by subject what you want and presto
you will be referred to all pages that have software
dealing with that topic.
You must print this (170 pages) right away, it is
terrific (thank you Ralph Stamerjohn). The file
is in [346,100].
2. SRD - several new versions of SRD appear on the tape.
see [300,57],[330,005],and [373,4].
3. DOB - new versions of the disassembler, one gives macro
output. see [300,23] and [373,23].
4. CCL - new version of CCL from Jim Downward [344,61],
plus enhancements to the old one. A recursive
CCL appears in [373,013].
5. NETWORK software from DEC. There are new RMDEMO pages,
(they don't need DECNET) and many other goodies.
see virtual devices (11M without a disk ?). [336,200]
6. RUNOFF - this version of runoff has a table of contents.
[300,101]
7. SOFTWARE TOOLS - a new version of software tools [307,30]
8. LUT - display of the logical unit table and open files
[300,70].
9. FORTRAN and F4P debuggers - FORTRAN debugger [312,315] and
F4P debugger [330,002].
10. DSRDMP - prints out the structures in 11M pool. [373,17]
11. VSDRV and SRDRV - core-to-core message drivers.
VSDRV [300,120] and SRDRV [307,034] (part of software tools)
12. RMSINT - a 200 page document on the internals of RMS [373,310].
13. CRASHW - a 100 page manual on analyzing crash dumps [346,101].
14. MUPGMS - multi-user programs (shared code) on 11M [301,60].
15. HELP - many new help files [301,54] and [373,302] (there is
a help file for TKB in 373,302).
16. DDT - problems with ODT because it takes too much space in your
task ? Well this version of DDT only needs 200 words in
your task !!! [312,315].
Now we start, uic by uic of what is out there. For your information
the collection contains approximately 48,000 blocks of material in 2900 files.
[300,001] Account contents:
This uic contains useful files that serve as a roadmap to the
tape.
BEGINS81.TXT - The first file on the tape you should print, it
explains how to extract material from the tape.
RSXSP81.DIR - A directory of the entire tape.
READ1ST.ALL - this file contains a concatination of all of the
README.1st files on the tape.
UICSETS81.CMD - a command file that creates the UFDs. You can edit it
to do a partial restore of the material on the tapes.
RSXSP81.DOC - an annotated directory that documents the material
on the tape. (this document )
SUBMIT.DOC - a guide for people who intend to make a submission
to the RSX/IAS SIG tape.
[300,002] Account contents:
BIGTPC - a program that is very useful in making a copy of the
tape. It will copy FLX,DSC, or BRU tapes.
[300,021] account contents:
RSX BASIC - Michael Reese Version
BASIC as implemented in the following files under RSX-11D is
a highly upgraded version of what used to be a DECUS library
program for DOS. Full FILES-11 I/O is supported (although
in a bastard sort of way). String functions and user de-
fined functions are much more flexible than in either the
original version or in DEC's BASIC-11. A number of defici-
encies and bugs in the original have been fixed. Multi-user
implementation is supported with separate pure and impure
areas. This implementation of BASIC is highly interpretive
and thus is slow for number crunching (that's why you have
FORTRAN). It is however, highly interactive and includes
special debugging commands: STEP, CON and SET TRACE.
OVERLAY and a data preserving CHAIN are also supported. A
clean "break" feature is implemented via the TT handler
(more below). In the latest version, a number of RSTS-like
features have been added including: virtual arrays, integer
and byte variables, continued lines and IF-THEN-ELSE.
[300,023] account contents:
This submission comprises a maintenance release of the REESE
Dis-assembler, (DOB.) This program can take object modules
or modules automatically extracted from object librarys and
create pseudo listings. These lisings can be edited back to
MACRO source code.
The main difference between this and other routines around,
(a la TASKER and DDT,) is that GLOBAL lables that are unde-
fined in the object code have their GLOBAL SYMBOL inserted
in the text. The end result, especially for EXEC type mo-
dules, is much clearer.
[300,026] account contents:
Following is a description of the modifications made to MCR
to allow DCL type commands on a small MCR type IAS system.
The essence of the change is:
1. Dec's MCR is modified by using PAT so that instead
of changing the task name from ...XXX to $$$XXX if
MCR gets an error of "TASK NOT INSTALLED" MCR
changes the task name to "...MAS".
2. The reese task "...MAS", receives the command line,
modifies the syntax, and sends it to the necessary
utility. For example, a command line of "USE 1,1"
has "SET /UIC=[" added to the front, and "]" added
to the back. The command is then sent to "SET".
3. If MASsage does not recognize one of the pseudo
commands, it instead trys to install "XXX.TSX" from
LB:[11,1]. If this is successful it sets the "Re-
move-on-exit" flag for the task, and spawns the
task. If unsuccessful, massage issues the normal
"TASK NOT INSTALLED" message.
4. Massage causes INS to run with a TI: of CO: so
users are not confused by any INS error messages if
they issue an incorrect command.
[300,057] account contents:
This is a re-issue of this SRD, which was originally on
the SanDiego Fall78 tape.
Please take note that this is *NOT* the same as the "working group" version.
(I am inclined to say that it works better, but since I never tried the
"working-group" version myself, I don't know from personal experience.
Others who have used both, though, tell me that this one is the better.)
Since the San Diego release, this has had some minor fixes and some small
modifications as requested by other users.
I have added some more switches, and included "grand-totals", for
wild-card directories.
The operation concerning omitting system directories has been changed
slightly ( /SY switch).
Included also in the release are 2 TECO macros, to do a mass-edit on
an SRD output file. In TECO, do EISTRIP$$ and EIGET$$
This is a vastly improved version of SRD, dated 6-DEC 1979.
It has been sporadically updated over the last year and a half, so long
that I forgot what it's original base was. I think it was the
version distributed with our IAS V 2.0 system.
The original SRD had a massive number of bugs, the worst one was
probably that it would loose files on a Write-back (but mostly only
on very large directories) !!!!!!!
This version has had all the bugs I could find fixed, plus a lot
of enhancements.
Some of the key things are:
Write-back works properly, and it also re-tries if it gets
an I/O error on the write.
It will dynamically expand it's core buffer (via EXTSK$ directive)
if the directory won't fit.
If the entire directory won't fit, it will do it in
chunks, each of which is sorted properly; along
with a warning message.
A trailer line for each directory is given, showing the number
of files in the UFD, and the number of files selected.
If a switch that causes the file header to be read is
specified (/MI, /FU, or a date switch), the number of
blocks used/allocated is also given.
A "nolist" switch is added, so you can get the above trailer
info, without getting a list of the individual files
(very handy for scanning an entire disk).
The user write-up has been re-written.
On date select, the "before" and "after" dates can now
both be specified.
The format of the output has been slightly modified so that
a simple TECO macro can make a PIP command file to
copy (or whatever) the selected files. Two macros to
do just that are also on this tape, in my UFD for
TECO macros. To use them, do:
SRD XXX.CMD=????
TEC XXX.CMD
*EISTRIP$$
*EIGET$$
* (any other edit you may want to do)
*EX$$
This SRD has been built and run on 11D 6.2, IAS V2.0, and 11M V3.1
[300,070] account contents:
LUT.MAC This program will display the Logical Unit Table
LUTASM.CMD of a running task including open file information.
LUTKB.CMD
FCB.MAC This program will display the the FCB chain of
FCBASM.CMD a disk device i.e. all open files on that device.
FCBTKB.CMD
WHE.MAC This program will display the trace back chain of a
WHEASM.CMD running task compiled with Fortran IV plus and /TR
WHETKB.CMD
TTPOOL.MAC This program will display the number of free buffers
TTPASM.CMD in the RSX-11M V3.2 full duplex terminal driver of
TTPTKB.CMD a running system.
[300,100] account contents:
This UIC contains system-management programs for RSX-11M V3.2
I put in SYSLOG (KMS-FUSION), but I mainly use it to count
user QIO's, and to keep the double -word time.
I fixed RMDEMO (DEC's copy for 3.2) to use this time counter
to display the system up time (RMD starts up on a dedicated
CRT when the system boots up). I also use the idle-time
count to compute and display the system percent idle time, which gets
displayed on line 1 of the RMD screen. Since we lost the rotating
lights with the remote front panel, this % display is a handy crutch
to see how loaded the system is.
OPERATOR THINGS
---------------
We do not have an operator, and there is a need for regular
non-privileged users to be able to do certain restricted privileged
things, so they don't have to bug me to do it. These are things like
changing the spooler form number for the LA-120, log off terminals,
etc. I set this up by having a special user "OPERATOR", who is
privileged, but gets a program run (MONOPR) when he logs on. This
program spawns "@OPERATOR.CMD/-LI", which I have asking questions, so
that anyone can do them, but only those things which I have set up the
command file for them to do. Needless to say, the password for
OPERATOR is known to everyone.
TERMINAL TIME-OUT (TT.KGB)
--------------------------
One feature that IAS had that we missed was that IAS would time out
the PDS prompt and log off the user after a while. Since we have many
more users than terminals, this would take care of someone who walked
away from his CRT and didn't come back for an hour, but of course, if
you just did a PIP, it wouldn't time out. I liked this featue so much
that I added a program to do almost the same thing, called "TT.KGB".
If a terminal is idle for 9 minutes, it blasts out a message asking if
anyone is still there. In another 30 seconds, it sends out another
one, and 30 seconds after that, forces a BYE (via RMC). I use the QIO
count that SYSLOG (KMS-FUSION) keeps in the
terminal UCB to decide if it is idle or not.
TIME-SCHEDULED SYSTEM TASKS
---------------------------
It is handy to have MCR commands kicked off periodically. For example, you
might want to clean-up your disks (delete .LST files, etc) once a day,
fire off programs at midnight, etc.
It is really yuchy to do this by running the task with a /RSI, and anyway
you can't give it an MCR command line if you do that.
Therefore, I wrote RECUR, a program which does just that.
SETPRV
------
It is handy to be able to do privileged operations without having to
log off & log back into a privileged account. The task SETPRV does this,
asking for the "magic word", and if you type in the right one, it
will do one MCR command with the terminal privileged, then set it back
to non-priv. If you want to stay priv, this command would be: ABO SETPRV.
RMD
---
The fixes to RMD are also here. The RMDRIV is the one that does the
one huge QIO (DEC/JAN 81 Dispatch), with a fix.
UFD
---
I have each user as an entire group by himself. So that user's can
seperate their stuff however they choose amoung UIC's, I put a fix
into UFD, which will let a non-priv user create a UFD on his
LOGIN SY:, with the same group-number and a higher member-number as
his LOGIN UIC.
For example:
User DR1:[21,1] can create DR1:[21,2] but not DR0:[21,anything]
User DR1:[25,100] can create DR1:[25,105] but not DR1:[25,10]
User DR0:[300,377] cannot create any UFD's
This is UFD.COR
CKP
---
This is a KMS program (fixed by me for virgin 3.2) to checkpoint out
all stopped tasks. I set it up to run every 5 minutes. This is seperately
on this tape.
MVI
---
Tells you who has mounted what volumes.
FRC
---
Force MCR command line to a terminal.
RMC
---
A KMS-FUSION program to do the same thing.
[300,101] account contents:
The following modifications have beene made to the standard
RUNOFF. This is a re-issue of the RUNOFF on the last SIG tape.
1. Change bar support. This allows modified areas of text
to be marked by a bar (| ASCII code 174 octal) on the
left hand side of the page.
2. .BLANK and .SKIP may use negative arguments. This allows
text to be moves some number of lines from the end of the
page.
3. A shift right switch /RI. This switch allows text to be
shifted right some number of columns.
4. Allowing tabs to be relative as well as absolute column
number.
5. New commands for emphasis. Dark (via overprinting) and
underlining, and commands to easily set the "underscore
lock".
6. Support to change character-size for an LA-120 instream.
7. Some misc fixes, to FIGURE, FOOTNOTE, and others
8. New switches to set the default margins and page size on
the command line, plus the ability to set some defaults
in the task-build command file.
9. FINALLY, after long at last!!!!!!! Table of Contents.
Table of Contents is generated by Header-level commands,
and is automatically generated at the end of the file.
10. The /PA switch now goes by absolute page number, so that
it is usable even on chapter-oriented files.
11. Special characters within a ".LITERAL" are treated as
they should be, literally.
12. The defaults for the output file have been changed
slightly, so that it is easier to use. Also, the output
device is attached (if it is record-oriented), so that ^O
will work.
13. The document has been fixed up & several corrections
made.
[300,102] account contents
This is some modifications to TECO version 36.
I plan to support these changes as long as I can, so if there is a new
version of TECO, I will modify these mods correspondingly. The only ones
liable to cause anyone else problems are the patches to OBJ modules. I use
DISOBJ to find out where the patches go.
They do the following:
1) On an EI command, if the device/UIC are omitted, it will
first look in the default device/UIC, and if not found, it will
then look in LB:[2,2].
We use this to keep a "library" UIC of handy TECO macros, including
a nifty TECO.TEC.
2) Some fixes to terminal I/O for RSX-11D and IAS.
3) When creating a file with an explicit version of ";1",
it will re-write it if it is already there. (I would prefer to
have a switch on the file-spec to specify /UP, but I didn't want to
add things to the external interface.
4) When doing a normal edit (EB), it creates the output file with
an initial allocation equal to the size of the input file. This
makes the file creation slightly faster.
5) Patch to SCREEN.OBJ and TECO.OBJ, to take better advantage
of "split-screen" mode.
The one in SCREEN will paint a blob for a space on a HAZELTINE (or
another CRT which doesn't have inverse video.
The one in TECO.OBJ turns off the automatic echoing of the current
text line on <LF> and <BS> if split-screen is on.
6) If invoked as MUN, TECO will lower it's priority after 30 seconds.
This will keep a long MUN (for example: MUN SEARCH) from screwing
up the other users.
7) Changed the "IMMEDIATE" mode operation slightly. This is
un-documented, but very handy. It lets you write a TECO macro to
get control when you use the alternate keypad, like the immediate
mode <LF> and <BS>.
This TECO macro is invoked by "n1,n2MA$$", where n1 & n2 tell
you which key was pressed.
I fixed it so that it would pass the entire 7-bit value to the macro
(if I want it stripped to the lower 5 bits I can do it in the
macro), and took out the echoing of the command.
8) To go along with this, I have a macro (KEYS.TES), which
has most of the useful functionality of VTEDIT, but which
is much, much faster and smaller. This works even on a tube (such
as a hazeltine) that doesn't have an alternate keypad, you just to
give it a double keystroke.
9) To go along with these, there is a TECO.TEC, which will load-up
KEYS, go into split-screen mode, and fly!
I fixed up CRTRUB.MAC, to add CRT support for a HAZELTINE 1500
(we have both VT-100's & HAZELTINES), and split-screen even works
on a tube that does not have it!!!!!!! (Albeit not as nice as on
a VT100).
10) The TECO.TEC supplied here set up these things, and if invoked
as TES, starts up in split screen mode, if as TEV, starts up
in VTEDIT mode. We keep this in LB:[2,2], so everyone gets started
with it, unless they over-ride it with their own TECO.TEC.
[300,103] account contents
Enhancements to some MCR commands.
I put these on my system (M3.2) along with the KMS-FUSION stuff.
Jim Downward has these changes and they might possibly be included in his
next release.
These, however, are totally independent of his stuff.
ATLOV--Changes to the "ACT" command so that it also tells you the state
of the task, without all the garbage that the full one gives.
Finally ACT gives you useful info (such as if your task is
out or not).
BYE -- Mod's to Downward's BYE. This does a spawn & wait to UPD (to avoid
a race condition with setting the logoff bit). It also spawns
"@LB:[1,5]SETTERM.CMD" to reset the terminal characteristics
when the user logs off.
HELLO --An enhancement to HELP, to give you a list of the available
help-file keywords at a particular qualification level. This is
only useful if you have HELP files, and Jim's are the only ones
that I know of. (Unless and until DEC puts them in the next release.
I know they have them, because our M+ system had over 6000!!!!!!
blocks of HELP files. Although in a lot of cases, Jim's are better.
DEC seems not to have heard of lower-case letters.)
MCRDIS - A fix to allow a time scheduled task (which by definition has
CO: as it's TI: device) to spawn tasks via MCR. My RECUR task
(supplied on another UIC) uses this.
[300,104] account contents
This is the BROOM program that I found on an earlier DECUS tape, modified
so it does only rational things. The old one did weird things and had
to be a privileged task, and it only worked on IAS.
[300,105] account contents
These are corrections/enhancements to two programs in the KMS-FUSION set.
Jim Downward has been given these and he might also put them in
his next release.
My system is a straight base-level 26, so his CKP did not work. Thes one
here does, and it does a good job, especially in kicking out
the que-manager, de-spoolers, and RMD if they happen to be in the middle
of GEN (which does happen when there are a lot uf users working.)
This CVL adds another switch, plus will let a non-privileged user read a
volume label of a disk he has allocated. When we switched from IAS to
M3.2, nobody could mount their RK-05 disks, because of not knowing the
volume labels.
[300,107] account contents
PULLED THIS TYPE OFF THE FALL 80 SIG TAPE [313,20]
I modified it to improve the logic for locating comments,
and fixed it so that it will type out control characters in a readable
format.
It goes into this mode by a switch, or if the file is ?.TEC or ?.TES.
This is really a nifty program.
[300,110] account contents
I got an older copy of TRUNC, which I fixed up so that it is better
that PIP /TR. I have set up the default file extend on my public
disks to 10 blocks (to get better performance for editting, etc), and
TRU runs every night to chop off all the wasted space.
This version of TRU handles wildcard UIC's, plus it preserves the file
creation & revision dates.
[300,120] account contents
THIS UIC CONTAINS MISCELLANEOUS RSX-11M TASKS THAT MAY BE OF SOME GENERAL
USEFULNESS. INCLUDED ARE:
CATEST -- A CAMAC TEST TASK DESIGNED FOR USE WITH ANY DRIVER COMPATIBLE
WITH PHIL CANNON'S 3912/3992 DRIVER FOR RSX-11M. THIS TASK
ALLOWS A LIST OF COMMANDS TO BE ENTERED AND EXECUTED, AND
PROVIDES ACCESS TO THE DRIVER'S LAM HANDLING, LOOPING, AND
CRATE CHANGE COMMANDS, AS WELL AS PROVIDING FOR THE SAVING AND
RESTORING OF COMMAND LISTS, AND PRODUCING A HARDCOPY RECORD OF
THE TEST RUN (BY CREATING A LISTING FILE). CATEST.TXT IS A
BRIEF COMMAND SUMMARY.
DNDISPLAY -- A PROGRAM THAT BREAKS UP THE DISK REGISTERS FOR THE XYLOGICS
DISK CONTROLLER INTO INDIVIDUAL FIELDS, FOR USE IN EXAMINING
LOGS OF DISK ERRORS.
LOGLST -- A PROGRAM THAT CAN BE USED IN CONJUNCTION WITH JIM DOWNWARD'S
SYSLOG PACKAGE, TO PRINT OUT THE CONTENTS OF THE LOG SORTED BY
ENTRY TYPE, WITH USER NAMES INSTEAD OF ACCOUNT NUMBERS.
NSPOOL -- A SMALL TASK THAT ENABLES UNPRIVILEGED USERS TO STOP AND START
THE SPOOLER. THIS TASK ACCEPTS A "SEND DATA" MESSAGE FROM ANY
TASK, AND RESPONDS TO THE COMMAND IN THE MESSAGE BY SPAWNING
COMMANDS THAT START OR STOP THE PRINT SPOOLER. IT THEN
UNSTOPS THE SENDING TASK, WHICH CAN GO AHEAD AND USE THE
PRINTER (E.G. FOR PLOTTING OR OTHER DIRECT OPERATIONS).
THE ONLY REQUIREMENT IS THAT NSPOOL RUN ON A LOGGED-ON AND
PRIVILEGED TERMINAL (IDEALLY ON CO:).
ANOTHER TASK, "UNSPOOL", IS INCLUDED. THIS TASK SENDS THE
START AND STOP SPOOLER MESSAGES TO NSPOOL, AND MAY BE USED
TO IMPLEMENT START/STOP SPOOLER CCL COMMANDS FOR UNPRIVILEGED
USERS.
NSPOOL ALSO CAUSES A "TIME" COMMAND TO EXECUTE ON CO:
EVERY NIGHT AT MIDNIGHT, THUS MARKING THE DATE WHEN THE
CONSOLE LOG WAS PRODUCED. (THIS FUNCTION WAS ADDED TO NSPOOL
BECAUSE IT HAPPENS TO ALWAYS BE ACTIVE BUT ALMOST ALWAYS
STOPPED, AND IT'S A VERY SMALL TASK ANYWAY.)
FINALLY, A FORTRAN SUBROUTINE "SPOOL" IS INCLUDED. THIS
SUBROUTINE ALSO SENDS THE START/STOP MESSAGES TO NSPOOL.
CALL SPOOL(0) STOPS THE SPOOLER, ASSIGNS LUN 6 TO LP:, AND
ATTACHES THE PRINTER. CALL SPOOL(1) (OR ANY OTHER NONZERO
VALUE) DETACHES THE PRINTER AND RESTARTS THE SPOOLER.
SORT -- A TASK THAT WILL SORT ANY FILE OF FIXED-LENGTH RECORDS (UP TO
NINE LINES PER RECORD) INTO ASCENDING OR DESCENDING ORDER ON
UP TO NINE FIXED-LENGTH, FIXED-POSITION ASCII KEYS. A MAXIMUM
OF ABOUT 300K BYTES OF DATA (ABOUT 600 BLOCKS) MAY BE SORTED,
AS THE PROGRAM USES WORK FILES TO STORE PARTS OF THE SORTED
FILE. A QUICKSORT METHOD IS USED.
TSKBUILD -- THIS IS A MODEL COMMAND FILE THAT CAN BE CUSTOMIZED TO PROVIDE
EASIER TASKBUILDING FOR PROGRAM DEVELOPMENT SYSTEMS. BY
SUPPLYING PARAMETERS ON THE COMMAND LINE (E.G. "NOF4P" FOR NO
F4P RESIDENT LIBRARY, "TEK" FOR INCLUDING TEKTRONIX PLOTTING
PACKAGE) THE USER CAN SPECIFY STANDARD LIBRARIES THAT ARE
REQUIRED, WITHOUT HAVING TO REMEMBER WHERE THEY ARE LOCATED
OR WHICH MODULES MUST ALWAYS BE INCLUDED. THE COMMAND FILE
PROMPTS FOR INPUT AS TKB WOULD (E.G. AFTER "/" IT PROMPTS FOR
OPTIONS), AND CREATES A TKB COMMAND FILE WITH VERSION 777,
WHICH IT SUBMITS TO TKB. THE COMMAND FILE CAN BE DELETED
AUTOMATICALLY AFTER IT IS USED, OR IT CAN BE KEPT FOR FUTURE
USE DIRECTLY WITH TKB (E.G. TKB @CMDFILE).
WHEN -- A SMALL TASK THAT ACCEPTS COMMANDS OF THE FORM:
WHEN <TASKNAME> <MCR COMMAND>
WAITS FOR <TASKNAME> TO FINISH, AND THEN SPAWNS THE COMMAND.
IT CAN ALSO BE INVOKED AS: WHEN <TASKNAME> IN WHICH CASE IT
PROMPTS FOR COMMANDS TO BE SPAWNED IN SEQUENCE. EMBEDDED
"WHEN" COMMANDS OF THE SECOND TYPE CAN BE USED TO WAIT FOR
OTHER TASKS TO FINISH. THIS ALLOWS YOU TO "TYPE AHEAD" TO
SOME EXTENT; E.G. YOU DON'T HAVE TO WAIT FOR A COMPILATION
TO FINISH TO TYPE "TKB XYZ=XYZ,[1,1]F4POTS/LB"; YOU CAN TYPE:
WHEN F4P TKB XYZ=XYZ,[1,1]F4POTS/LB
(NOTICE THAT THREE CHARACTER NAMES FOR ...-TASKS ARE ACCEPTED,
AND WILL BE HANDLED CORRECTLY AS LONG AS THE TASK IS ACTIVE
ON YOUR TERMINAL.) ADMITTEDLY, THINGS LIKE THIS CAN BE DONE
BETTER WITH COMMAND FILES, BUT SOME OF US ARE TOO LAZY TO
CREATE AND MAINTAIN COMMAND FILES FOR ALL THESE LITTLE THINGS!
(BE A LITTLE CAREFUL WHEN USING THIS TASK WITH CCL; IF YOU
DO A SEQUENCE OF COMMANDS LIKE THIS:
>MAC XYZ=XYZ
>WHEN MAC
WHEN>LBR XYZLIB/RP=XYZ
WHEN>WHEN LBR
WHEN>TKB @XYZBLD
WHEN>^Z
AND ...LBR IS NOT INSTALLED, BUT IS IMPLEMENTED VIA CCL, THEN
CCL MAY NOT START THE TASK BEFORE THE "WHEN LBR" EXECUTES;
SINCE LBR IS NOT ACTIVE, IT IS ASSUMED THAT IT HAS COMPLETED,
AND THE TKB @XYZBLD IS STARTED. ONE WAY THIS PROBLEM CAN BE
SOLVED IS TO DO A "WHEN CA." BEFORE THE "WHEN LBR".)
XPR -- THIS TASK IS DESIGNED FOR USE WITH A VERSATEC 1200A PRINTER/
PLOTTER (OR EQUIVALENT), AND WITH VERSATEC'S LP: DRIVER. THE
1200A PRINTS 132 COLUMNS ON PAPER THAT IS 11" WIDE AND 8.5"
HIGH, WHEREAS NORMAL PEOPLE ARE USED TO PAPER THAT IS 11" HIGH
AND 8.5" WIDE. THIS TASK USES THE PLOTTING CAPABILITIES OF
THE 1200A (200 DOTS PER INCH) TO PERFORM SOFTWARE CHARACTER
GENERATION, AND PRINT MORE READABLE DOCUMENTS. AS LONG AS
CHARACTERS ARE GENERATED BY SOFTWARE, IT WAS DECIDED TO LET
THE PROGRAM ALSO DO PROPORTIONAL SPACING, RIGHT MARGIN
ALIGNMENT, SPECIAL SYMBOLS, UNDERLINING, DOUBLE-WIDTH
CHARACTERS, AND DOUBLE-SPACING. 59 LINES OF 80 CHARACTERS CAN
BE PRINTED WITH "NORMAL" SPACING; LINES MAY BE SHORTER OR
LONGER WITH OTHER SPACING OPTIONS. AUXILIARY TASKS ARE
SUPPLIED TO BUILD THE CHARACTER SET DATABASE, AND TO CREATE
SPECIAL SYMBOL FILES FOR GENERATING MATH OR SCIENTIFIC
SYMBOLS, SUBSCRIPTS AND SUPERSCRIPTS, ETC.
VTEDITMIM -- THIS .TEC FILE, ALONG WITH TECOMIM.INI, HAS BEEN USED TO
IMPLEMENT A SCREEN-EDITING TECO SYSTEM FOR MIME-I TERMINALS.
THE MIME-I EMULATES THE DEC VT52, BUT HAS NO ALTERNATE
KEYPAD. CONTROL CHARACTERS AND ESCAPE SEQUENCES THAT ARE
EASY TO REMEMBER ARE USED TO IMPLEMENT EDITING COMMANDS.
THE FILE TECOMIM.TXT CONTAINS THE COMMAND LIST. TECOMIM.INI
IS USED WITH A SLIGHTLY MODIFIED TECO.TSK, WHICH PICKS UP
TECO.INI FROM A STANDARD UIC (E.G. [1,2]) INSTEAD OF FROM
THE USER'S UIC. IT PREPROCESSES THE COMMAND LINE, AND THEN
STARTS UP THE VTEDIT MACRO FILE.
VSDRV -- THIS IS AN RSX-11M V3.2 DRIVER THAT IMPLEMENTS SOME OF THE
FUNCTIONALITY OF THE "VARAIBLE SEND DATA" FUNCTION AVAILABLE
ONLY WITH M-PLUS. THE DRIVER MANAGES ANY NUMBER OF NAMED
QUEUES OF VARIABLE-LENGTH MESSAGES. FUNCTION IMPLEMENTED
INCLUDE: CREATE QUEUE, DELETE QUEUE, "WRITE" (SEND MESSAGE),
AND "READ" (RECEIVE MESSAGE). IT IS POSSIBLE TO EMULATE THE
RECEIVE-DATA AST BY PERFORMING A "READ AND WAIT" QIO WITH AN
AST SPECIFIED. ALL STORAGE IS ALLOCATED FROM A PRIVATE POOL
AREA WITHIN THE DRIVER, SO THAT ONLY THE NECESSARY QIO PACKETS
COME OUT OF EXECUTIVE POOL SPACE.
INCLUDED WITH VSDRV IS A UTILITY PROGRAM "VSUTIL", WHICH CAN
BE USED TO CREATE AND DELETE QUEUES, AND TO DISPLAY THE
CURRENT CONTENTS OF THE DRIVER'S POOL.
[300,210] account contents
The following is a description of the Michael Reese Medical
Center RUNOFF oriented submission to the SPRING 81 DECUS
lug.
1. Reese RUNOFF.
Although this departs from the RUNOFF SIG'S attempt
to cut down on the profligate versions of RUNOFF
floating around, this has some features which I
think are of interest to users, and inclusion in
the RUNOFF sig version would probably take too
long. In addition, some of the documentation for
REESE programs takes advantage of these features,
and would not work on standard RUNOFF. Changes
made include:
1. Support for DIABLO type printers, including
1. Text enclosed in "{}"'s is printed in RED.
3. Capablility of inserting any octal char-
acter in the source, for things like half
line-feed's etc.
2. Seperate margin commands that only affect the
HEADER lines, so that they don't move if you
temporarily change your margins.
3. Output is standard "record" oriented, rather
that with embedded carriage control, so EDIT
can edit it and output can be used as log-on
message files. (Unfortunately underlining via
an overprint no longer works, but you can't
have everything.)
2. RNP
RNP is a pre-processor for RUNOFF. It has the
ability of inputing a file containing normal RUNOFF
source and:
1. Opening secondary files and inserting the text
in the main RUNOFF source.
2. Opening TI: as a secondary input.
3. Issuing prompt lines to TI to that you know
what you should enter.
(This letter was prepared from a stock form using
RNP.)
3. DIABLO
After we got our DIABLO printer, we discovered we
couldn't do all the nice things because the TTY
handlers kept stripping off most of the control
codes we sent. DIABLO reads in an input file and
sends it to the DIABLO in WRITE-PASS-ALL mode.
(You will have to change the TKB command file to
reflect which terminal line the DIABLO printer is
connected to.)
4. TWOPAGE
TWOPAGE is a REESE BASIC program which seperates
seperate pages into "left" and "right" files. The
resultant files are printed seperately on the
printer, (reloading the paper at the same starting
point, but reversed before the second printing.)
The result is a manual printed on both sides of the
paper. Provision is made to ensure that chapters
are always started on a "right-hand" page, etc.
5. RNOMAN
RNOMAN is an example of using REESE BASIC to ex-
tract CHAPTER, APPENDIX and HEADERLEVEL data from a
first pass of the RUNOFF manual, and creating a
TABLE OF CONTENTS. The table of contents is then
inserted into the RUNOFF source, and a second pass
of RUNOFF produces the final manual herein present-
ed.
[300,375] account contents
This version of ADVENTURE is the same as the one on an earlier
New Orleans Spring 1979 [307,7] DECUS Symposium tape.
This has been converted to upper/lower case, so the messages are
more tasteful (a la DUNGEON), some minor bugs have been fixed,
it has SAVE/RESTORE, so you can interrupt a game to continue
later, and it has pre-loaded data files, so it doesn't
have to go thru a long initialization phase when you start up.
This version **DOES NOT** use any floating-point or long integers.
[301,001] account contents
A readme.1st file describing SCLUG's spring 1981 submission.
[301,002] account contents
Contains information about all programs that SCLUG
has submitted. LIB.DIR is a brief 2-line description of each program, and
LIB.DOC is the complete description of the programs.
**********************************************************************
editor's note: All LUG librarians should study the style with
which SCLUG maintains its library. Jim Neeland does a very good
job.
**********************************************************************
[301,053] account contents
Description: A SLP file to convert DEC's copyrighted HELLO distributed
source into a FORTRAN-callable subroutine to read and display standard-formatted
.HLP files (see MCR OPERATIONS MANUAL under HELLO for format, or read comments
in MACRO source code.) Also included are a sample/test FORTRAN program, along
with a little utility subroutine to cancel control-O's (in case terminal was
attached by main program).
[301,054] account contents
This file describes the rather meager contribution from the Computer-Aided
Technology and Acoustics groups at Douglas Aircraft for the Spring 1981
DECUS symposium. Included in this directory are a number of small, but
useful utility programs and subroutines. Each group of routines has its
own .DOC (and maybe even a .RNO) file to describe its function. The programs
are:
RID - Multi-purpose program TAB/deTAB, de-trailing blanks/
sequence number-er program written in RATFOR
FLABEL - Uses HP-7221 plotter to plot "fancy" (multi-font)
character labels
CCLDSP - Reads your CCL command file(s) (if you have CCL, that is)
and prints a nice, sorted & formatted listing of available
commands
HELPDSP - Reads your HELP.HLP file for all top level entries and
outputs a nice, sorted & formatted list file suitable for
use in creating help indices
SCR - Reads your Teleray-12 screen memory into a disk file and
optionally spools it the the print spooler or adds a time-
stamp or appends it to an old file
CLEAR - Clears the screens of VT-52, Teleray-10/12 and VT-100 CRT's
(as well as any 24-line CRT) as well as enables the buffering
features of Teleray-10 & Teleray-12 terminals
GETMCL.MAC - Fortran-callable subroutines used to read input like all
those fancy MACRO utility tasks
MAG.MAC - Fortran-callable subroutines to handle magtapes
SHIPS - Teleray-12-oriented CRT game
We have also included our help files, which involved no little time and effort
to type in, in the hope that they may save someone else the trouble.
[301,055] account contents
THIS IS FOR IAS USERS USING HI LEVEL LANGAUAGES (F4P,FOR,PL1)
AND WANTING TO USE (TCP) SUBTASKS.
"SUBTASK.BIS" IS A BATCH STREAM THAT WILL BUILD A LIBRARY OF MARK LEWIS
OLD TIMESHARING CONTROL SERVICES SUBROUTINES. THERE WERE A NUMBER OF ERRORS
IN THE OLD SUBTASK PACKAGE (PROGRAMS THAT HAD BUFFER OR ARRAY ADDRESS
THAT WERE 16K OR OVER WERE REJECTED AS ERRORS), ALSO THE LIBRARY WILL
ALLOW THE USERS TO EXTRACT ONLY THE ROUTINES THAT ARE NEEDED, NOT THE
HOLE PACKAGE.
THE USER SHOULD EDIT SUBTASK.BIS TO REFLECT THE DESIRED UIC'S.
[301,056] account contents
This directory contains a new version of the mp macroprocessor.
See readme.2nd for details.
Using the above along with some C routines is a C runtime library
(pieced together from other librarys) which supports the unix
standard i/o package plus a true disk cache (this is new). Thus it
efficiently supports the array of bytes concept. With this, one
can treat a file as a 32 bit address space. So, along with this
are some C routines which implement a HEAP structure on disk. This
was for a school project on relational database storage structures
and also includes a package of doubly linked list routines for the
disk heap.
[301,057] account contents
This directory contains a front end to a command line interpreter
written in MP which permits the user to spawn commands that can be
saved, edited, and invoked via control characters. It is under
3.3K with just 2 buffers, but it expands. See the file jpl.doc.
[301,060] account contents
This directory contains two programs which I use to create
multiuser programs. See the feb 1981 multi-tasker for details. The
programs here just copy portions of files from one file to
another. The rocpy also needs to know the address of the disk r/o
code pointer cell. The program echos the expected value if no
parameters are given on the command line.
The file f3 is just a module with an fsrsz of 3. The remainder of
the build files (.ftb) references the mp macroprocessor directory.
[301,061] account contents
This directory contains an Evans and Sutherland PS2 Help file. It
was created from the manual sources.
[302,001] account contents
a readme.1st describing the SEATTL LUG's contribution to the
1981 spring RSX/IAS sig tape.
[302,211] account contents
TMON is a task monitoring utility based on Bob Stodola's PC monitor (PCM).
It is very useful either as an optimizing tool or as a debugging aid.
TMON monitors a target task's PC and overlay segment descriptor table
and accumulates statistics. A reporting task will read the accumulated
data a provide a histogram of the task'S PC and overlay segment activity.
The histogram report can be for the entire task or just selected portions
of it.
This program is dependent on the structure of IAS. Some revision
would be necessary for it to work under RSX-11M or M+.
[302,304]account contents
INDEX is a highly enhanced FORTRAN cross referencing program. In addition
to regular cross referencing, it can cross reference entry points with
modules or all variables for an entire program. It also has numerous
switches for such things as identifing defined, but not referenced variables.
INDEX was originally written by Michael Levine of China Lake and has been
submitted to Tape Copy at least twice, the last being the Fall of 1980
in San Diego (I understand he plans to submit it with further revisions at
the 1981 Fall Symposium in Los Angeles).
This version contains the following changes to the San Diego version:
1) Capabilility of merging the cross reference listing of a file with
the actual F4P listing.
2) Capability of differentiating between FUNCTION calls and array references.
3) A few bug fixes and miscellaneous clean-up.
[303,120] account contents
TEDI EDITOR FORMATTER
TEDRSX.TED is the source text for the documentation, set
up for the TEDI formatter, while TEDRSX.DOC is a formatted
version ready for a printer.
TEST.TED is an example of TEDI formatter source text which
exercises most of its features, including equation formatting
to daisy-wheel printers.
************************************************************
editor's note: sorry, but the documentation file "TEDRSX.DOC"
was not run through TECO. So it has a bunch of unwanted carriage
return,line feeds in it. FLX eats another one. For those who don't
know it, the DOC file output of RUNOFF (or TEDI ??) must be passed
through TECO before FLX will correctly copy it to magnetic tape.
**********************************************************
[307,030] account contents
TOOLGEN.CMD, the command file for building
the LBL Software Tools Virtual Operating System. The
release notes for the VOS are also contained in this
UIC.
[307,031] account contents
Fortran and macro sources for VOS.
[307,032] account contents
Manual entries for VOS utilities.
[307,033] account contents
Ratfor source files for VOS utilities.
[307,034] account contents
Source files for variable-length send/receive driver.
[307,035] account contents
Source files for virtual aether driver.
[312,315] account contents
DDT is able to debug one task from another now, symbolically.
That is, this version of DDT may be built as a separate task and
the task being debugged is then linked with a small (about 200 words)
DDT "kernel" (called, appropriately enough, DDTKNL) which controls
execution. DDT itself does most of the work of debugging and supports
a very large and powerful array of commands. These include the
ability to read symbol table files, memory watchpoints, breakpoints,
single stepping, examining numbers in any format including floating
point, setting variables, defining labels, looking at assembly language,
or examining memory image files. The ability to examine core image files is
present to permit DDT to be used (reading the symbol table) to allow
an effective SYMBOLIC OPEN on RSX11M.SYS files or TASK IMAGE files or
crash dump files. (Yes, you may now pretty much discard ZAP once
DDT is installed.)
A simple code profiler is included in DDT which allows you
to specify a base address to "profile" (PRFBAS) and a shift count
to shift addresses down. If DDT is put into single step mode, a
DDT array called PRFHST will be incremented in the bin determined
by the PC of the instruction. The calculation finds the word to
increment by computing (PC-PRFBAS) shifted down by the "shift count"
bits. If the result fits into PRFHST, the PRFHST word is incremented.
Normal DDT commands may be used to display PRFHST after you get
the program to stop.
Because DDT is now able to run in a separate task, it is possible
to do some fancy things like reading symbol tables that were too
costly before. Since many users are FORTRAN programmers, a
public domain program is included which was slightly modified from
a version at Institute for Cancer Research which can read FORTRAN
.LST files and taskbuilder .MAP files and construct a symbol table
file. This program is designed primarily to make the symbol file for
FDT, a FORTRAN debugger that is also included. However, it can also make
up a pseudo- .STB file that DDT can read and which gives separate
names to each FORTRAN line number, statement label (where possible),
and makes all variable and array names effectively global so DDT will
know, once it opens the symbol table file, where they are. This
permits DDT to set a breakpoint symbolically on any FORTRAN line number
so that you can debug FORTRAN programs symbolically at a fairly high
level if you can find as few as 200 words in your task's space for
the debug kernel.
DGT and DGWRIT are separate tasks; see the DGWRIT.FLX source
(in FLECS, so the FORTRAN version is supplied for those not having the
FLECS preprocessor in house) to see how to put DGWRIT together. DGT
works off console commands and is pretty much a manual tool. Its
only virtue is that it works. There is a file called DGWRT7 whose
intent is to allow you to write 7 track tapes a DG machine can read.
Don't mix them; DG does some things differently from DEC on 7 track
tape and messing around was needed.
The FDT and FDTSYM files on the tape are the FORTRAN debugging
tool from Institute for Cancer Research, but with the FDTSYM as a mod
of their TRACE program to generate symbol tables. There are rumors the
package has some problems on 11M systems; should be solid in IAS.
For amusement, I have added also another F4P debug aid called
FPODT (FPODT and PREPRC are the sources pertaining to FPODT) which
goes through a task image and changes all references to $SEQC to
traps. A module in your task (which must first call it to initialize
and all of whose FORTRAN programs must be compiled with /TR:ALL) then
allows single steps, variable examine/modify, or breakpoints. It
does not allow starting from new places as FDT and DDT do. However,
it can be used along with either FDT or DDT. Just depends how much you
are willing to do to find the bugs. FPODT single stepping would go from
one trap to the next, so does not need to set the T bit and ought to
be faster than single stepping with FDT or DDT, even in your task
image, since it steps FORTRAN lines, not instructions.
The TPCDIR program here can list directories of TPC container
images of FLX tapes, and even extract files from such images, one at
a time. The main use is to avoid a separate tape pass to make a directory
when making tape copies.
A new version of the MSX system is included. It works under RSX11M+
now and has several bugs fixed so that it is much better tested. Note when
building it that the security categories in MXSTLS should be set up as
you really want them...the system will block data transfers where military
security policy is violated and the present sources are experimental and have
several invalid possible transfers with those security categories. The
MXTST module is the test program which illustrates its use. It should be
interesting to those who want to experiment with what trying to use a
secure kernel is like, especially on distributed systems. Note the stand
alone version is considerably less tested than the RSX version, but the
basic mechanisms, mapping, and I/O are known to be good. Be careful of
using directives like MODSTT in secure kernel systems; they only work
from libraries or if your task has "ignore multilevel security"
privileges. Also, you need the asynchronous I/O privilege to allow any
I/O to be logically asynchronous in a secure system. Device ownership
may be implemented by adding a special category for "device" to the
special category list of the device and that of its owner.
Also included is a version of Michael Lampi's XMIT program that has
been built to run under IAS V3. (Should be OK in V3.1 but can't test yet.)
The IAS code has been mostly conditioned, and the RSX11M version also took
less CPU than the original. However, RSX11M and RSX11M+ users may have to
do a bit of retrofitting. It may be the IAS version will work in 11M too
but it's terminal I/O is via read/write, not via AST for input, relying
on the full duplex driver to keep characters around. Reducing CPU load
has been a primary concern.
[313,001] account contents
The readme file for the ORNL version of virtual disks.
If you are unfamiliar with Virtual Disks, you should find out about them.
Unfortunately, I didn't have time to write up a thorough description of
what VD's are. They are described by Ralph Stammerjohn (the original
inventor/author to whom all praise and glory belongs) on the Spring '79
DECUS tape. I presented a paper on VD's at the Fall '80 DECUS in San
Diego and the write-up in the Proceedings of that Symposium describes
VD's, how they work, and how great they are.
[313,010] account contents
HELLO and BYE
***********************************************************************
editor's note: The modified sources to HELLO and BYE were deleted. This
is liscensed software and as such cannot be distributed. Submitors are
encouraged to submit SLP files when they wish to modify a DEC source module.
I ran CMP and compared HELLO and BYE as modified by Ken with the HELLO
and BYE supplied by DEC on an 11M v3.2 release kit. The results of the
comparison are in HELLO.CMP and BYE.CMP.
*************************************************************************
The modifications to HELLO and BYE are not required to implement the
---
rest of the virtual disk package. The mod to HELLO checks for a
command file in LB:[201,1] that corresponds to the UIC of the person
logging on, e.g., if they are logging onto UIC [123,321], then HELLO
looks for file LB:[201,1]123321.CMD. If it finds it, then it is executed
with the terminal privileged and slaved.
The modification to BYE simply runs DVD to dissociate any virtual
disks the user had while he was logged on. We have found this to be
an essential clean-up operation when using virtual disks. (The changes
to BYE would have been unnecessary if RSX-11M would run a command file
on logout like it does on login.)
[313,022] account contents
Virtual Disk Driver
This account contains the source for the virtual disk driver (VDDRV). It
has changed very little
from the original Stammerjohn version, except for some protection
mechanisms to make sure the host disk (i.e., the disk on which the VD
is allocated) has not been changed. VDDRV is really trivial (which is
the beauty of the entire VD package).
[313,020] account contents
AVD and DVD
There are two other utility programs called AVD and DVD. (I have changed
the meaning of these from the original "Allocate Virtual Disk" and
"Deallocate Virtual Disk" to "Associate Virtual Disk" and "Dissociate
Virtual Disk" since there was some confusion with the MCR ALLocate
operation.)
AVD and DVD have been changed considerably from the original Stammerjohn
version. The reason for the changes were to enable a large number of
unsophisticated users to use VD's easily.
[313,030] account contents
TRMSET
Our system has many people connecting with different types of terminals
on different lines all the time. So we wrote this simple little
program called "TRMSET" that sets the terminal characteristics for
users when they log in (TRMSET is automatically run by HELLO when
you log in).
TRMSET sends the escape sequence to request a terminal to identify
itself. Many terminals respond with an identification string of some
sort. TRMSET then sets the terminal characteristics for that type
of terminal. Currently, TRMSET recognizes VT100's, LA120's, VT52,
and VT55.
If a terminal doesn't respond within a time-out period, then TRMSET
sets some default characteristics based on the baud rate of the line.
[330,001] account contents
LIST is a utility for displaying selected portions of files. It
is possible to display lines from a file in several modes, including a
screen at a time, as well as to search for character strings and write
out portions of files to disk. In addition, LIST can use SRD to se-
lect multiple files, making it easy to do such chores as looking
through a group of files for a particular string.
LIST is very fast, as it uses a macro interface to FCS to read
and write files; in addition it buffers the disk address of every 100
lines to simulate random access on the file.
This version of LIST includes the capabilities to subtask SRD and
subtask itself to list a help file. The help file is included
(LIST.HLP) and should be placed on LB:[1,2]. The routine which gets
file names from SRD is called FNDFIL.FTN; if for some reason SRD
doesn't work with LIST (if your SRD has a different output file format
from our SRD, for instance) then modify this file. If you would rath-
er use PIP than SRD, there is a file called PIPFIL.RAT (and also PIP-
FIL.FTN) which contains the FNDFIL subroutine, but for PIP.
LIST is written almost entirely in Ratfiv; the .FTN files are
included, however the Ratfiv compiler is also included on this tape,
and should be very simple to build. (See README.1ST on the Ratfiv
compiler directory.)
This program works on IAS and RSX-11M.
[330,002] account contents
This UIC contains the Institute for Cancer Research
FORTRAN IV Plus Symbolic debugger. It is designed for IAS, but
should also work under RSX-11m. Please read FDTEXP.DOC for more information.
[330,003] account contents
CPL is a utility which compiles your programs in a compiler independent
manner. It alleviates the headache of having to remember the syntax quirks
of all the compilers on your system. CPL also aids in the maintenance of
programs consisting of many modules since it compiles files based on the
dates of the source and object files and maintains a user library.
Features include:
The same command line syntax is used for all compilers.
CPL selects the compiler to use based on filename extension.
CPL will only compile a given file if the source file post-dates
the object file.
User library maintenance. CPL will replace in the users library
only those files on a command line which were actually compiled.
Multiple pass compilations are done with a single command line.
Implementation of prefix files in a compiler independent manner.
Compilers supported in the current implementation:
FORTRAN 4 PLUS
MACRO
BASIC PLUS 2
RATFOR
NBS PASCAL
SWEDISH PASCAL
WHITESMITHS C
[330,004] account contents
SRD command line generator
SRDCMD generates command lines which contain file names selected
by SRD (or PIP - see below). For example, the following SRDCMD com-
mand generates command lines to copy all files created today to UIC
[6,120]; the command lines are written to the file T.CMD:
SRDCMD> *.*/DA COPY 'N'E [6,120]*.* >T
The string 'N'E specifies that the name and extension of each file
name be expanded at that point in the command line. Version, UIC, and
device information can also be placed anywhere in the command line.
For further information, see SRDCMD.DOC.
The routine which gets file names from SRD is called FNDFIL.FTN;
if for some reason SRDCMD doesn't work correctly (if your SRD has a
different output file format from our SRD, for instance) then modify
this file. Another alternative is to build the version of SRD includ-
ed on this tape, which is an enhanced and debugged version. If you
would rather use PIP than SRD, there is a file called PIPFIL.RAT (and
also PIPFIL.FTN) which contains the FNDFIL subroutine, but for PIP.
SRDCMD is written almost entirely in Ratfiv; the .FTN files are
included, however the Ratfiv compiler is also included on this tape,
and should be very simple to build. (See README.1ST on the Ratfiv
compiler directory.)
This program runs on IAS or RSX-11M.
[330,005] account contents
This is an enhanced version of SRD.
Enhancements include:
1. The selection string wild character '*' will match any
string or the null string.
A*B.* will match AB., AAB.FTN etc.
2. A /RV[:date] switch which will select files by revision date.
It may be combined with /BE, /AF.
3. A /SU switch which will give a storage allocation summary
for files selected.
4. A /OW[:selectionstring] switch which will select files by
owner. The selection string syntax is the same as file
selection strings. The wild characters '?','*' may be used.
/OW:[2??,1*] will match [200 thru 277,1],
[200 thru 277,10 thru 17],
[200 thru 277,100 thru 177].
[330,006] account contents
TAPE instructions.
Tape is designed to run under version 3.00 of IAS on a PDP-11/70. On an
IAS system it may be linked using the command file EXPORT.CMD. This program
has not been tested on an RSX-11M system, but should require little or no
modification. The command file RSXBLD.CMD may build TAPE correctly under
RSX-11M, but this also has not been tested. Note that the build files assume
that the FORTRAN-IV Plus library is in the system library. Begin by copying
all files to the system disk. NOTE: QIOSYS.FTN MUST BE PUT ON [1,1] AND
ALL OTHER FILES ON [22,11] TO COMPILE CORRECTLY.
[330,011] account contents
This program resequences statement numbers in FORTRAN IV-PLUS
programs. See RESEQ.DOC for more detail.
[330,012] account contents
TRU truncates files. Though PIP will also truncate files, PIP will
change the revision date on all files specified, even if they did not
require truncation. This will cause an incremental backup program (such as
BRU) to needlessly backup files which have not been changed. TRU first
ascertains whether or not a file requires truncation before invoking
PIP to actually do the truncation.
[330,014] account contents
This uic contains three programs for IAS v3.0.
The GAME program allows a common interface to games. Used in
conjuction with the SCHEDULE program, it also allows the system manager to
restrict game playing to specified hours.
See GAME.DOC for more information.
RUNNL is a program which will spawn a task on the privileged TI of
device NL0. In order to use, you must have included this device in your
SYSGEN command file, though it need not have any device handler associated
with it. RUNNL handles all task scheduling requests sequentially, and
will not schedule a second task until the first has exitted. Where possible,
therefore, tasks which may take more than several seconds should be started
by spawning ...RUN with a command line to run the desired task.
see RUNNL.DOC for more information.
The program SCHEDULE allows a system manager to schedule tasks to
be run at specified times and days. For example, you may schedule a
timesharing parameter be changed each weekday at 08:00, and again at 18:00
on weekdays, or that a file be printed each saturday at 20:00, etc.
See SCHEDULE.DOC for more information.
[330,015] account contents
Ratfiv is a structured Fortran preprocessor providing
SWITCH, IF - ELSE, WHILE, FOR, DO, REPEAT - UNTIL, STRING,
and BREAK and NEXT constructs. Also supported are INCLUDE
files, DEFINE for symbolic constants and macros with
arguments, conditional compilation, use of >, <, etc.
instead of .GT., .LT., etc, and the RETURN VALUE construct
in functions.
Ratfiv was developed from the Ratfor compiler
distributed by Lawrence Berkely Labs; the major
enhancements in it are:
1. Ability to specify a format statement inside READ,
WRITE, ENCODE, and DECODE statements;
2. True line continuation using the underline
character;
3. Production of properly indented upper case Fortran
code with comments passed through;
4. Addition of a /SYMBOLS switch to the command line
to specify reading of the SYMBOLS file;
5. Incorporation of a powerful, easy to use macro
processor;
6. Conditional compilation of large blocks of code;
7. Correct line number reporting;
8. Exit with error status if an error occurs during
compilation;
9. Use of character constants in case labels.
10. Ratfiv keywords need not appear at the beginning of
a source line in order to be recognized.
This version of Ratfiv may be used if you want to
compile the Ratfiv source to the other programs written in
Ratfiv on this tape.
Documentation is in the file RATFIV.DOC.
[330,016] account contents
(the IAS version of MULTI-TREK)
MULTI-TREK is a game similar to traditional Star-Trek games
except that it involves several players (currently up to 8), who
play against each other rather than against imaginary Klingons,
Romulans or whatever. The MULTI-TREK universe consists of a 100
by 100 matrix, of which an approximate 20 by 20 matrix of each
player's immediate area is displayed and continuously updated at
their terminal. MULTI-TREK was originally submitted to the New
Orleans collection, then to the 1979 San Diego collection with
the addition of robot ships. This version has been recoded in
RATFOR and has significant enhancements and improvements, most
notably the addition of energy nets, an improved screen display,
much better input-output (any terminal with clear screen and
cursor positioning may be used at speeds from 1200 baud on up),
use of default ship and direction with most commands ("L" sets
the default direction), more competitive scoring, smarter robot
ships, and fewer bugs. Complete instructions for starting and
playing MULTI-TREK are in the file MTREK.DOC.
The game consists of a driver task (MTREKD), a player task
(MTREK) for each terminal, and a shared global common area for
inter-task communication. This version runs on IAS 3.0, however,
it should run under any system with shared global areas and
rudimentary task switching or time-sharing. Real-time systems
without time-sharing may have to play with priorities to insure
that the driver task gets sufficient CPU time and that all
players get an equal amount of time.
MULTI-TREK was written with a structured Fortran called
RATFOR. This UIC contains all of the Fortran produced by RATFOR,
as well as the RATFOR source (file type .RAT). The RATFOR
compiler is also included on the distribution tape.
MTREK will work properly with any terminal with clear screen
and cursor positioning. The source code is set up to drive
ADM3-A's or VT100's. if you are using different terminals, you
should modify the file "TERMINAL.RAT" (you could modify
"TERMINAL.FTN") to drive your terminal. Modification should be
straight-forward as the code is commented on where to change it.
[330,017] account contents
(the VAX/VMS version of MULTI-TREK)
MULTI-TREK is a game similar to traditional Star-Trek games
except that it involves several players (currently up to 8), who
play against each other rather than against imaginary Klingons,
Romulans or whatever. The MULTI-TREK universe consists of a 100
by 100 matrix, of which an approximate 20 by 20 matrix of each
player's immediate area is displayed and continuously updated at
their terminal. MULTI-TREK was originally submitted to the New
Orleans collection, then to the 1979 San Diego collection with
the addition of robot ships. This version has been recoded in
RATFOR and has significant enhancements and improvements, most
notably the addition of energy nets, an improved screen display,
much better input-output (any terminal with clear screen and
cursor positioning may be used at speeds from 1200 baud on up),
use of default ship and direction with most commands ("L" sets
the default direction), more competitive scoring, smarter robot
ships, and fewer bugs. Complete instructions for starting and
playing MULTI-TREK are in the file MTREK.DOC.
The game consists of a driver task (MTREKD), a player task
(MTREK) for each terminal, and a shared global common area for
inter-task communication. This version runs on VAX VMS, however,
it should run under any system with shared global areas and
rudimentary task switching or time-sharing. Real-time systems
without time-sharing may have to play with priorities to insure
that the driver task gets sufficient CPU time and that all
players get an equal amount of time.
MULTI-TREK was written with a structured Fortran called
RATFOR. This UIC contains all of the Fortran produced by RATFOR,
as well as the RATFOR source (file type .RAT). The RATFOR
compiler is also included on the distribution tape.
MTREK will work properly with any terminal with clear screen
and cursor positioning. The source code is set up to drive
ADM3-A's or VT100's. if you are using different terminals, you
should modify the file "TERMINAL.RAT" (you could modify
"TERMINAL.FTN") to drive your terminal. Modification should be
straight-forward as the code is commented on where to change it.
[332,200] account contents
CONTRIBUTION FROM HOUSTON AREA RSX/IAS LUG
DSCDIR - FORTRAN PROGRAM TO PRODUCE A DIRECTORY OF THE CONTENTS
OF DSC (DISK SAVE AND COMPRESS) TAPES.
FILES: DSCDIR.FTN, DSCDIR.CMD, DSCDIR.DOC
DSCCPY - FORTRAN PROGRAM TO COPY JUST A FEW FILES FROM A DSC
TAPE WITHOUT RESTORING THE ENTIRE VOLUME OR READING
ALL REELS OF A MULTIPLE-REEL SET.
FILES: DSCCPY.FTN, CPYFTN.FTN, CPYMAC.MAC, DSCCPY.CMD
DSCCPY.DOC
DSCTAPE - TEXT FILE WHICH SUMMARIZES WHAT I KNOW ABOUT DSC TAPE
FORMAT. FILE: DSCTAPE.TXT
[336,200] account contents
******************************************************************
editor's notes: This software was given to the NETWORKS SIG at
the Spring 1981 DECUS meeting. The new pages for RMDEMO work even
if you DO NOT HAVE DECNET.
******************************************************************
The following programs were submitted by the Digital Equipment
People as Network software for RSX-11M
1) New page displays for rsx-11m rmdemo display to display
a) active task list
*************************************************************
editor's note: The active task list page looks like it might
show the names of the files that a particular task has open.
It could be a very valuable debugging tool.
*************************************************************
b) remote node display (maps)
c) logical link display
d) help page
2) Vms compatible network mail for rsx-11m
3) Remote command terminal from rsx-11m to
a) tops-20
b) rsts
c) vms
4) Poor mans node route through, allows phase II DECNET messages to
route throught a phase III node or thur another phase II node.
5) Remote virtual devices using DECNET
***********************************************************
editor's note: This allows an RSX-11M system to run without
a disk (what about boot physically attached to the CPU.
The RSX-11M system uses the "virtual disk device"
via DECNET to connect to a physical disk on a host system.
Currently the host system is another RSX-11M system. Maybe
someone could work out the software to make the host system
a VAX/VMS system.
************************************************************
a) Virtual Disk
*************************************************************
editor's note: this virtual disk is something quite different
than the Virtual disk invented by Ralph Stamerjohn. An
implementation of Ralph's virtual disk appears on this tape
in uic [313,*].
*************************************************************
b) Virtual Magtape
c) Virtual Line Printer
[337,030] account contents
This submission is a revision to the graphics package
that was submitted on the Chicago spring 1980 tape.
1. a bug in the cross hair cursor support has been eliminated.
2. a bug in SCALE.MAC which caused an occasional floating point
to interger overflow error has been eliminated.
3. to run under RSX11M you should just assemble all the *.MAC'S
and build your library.(the original version incorrectly directed
you to do some conditional assemblies)
4. there is a conditional assembly statement to change in PLOTPRIMS.MAC
and PLTDAT.MAC to allow the package to support a Tektronix 4025
(the user program must call TK4025)
5. for users in an overlayed enviroment only PLTDAT.OBJ needs to
be in the root.
6. a bug in the call to 'HOME' has been fixed to allow it to work on
all terminal types.
SFGL70 is a graphics package which supports Tektronix
4006,4010,4014 AND 4025 terminals. It's useful feature is that in
most cases it will only add approximately 1K words to the user task.
The SFGL70.DOC file explains the SFGL70 package submitted to the
1980 FALL DECUS Symposium by Ken Demers.
The package can run under IAS and RSX-11M.
The package will work on a VT100 (if it has the hardware
to make it look like a tektronix 4010). There will be a future enhancement
to take full advantage of the VT100's additional capability.
[344,001] account contents
This uic contains the readme files for the KMSKIT.
This KIT contains KMSKIT UPDATE #4. To use these files you
need the files on the Fall '79, Spring '80 and the Fall '80 RSX SIG
DECUS symposium tapes.
1. A flashy, super enhanced CCL. [344,61]
2. New features and reports for system accounting programs. [344,65]
3. New features for transient libraries. [344,51]
4. New HELP feature from Ray Van Tassel [344,40]
**********************************************************************
editor's note: the uic [344,60] was submitted by Ray Van Tassle and
appears here with Jim Downward's permission.
*************************************************************************
KMSKIT's next major release, is scheduled for release with
RSX11M V4.0.
[344,040] account contents
W A R N I N G W A R N I N G W A R N I N G
1.
The ability to pass command lines to non-installed tasks will be
modified. Only commands passed in quoted strings will be supported.
ie /PRM ="command line" is legal, whereas /PRM=command_line will
be illegal. /PRM may or may not be supported. If it is not supported
a modification almost identical to it will be. ie /CMD="full_command_line".
The difference between the two syntaxes is that "full_command_line" requires
that a 3 letter MCR like command name occupy the beginning of a command
line, whereas /PRM="...." automatically inserted a dummy three letter
command name into the first three characters of the MCR buffer. At this
time the /CMD="...." will definitley be supported, and the /PRM="..."
is also supported but may get dropped. Please contact me if dropping
support for the /PRM form will cause undue hardship.
2.
HELP processing was previously aided by the inclusion of a matchall
matchall character,'#' which if found in position 1 would match all keywords
at all levels. It could only be placed at the very end of a file. The
matchall was created before V3.2 used the # for synonyms. This causes a
conflict and eronious behavior at times. For HELP for V4.0, the # matchall
will be dropped in favor of a '*' wildcard which will be able to provide
a matchall at any level rather than just being at the end of the file.
This SLP file for HELP is so good, I just had to include it.
It got passed on to me by Ray Van Tassel. It is designed to SLP either
the unmodified HELLO or the HELLO modified by the KMS modifications.
When applied a ? in place of any keyword identifier is taken to
request a formatted listing of all keywords available at that level.
Hence if a help file looks like
1 FOO
...
2 key1
....
2 key2
....
2 key3
....
1 BAR
...
1 PIP
...
1 UTL1
...
...
1 UTLn
....
then HELP ? would produce a listing of the format
FOO BAR PIP UTL1 UTL2 UTL3 .....
... UTLn
and HELP FOO ? would produce
key1 key2 key3
The one restriction with this is that matchalls '#', are not handled
correctly.
This feature will be expanded for use with RSX11M V4 HELP processing to
include the ability to automatically produce listings of keywords if
invalid keywords are entered.
[344,050] account contents
One problem with transient libraries is that if memory is tight
the region required for the library sometimes can not get created and
LDLIB produces an annoying message that the task can't run because
the region could not be created. Several steps are taken to solve this
problem (at least partially)
1. The SUSPEND/RESUME logic/handshaking between LDLIB.TSK and LDLIBS
is replaced with STOP/UNSTOP for systems having parent/offspring
tasking GENed into their system. A STOPed task has 0 effective priority
and is immediately elegible for checkpointing. A suspended task only
can get checkpointed by a higher priority task. This makes it easier
to checkpoint the initiating task while LDLIB is doing its thing.
2. CKP... has been modified to accept requests from LDLIB to clear
out core. LDLIB now sends off requests to CKP... to checkpoint
everything (!!!) it can when LDLIB needs to load a loadable library.
Clearly this would cause a lot of thrashing on a system in which
tasks with loadable libraries were bouncing in and out of core constantly.
However, in many systems, the tasks built with loadable libraries are
infrequently used.
3. CKP... has been modified so a privileged user can RUN CKP... and
checkpoint everything out of core.
[344,060] account contents
<<this submission is from Ray Van Tassle and it appears here
with Jim Downward's permission>>
This is a set of fixes and enhancements to the KMS-FUSION CCL.
I have sent them to Jim Downward, and they appear here with his blessing.
He is supposed to be working on an entirely new release of CCL, which
will fit in with M4.0 (whenever it comes out). That one will undoubtedly
be better with 4.0 than this one, but this one works better with 3.2 than
his old one.
Also in this UIC is a seperate "HEY" task. I find that HEY is
so useful, that I want to do it even while CCL is busy, so I yanked
it out and made it stand-alone.
CCL is great !!!!!!!!! However, it had
some bugs and some awkward things in it's operation. I have spent
quite some time of fixing it up, as follows:
Fixed up some bugs (there are some error messages that go out
LUN 5, which is difficult since there is a UNITS=2 in the TKB
file).
1. I made it privileged (but it sets itself non-priv), so that
it links to the EXEC. it can now use $GTTSK for "HEY" and
"-taskname", so you can just say "HEY TKB"
2. Added some validity checks for the ?.CCL file, so that it
will catch you if you mess it up (which is very easy for a
non-experienced CCL user to do in making his own
USERCCL.CCL). Also allowed for null lines comment lines, so
that you can read a ?.CCL file without going blind.
3. If it has no command line, it will prompt for one. If it
does have to prompt for one, it will also print out each
command line before spawning it to MCR. Sometimes this is
about the only way to see how to set up your CCL control
line, to make it come out right. In essence, for debugging a
new CCL command, you type "RUN ...CA." and enter your command
to the CCL prompt.
4. The last line of a CCL command does not wait, it passes the
OCB (if any) and MCR prompt status (bit T3.MCR) to the
spwaned task. This lets CCL get out of the way as quickly as
it can, especially when you do a COPY or DEL, etc, which will
take a long time, and you can start up another CCL thing
(many of our infrequently used programs are run by CCL).
This removes the need for the %$ parameter (although it is
still in). For those times when you want CCL to stay active
for all the programs (for example if you specifically want
CCL to tell you about the exit-status) you merely make the
last command a null line (which did not used to be
legitimate).
5. CCL now prints out a message if a tasks exit status is not
SUCCESS, and will quit (with a message), unless it is only
WARNING.
[344,061] account contents
CCL V7.0B
This version of CCL is a composit of the work of Angle Li at the
University of Miami, Paul Sorenson at Michigan State University, and Jim
Downward at KMS Fusion. Angle Li edited the code to remove all the funny
F4P lables and to comment the code. In addition he conditionalized the
internal commands so that section could optionally not be included.
Look for the symbols $DCL and $PIP in CCL.MAC to see what effect the
deleting the conditional code would have. Paul Sorenson then provide
a substanitial rewrite to CCL to provide significantly more flexible
command string parsing. I then fixed some bugs with the the changes,
added some additional special characters and packaged this latest release
for DECUS.
To simplify use of CCL for user's(like myself) who often assign a local
LB:(to do a sysgen). The location of SYSCCL.CCL has been moved to LB1:
(unless symbol $LB0 is defined in PRECCL.MAC). To use CCL, please do a global
assign of LB1: to a physical disk either in VMR or in STARTUP.CMD.
As was previously mentioned, this version of CCL has been extensively
rewritten to provide a more flexible command format. The major changes that
must be kept in mind for this newest version are,
1) All conditional assembly directives are now in PRECCL.MAC
for easier access (see comments for their affects).
They are $LB0,$PIP,$DCL, and $TABLE.
2) An internal table lookup has been added to offer a compromise
between DCL and the .CCL files. (see comments in CCL.MAC)
3) Several new parameters have been defined -- %A%, %B%, %C% and %0%
%A% is set the the entire character string past the command.
%B% is set to character string up to first space
%C% is defined as everything not in %B% (excluding delimiter)
%0% is set to values of any leading switches
(e.g. BRU/REW/VER, %0%=/REW/VER)
4) Two options have been added for command building. If a
parameter is null, it may be replaced by a string starting with
the key character "\" and extending to the % delimiter, also,
a parameter may be tested by %2?S1\S2% which will insert string
"S1" if parameter 2 is defined, otherwise string "S2" will be
inserted (either S1 or S2 may be null).
5) A command line may extend to the next line by including the "+"
character as the last character in the line.
6) USERCCL.CCL is opened on the user's logon UIC rather than the
current default UIC.
7) If CCL can not find a command, it will try sending off its
command to another catchall ...CA2.
8) %R is a special character signifying that CCL is to ring the
terminal BELL when it exits. A previous version used %B for this.
Sorry for the incompatability.
9) %D is added to aid in debugging CCL command lines. When it is
encountered, the command up to that point is displayed in
expanded format at the user's terminal rather than being sent
to MCR.
9) %P is added to signify that an action line is to be printed out
on TI:. This makes it very easy to create CCL commands which
send escape sequences to control terminal characteristics.
10) %Q is added to signify that CCL should exit without sending on
the command to MCR if a required input field is defaulted with
a null (ie %1\%) and the input parameter is also a null.
This prevents various PIP like commands from occuring with
no files specified which often results in defaulting to the
entire directory.
11) Two possible TABLE macro files are offered. TABLE.BIG has all
kinds of CCL commands and results in a significantly larger
CCL. TABLE.KMS includes only those commands which used to
be processed by the internal DCL code. If the PIP-like
internal CCL commands are not selected for assembly and the
TABLE.KMS commands are selected, CCL stays about the same
size and has identical functionality. TABLEGEN.CMD allows
the user to create an internal CCL file.
12) CCLGEN allows the user to select amoung the ever widening
variety of options. If the user desires to use an internal
table, CCLGEN will allows the user to link to TABLEGEN.CMD
to create an internal table from a standard CCL file
(USERCCL.CCL or SYSCCL.CCL for example).
13) CCL now checks to see if the keyword typed in is longer than
the maximum specified in the file (previously was ignored). This
means that commands FOO and FOOY can be different and distinct.
14) Documentation and HELP files have been updated to reflect
the many, many new features
[344,065] account contents
This UIC contains updates to the SYSLOG accounting system.
More elaborate reports now exist. ACCREP is amost suitable for a
charge back accounting system. SYSLOG now outputs a special flag shortly
after midnight so the system will know when one day ends and another
begins. This packet enables LOGTSK to automatically append a day's
data to a monthly backup file and only keep the current day's data
in SYSLOG.DAT.
[346,100] account contents
THIS ACCOUNT HAS TWO SUBMISSIONS. FIRST IT HAS A GOOD SUBMISSION OF THE
PREVIOUS LOADABLE XDT. THIS ONE HAS ALL KNOWN PROBLEMS FIXED.
SECOND, IT HAS COMBINED DIRECTORIES OF THE PREVIOUS SEVEN SIG TAPES AND AN
ABSTRACT OF THOSE TAPES. THE FILES ARE AS FOLLOWS:
TAPESB.RNO - ABSTRACTS OF PREVIOUS TAPES
TAPESB.MEM - OUTPUT FILE OF TAPESB.RNO
RSXSIG.UIC - LIST OF ALL UIC'S ON ALL TAPES.
RSXSIG.FIL - LIST OF ALL FILES ON ALL TAPES, SORTED BY FILENAME.
RSXSIG.TYP - LIST OF ALL FILES ON ALL TAPES, SORTED BY EXTENSION.
RSXSIG.COM - LIST OF ALL FILES ON ALL TAPES, SORTED BY UIC.
***************************************************************************
editor's note: This index is a fantastic piece of work. I have used it
extensively and find it invaluable. The first thing you should do with this
tape is print the file TAPESB.MEM. note the file TAPESB.RNO is in
DEC Standard Runoff (DSR) and the runoffs we have give "illegal command
error messages". DSR has more powerful indexing features and Ralph put
them to good use.
Do not overlook the usefullnes of RSXSIG.COM. It contains a sorted
directory of EVERYTHING. Print it if you like, but I keep it on-line
and use TECO to find things in it. Ralph what can I say ...
Thank you so very much
Philip Cannon
******************************************************************************
[346,101] account contents
This submission arose out of the pre-symposium seminar
on RSX-11M crash dump analysis. A major item of interest
is the seminar workbook, which is a written version of the
material covered in the seminar. The appendixes to the seminar
workbook discuss several of the other files. Another appendix
of interest is a description of RSX data structures, including
contents, usage, and a cross reference of other structures which
point to this one. The CRAxxx programs represent "homework"
from the seminar. These programs crash RSX systems in known
ways. The resulting crash dumps can then be used as references.
CRASHW.RNO;3 Workbook from CDA seminar
*******************************************************
CRASHW.DOC
editor's note: The author submitted the file
CRASHW.RNO which I believe to be in DSR
(DEC Standard Runoff) format. I used DSR on
a VAX to create the CRASHW.DOC file.
*******************************************************
[352,001] account contents
DISKUSAGE.FTN Produces a table summarizing disk space usage by UIC
(number of files, blocks used and allocated), totals
for each group, and compares the total group allocated
space with an assigned disk space allotment value. It
reads a file DISKUSAGE.DAT containing the disk unit/
group/allotment numbers, and a file TOTALBLK.DAT which
is a PIP /TB listing file. Written in F4P. Please see
the comments at the beginning of the file and within
DISKRPT.CMD and TB.CMD for more information.
DISKRPT.CMD Reads the DISKUSAGE.DAT file mentioned above, creates
TB.CMD and executes a PIP command file to create TOTALBLK.DAT,
and executes the DISKUSAGE program. It executes TB.CMD.
It renames the input and output files to TByymmdd.DAT and
TByymmdd.TBL where yymmdd is the current date. Examine
both command files and change the UIC assignments for your
system. Note that the data file and task file are expected
to be in the same UIC.
CONSOLE.FTN Creates a copy of CONSOLE.LOG files with repeated records
replaced by a message indicating the number of occurences
of the record. Good for eliminating many device NOT
READY messages before printing these files. The user may
merge multiple file versions into the output file. Written
in F4P.
CONSOLBLD.CMD TKB command file to build CONSOLE.TSK.
FT.FTN A tape utility program for determining tape record sizes,
copying tapes, etc. This program is written in F4P
and is not structured well. No doubt better utilities
exist, but this has been helpful to us at times. I had
a request to submit this program to this tape. FT needs
to be built with the conversion/byte manipulation
routines mentioned below. In addition, it also needs
MTBLOK.MAC, which appeared on at least two RSX SIG tapes
several years ago.
FTBLD.CMD TKB command file to build FT. Note that UIC and library
specifications need to be modified for your system.
FT.DOC FT program documentation.
TAPCVT.FTN Another tape utility program which I have not used. See
the documentation for information.
TAPCVT.DOC TAPCVT program documentation.
FPCONV.MAC Routines to convert between IBM 360/370 and PDP 11
floating point format.
SWABI.MAC Byte swapping routine.
DAYFLD.FTN Convert from day of week (numeric) to ASCII day of
week.
DAYWK.FTN Convert from day of year to month, day, and day of
week; or from month, day to day of year and day of
week.
BINHEX.MAC Convert to hexadecimal.
EBCASC.MAC Convert a string from EBCDIC to ASCII.
ASCEBC.MAC Convert a string from ASCII to EBCDIC.
MOVE.MAC Routines to move a string of bytes, to fill a string
with a byte, and to convert a string to upper case
ASCII.
FPCONV.DOC Documentation for the above conversion and byte
SWABI.DOC manipulation routines.
DAYFLD.DOC
DAYWK.DOC
BINHEX.DOC
EBCASC.DOC
ASCEBC.DOC
MOVE.DOC
FILL.DOC
CAPS.DOC
HEL.COR SLP correction file to allow a privileged user to
log on in a special way even if logons are disabled.
Creates HELNEW.MAC, which you may MAC and TKB
appropriately (preferrably with a non-ANSLIB SYSLIB).
Good if you log off by mistake after a SET /NOLOGON.
[352,002] account contents
This UIC contains files for the TALK program.
TALK is a terminal emulation program which enables a user on an RSX-11M V3.2
system with the full duplex terminal driver to communicate with a remote
host computer over a single line interface (such as a DLV11 or DL11). The
program begins in terminal mode, and has the capability of bi-directional
file transfers. See the documentation and program source for details.
TALK is currently being worked on, but we expect to make only relatively
minor modifications to this version as we continue testing. Note that
requesting a function menu (with ^W) does not currently work for a VT100
in smooth scroll mode. The documentation is preliminary. It closely
reflects the current state of the program, but has not been completely
updated and reviewed. I plan to submit an updated version of these files
to the RSX SIG tape in the fall.
Kindly advise me of any changes you make to the program.
TALKPRE.MAC TALK prefix file.
TALK.MAC TALK source file.
TALK.CMD Command file to assemble and build TALK.
TALK.DOC TALK documentation.
TALK.RNO Runoff input file for TALK documentation.
[352,003] account contents
XMDRV.MAC is a DMC-11 device handler for RSX-11D. See the comments at
the beginning of the file for information.
[357,001] account contents
Contains an assortment of changes to RSX-11M V3.2.
Each change should be evaluated on its merits for a given
site. All changes have been in use but of course are not
guaranteed. A good system manager takes a hard look at each
possibility and either uses it directly or as a hint for
another change. The following are included:
PAT.CMD is an indirect file which helps to make object
changes from the Dispatch articles.
CORAL.COR, INITL.COR, IOSUB.COR, SYSCM.COR,
SPROVFDT.COR, and MCRDEF.COR are correction files which im-
plement pre-allocated I/O packets and only retain the
pre-allocated ones on I/O termination. Most of this is due
to Steinberg's submissions on the last DECUS tape.
HELLO.COR is a minor change to the same file from the
Spring tape. There was a problem with 'matchall' characters
in lower level help sections.
INDNP1.COR removes some unused messages from the 11M
version of Indirect.
INDOPN.COR makes Indirect non-privileged during file
opens (so users can't write/append files that would normally
be inaccessible.)
INDSU2.COR prints MCR lines if quiet and tracing.
Double tracing is also eliminated for .WAIT, .IFINS, .IFACT,
...
TKTN.COR is changed to not print task registers if a
load failure occurred.
LBRLI.PAT modifies the listing format to allow for the
SYSLIB date routine which was changed from 3.1 to 3.2.
P5CEL.PAT modifies TKB CREF output when a symbol is de-
fined in the root and referenced by up-tree elements. The
value was formerly not relocated if the task was privileged.
SLPR11.PAT modifies the SLP page format to center the
listing thereby making it easier to bind the output.
SLPSB.PAT corrects a problem with the placement of the
delete audit trail (;**-n) in lines which are longer than
the audit position and have tabs in the last 'n' characters.
HELP.ULB is a complete copy of my help files. Many of
these originally came from J. Downward's contribution a few
SIG tapes ago. Many new ones have been added and many er-
rors have been corrected. Since we run a single user sys-
tem, some of the multi-user comments have been deleted.
Several BRU patches:
BRUREAD.PAT corrects the error recovery procedure on
TU10/TE10 drives (originally tried to do a read reverse fol-
lowed by a skip record).
COMLIB.PAT prints the tape label as part of the direc-
tory listing.
TAPEIO.PAT corrects a line in dispatch article
5.1.17.16 which caused BRU initialized discs to appear to be
created on the date of the backup but the time of the re-
store.
[360,001] account contents
README files for the A.R.A.P. Submissions to all sig tapes.
[360,200] account contents
ARAP.OLB, needed for all following software
Only new source modules included, but library contains
all modules from previous tapes also.
[360,212] account contents
IAS/VAXNET - This is a simple-minded virtual terminal
program with the ability to send/receive ASCII files
between two IAS systems, or an IAS and a VAX system, over
asynchonous lines (DZ11 like devices). This originally
appeared on the VAX SIG tapes and was modified to work on
IAS as the host system. Simple yet effective. Refer to
NETBLD.CMD to build the two programs that run on the local
and remote systems. See also VAXNET70.CMD. MEMO11.RNO
describes the steps needed to actually transfer files.
[360,213]account contents
UND - This program finds undefined variables in FORTRAN
programs. It subtasks to the FORTRAN compiler and uses
a /LI:2 listing file. It will report undefined variables,
variables appearing in subroutine calls that are not
explicitly defined, etc. Refer to UNDEF2.RNO, UNDEF1.RNO and
UNDMN.FTN for further instructions. UNDTEST.FTN is an example
test file for UND and UNDTEST.LST is the resultant output.
[360,214] account contents
C36 - conditional fortran pre-processor
C36 is used by ARAP to convert fortran source to both
IBM and CDC machines and is specific to our operations.
However, C36 also has the ability to conditionally
include or exclude lines of code by using a logical
definition file and IF-THEN-ELSE types of constructs within
the code to be converted. It is not a process which needs to
be run before every compile. Rather, it lends to maintaining
a base line version of a program on a development machine,
while allowing easy conditional conversion for specific
exportation to other machines which require different
coding algorithms in selected places.
Refer to C36.RNO for documentation and also C36MN.FTN.
[360,215] account contents
DSM from spring 1980 SIG tape - only updated modules included.
A couple of bugs have been fixed pertaining to disk drive
size recognition and maximum number of uic's checking.
DSX - Also included on this uic is a somewhat modified
version of DSM which is used to associate a specific
file to a given disk cylinder/track/sector address. This is
useful for finding the files in which disk errors occur.
Refer to DSX.RNO for further description and use.
[360,230] account contents
SFS fortran cross-referencer from Spring 1979 SIG tape.
Fixed to identify VIRTUAL statement.
Only modified modules included.
[360,245] account contents
SPY from fall 1979 - only updated modules included.
Bug fixed - problems when real time tasks alter priority
to below 100.
[ARAP.RUNOFF] account contents <note this account is NOT on the tape, the author
could not get it together in time>
ARAP modified Triangle RUNOFF. ARAP has added a number of
features to RUNOFF to enable us to produce technical reports.
We required support for greek characters, extended sub/super
script capability, ability to produce very complex
mathematical equations etc. I was unable to put together a
formal contribution of ARAP RUNOFF in time, but would like to
point out some of our extensions to anyone interested.
I might add that we have successfully been using RUNOFF in
conjunction with a screen editor to produce reports on a
production basis. Some of the features include:
a) Support for Greek/Math character set. Output is produced
on a DIABLO 1640 in two passes. Lines up amazingly well.
b) Super/subscripts adjustable in 1/8 line increments. Can
set to super/sub 3/8 line instead of 1/2 and change
amount througout document as needed.
c) Line spacing adjustable to 1/8 line (.SP) and also
.BLANK can use 1/8 spacing. e.g. .BLANK 1+5/8
d) Overbar function, implemented similar to underline.
Adjusts to highest superscipt to be overscored.
e) Equation mode command. Allows reasonable description
of equations, set spacing around equation, sets line
spacing (to 1/8), allows '#' to be horizontal half-space
for aligning 1 character above 2 (a/bc) for example.
f) Justified output uses a horizontal half-space to fill.
g) Overstrike ability using backspace, justifies properly.
h) String substitution definitions. Similar to the
numeric substituion, but allows strings within the text
to be identified and substituted for.
e.g .SUB ARAP = 'Aeronautical Research Associates of P'
Any occurence of ARAP within text will be substituted.
Please call me if your are interested in using RUNOFF for
this sort of work. I will be happy to send an example of
output, send a tape etc.
**************************************************************
editor's note: This looks pretty useful, I would like to
get a copy for the next sig tape.
**************************************************************
[370,001] account contents
TALK.MAC - TALK IS A INTER-TERMINAL COMMUNICATION PROGRAM THAT
ALLOWS SIMULTANEOUS CONVERSATION BETWEEN TWO VT52 TERMINALS.
TAPE.MAC - TAPE.MAC IS THE START OF A GENERAL PURPOSE TAPE COPY PROGRAM
THAT WILL ULTIMATELY USE SWITCHES TO SELECT DENSITY, MODE, AND
ACCEPT INDIRECT COMMAND FILE SPECIFICATIONS FOR FILE NAME OF
UNLABLED TAPES. IF YOU HAVE ANY COMMENTS OR SUGGESTIONS ON THIS
PROGRAM PLEASE FORWARD THEM TO THE ABOVE ADDRESS. CURRENTLY THIS
PROGRAM WILL READ UNLABELLED EBCDIC TAPES AT 1600 BPI. INDIRECT
COMMAND FILE SPECIFICATION OF FILE NAMES IS ACCEPTED ALSO.
EBASC.MAC IS A CONVERSION ROUTINE THAT IS USED WITH TAPE.MAC.
UP1.FTN - A FORTRAN PROGRAM THAT USES THE DATA PRODUCED BY THE PROPRIETARY
SRF (SYSTEM RESOURCE FACILITY) FROM GEJAC TO CREATE A HISTOGRAM
OF CPU UTILIZATION VERSUS TIME. SRFMAINT.CMD IS USED IN
COMBINATION WITH THIS.
WP.CMD - A TEXT PROCESSING SYSTEM BASED IN EDT V2.0 AND RUNOFF. THIS
SYSTEM IS MENU ORIENTED AND IS USED ON A VT52 OR VT100 TERMINAL.
THIS WAS A SYSTEM THAT WAS THROWN TOGETHER IN ONE DAY TO COVER
THE ACTING DP MANAGERS REAR END. (NOTE: THIS IS THE THIRD ACTING
DP MANAGER .... THE LINE FORMS TO THE REAR, PLEASE!).
PLOT.FTN - A PLOTTING ROUTINE FOR THE DIABLO 1750 LETTER QUALITY PRINTER
PLOTTER. 3D.FTN WILL PLOT A 3D FUNCTION USING PLOT.FTN AS AN
EXAMPLE. WE WROTE THIS IN ONE DAY.
WHOTT.MAC - A TASK THAT WILL RETURN THE TERMINAL NUMBER OF THE REQUESTING
TERMINAL AS THE EXIT STATUS <EXSTAT> IN INDIRECT COMMAND FILES.
(NO BIG DEAL, BUT USEFULL).
TSHEET.FTN - A PROGRAM TO KEEP TRACK OF WORK TIME AND PRINT OUT TIMESHEETS.
PUN.FTN IS USED TO PUNCH IN OR PUNCH OUT. A FILE TIME.CLK,
WHICH IS EDITABLE IS USED TO STORE TIME INFORMATION.
[370,020] account contents
The software contribution from the Washington, DC area LUG consists
of the following programs:
1. PIPLST - A system of command files and TECO macro's that reformat
directory listings into a 5 column format along with some summary
information. Uses PIP to create directory, then TECO and FLY to
mung the listing around.
2. PUG - A nice little task to find zero length and locked files.
Program will print out the names and locations and will optionally
delete ALL of them. Be sure that this what you want it to do.
3. GAB - A half duplex terminal connector. Everything you type will
show up on both terminals until the next CTRL C, then everything the
other person types will show up on both terminals. A great replacement
for BRO. Use CTRL Z to exit hookup.
4. DIARY - Create a timestamped diary of any thing that you care to
type in.
The infor is appended to SY:[default]DIARY.TXT.
5. FILES - Program to determine which file a specific block belongs
to.
6. TIMIT - Set of programs and subroutine for RSX11M-PLUS to do account
file report generation, user accounting by login and individual TASK
accounting. Requires F4P brcause of include files, I*4 support and parameter
definitions. Conversion to FORTRAN should be doable, but not simple.
Programs:
ACLIST - List accounting history file in formatted summary fashion
Depends on the Accounting history files being named
"ACddmmmyy.DAT", command file to start up accounting
and COLOG with this format filespec. Another program
ACRESET - runs at 2400 to stop accounting and re-start with
the proper name for the next day.
GETAB - Grab users' Accounting block format and print out
information. Written as test system to allow allocation
of users activity to many different job orders in an
automated way by writing out the info into different files
rather than to just the terminal.
TIMIT - Grab accounting info before and after some operation in
an attempt to regain some task accounting capabilities for
RSX11M-Plus. Works very well within the limitations of the
information that M+ keeps for us.
[370,140] account contents
MGT IS A MAG TAPE TRANSFER PROGRAM THAT ALLOWS A USER TO
TRANSFER DATA FROM MAGTAPE TO DISK,MAGTAPE TO MAGTAPE,MAGTAPE
TO TERMINAL,DISK TO MAGTAPE ETC.
MGT DOES A BLOCK DATA TRANSFER SO NO DATA TRANSLATION IS
PERFORMED. IT CAN BE USED TO COPY ANY TAPE. IT ALSO HAS SOME
NICE BLOCKING FEATURES.
[370,150] account contents
* HEX - Microprocessor hex file management utility *
The HEX utility runs under the RSX11M operating system. Its purpose is to
manipulate INTEL ASCII Hex-formatted files that are produced by many cross
assemblers (or linkers) for 8-bit microprocessors (8080, Z80, 8085, etc.).
Operations that can be performed on a hex file with HEX include:
FILL - fill address range with a value.
READ - read in a hex file.
COMPARE - compare specified addresses to another hex file.
COPY - copy values from one area to another.
CRC - compute the standard bisync cyclic redundancy check on a range.
DISPLAY - display portions of the file to the screen or to a listing file.
EDIT - examine and optionally replace values in the file.
MOVE - move values from one area to another.
SEARCH - scan the file for a particular byte or word value.
SUM - compute 16-bit byte-wise sum-check (as with DATA I/O prom burner).
TRANSFER - manipulate the transfer address.
WRITE - write out locations to create new hex files.
[370,170] account contents
A MACRO SUBPROGRAM TO CONVERT CLUNKS TO DATES AND DATES TO CLUNKS
The module CLUNK.MAC can be used to convert Datatrieve-11
clunks to system date and time and back again. It can be called by
a Macro, Fortran, or Basic-Plus 2 program.
This routine was inspired by the article on Datatrieve dates
by Chuck Watson in the Feb. '81 issue of the Wombat Examiner.
I saw it as desirable to have a application program using the FMS
package access a data file created by Datatrieve, hence this routine
came about.
[373,001] account contents
This account contains a version of the Universal Magtape Reader
origiannly written by D. Macre in DEC Paris. A number of changes
have been made to make it more usable.
[373,002] account contents
This account contains the disk fragmentation utility originally
based on FRAG by Michaud. I have made a few cosmetic changes,
such as allowing lower-case, and allowing the device to be
[373,003] account contents
This account contains a version of FRC, the utility to force a
command line to another terminal. It has been modified from
William Storey's original design to use the Spawn directive, and
so does not have to map the Executive.
*******************************************************************
editor's note: This uic contained the two files README.1ST and
README.DOC. I decided to combine them into one README.1ST file.
I changed the name of the README.DOC file to SRDMOD.DOC.
This account contains Version 6.0 of SRD. It is an attempt to
combine the modifications made by Bob Turkelson (Goddard Flight
Centre) and Ray van Tassle (Motorola) as well as to add a number of
new ones. For a detailed writeup on the changes made read the
following section (which is the author's original README.DOC file
now appearing as SRDMOD.DOC).
**********************************************************************
This is an improved version of SRD, dated 3-DEC-1980.
This version is intended to combine all the versions of SRD I know of
(Ray van Tassle's, Bob Turkelson's and my own) into one (easily
tailorable) unit, which can then hopefully be used as a base for
future modifications. The changes from 'standard SRD' fall into
three categories, as follows:-
1. Those made by:-
Ray Van Tassle
Motorola
It has been sporadically updated over the last year and a half, so
long that I forgot what it's original base was. I think it was
the version distributed with our IAS V 2.0 system.
The original SRD had a massive number of bugs, the worst one was
probably that it would loose files on a Write-back (but mostly only
on very large directories) !!!!!!!
This version has had all the bugs I could find fixed, plus a lot of
enhancements.
Some of the key things are:
Write-back works properly, and it also re-tries if it gets an I/O
error on the write.
It will dynamically expand it's core buffer (via EXTSK$
directive) if the directory won't fit.
If the entire directory won't fit, it will do it in chunks, each
of which is sorted properly; along with a warning message.
A trailer line for each directory is given, showing the number of
files in the UFD, and the number of files selected. If a
switch that causes the file header to be read is specified
(/MI, /FU, or a date switch), the number of blocks
used/allocated is also given.
A "nolist" switch is added, so you can get the above trailer
info, without getting a list of the individual files (very
handy for scanning an entire disk).
The user write-up has been re-written.
On date select, the "before" and "after" dates can now both be
specified.
The format of the output has been slightly modified so that a
simple TECO macro can make a PIP command file to copy (or
whatever) the selected files. Two macros to do just that are
also on this tape, in my UFD for TECO macros. To use them, do:
SRD XXX.CMD=????
TEC XXX.CMD
*EISTRIP$$
*EIGET$$
* (any other edit you may want to do)
*EX$$
2. Those made by:
BOB TURKELSON
NASA/GODDARD SPACE FLIGHT CENTER
Dave Parker, Henry Tumblin
Duke University Medical Center
Support for /CO/LO/MU switches (to select contiguous, locked, or
multi-header files).
Support selection of files created between two dates when both
/BE and /AF are used in the same command line.
Yes/No/Quit/Go responses for selective delete (obtained from
another version of SRD).
Support /OV:[n] switch to select obsolete versions of files,
which are those files which would be deleted if the /PU:[n]
switch were used, and those which would not be selected if the
/SV:[n] switch were specified. This provides a convenient
method for users to see their obsolete files, and for system
managers to determine which users are not regularly purging
their files.
Fixes for a few bugs (including the one reported in the October
1979 Multi-Tasker).
Remove the requirement that a period must be included in the /SE: