Mail Delivery Problem
SMTP MAILER
postmaster at sandia.gov
Thu Mar 29 04:24:28 AEST 1990
----Reason for mail failure follows----
Sending mail to host ceratiidae.cs.sandia.gov :
Fatal reply code to command 'RCPT TO:<jnjortn at ceratiidae.cs.sandia.gov>':
554 /home/u/jnjortn/.forward: line 1: jnjortn at chaos... Never heard of chaos in domain cs . sandia . gov
----Transcript of message follows----
Date: 28 Mar 90 09:33:00 MST
From: info-unix at BRL.MIL
Subject: INFO-UNIX Digest V9#104
To: "jnjortn" <jnjortn at ceratiidae.cs.sandia.gov>
Return-Path: <incoming-info-unix-request at sandia.gov>
Received: from SEM.BRL.MIL by sandia.gov with SMTP ;
Wed, 28 Mar 90 08:41:22 MST
Received: from SEM.BRL.MIL by SEM.brl.MIL id ab06484; 28 Mar 90 5:58 EST
Received: from sem.brl.mil by SEM.BRL.MIL id aa06480; 28 Mar 90 5:46 EST
Date: Wed, 28 Mar 90 05:45:42 EST
From: The Moderator (Mike Muuss) <Info-Unix-Request at BRL.MIL>
To: INFO-UNIX at BRL.MIL
Reply-To: INFO-UNIX at BRL.MIL
Subject: INFO-UNIX Digest V9#104
Message-ID: <9003280546.aa06480 at SEM.BRL.MIL>
INFO-UNIX Digest Wed, 28 Mar 1990 V9#104
Today's Topics:
Setting up slip
Displaying 8-bit characters in vi...
UNIX front end for DOS -- remote!
pipe() doesn't seem to work. (Pyramid OSx 4.4) (Summary)
Re: Awk-compiler
LU6.2 wanted for 386 system
Re: Default switches for lp?
C-Shell/rusers problem
Allowing users to remove files in /usr/spool/at
Re: How can I get the Frequently Asked Questions?
Re: AT&T source license
Re: Programmers Prefer 8.5 x 11
PC/IX
Archive Question
Re: Archive to tape software
Re: 4.3 System Administration Training Courses
Re: RCS: how do you freeze a configuration?
Problems with diff -e and ed?
Re: ANSI header files
Re: How to set TERM for Televideo?
``Office Automation'' under Unix?
Re: File server security
gas
Re: termcap or terminfo?
Availability of Korn Shell
-----------------------------------------------------------------
From: Nick Sayer <mrapple at quack.uucp>
Subject: Setting up slip
Keywords: Sun 2/170 & 3/??
Date: 26 Mar 90 07:49:49 GMT
To: info-unix at sem.brl.mil
I have a Sun 2/170 and have recently gotten slipware
for it. At school we have a couple sun3s and a vax on
an ethernet, and will soon be hooked up to the internet
(via a slip link from the vax, but that's not important).
The sun 2 is quack, and the sun 3 I have been trying to
connect to is uop. quack is 138.9.100.1, and uop is
138.9.200.1 (quack's address is arbitrary - in future
it may need to be changed to something totally different
for various reasons). Using a 2-3/3-2 cable and two serial
ports on quack, I have gotten telnet to talk to itself
using the instructions in the slip 4.0 README, thus:
quack% sliplogin 138.9.100.10 138.9.100.11 > /dev/tty03 < /dev/tty03 &
quack% sliplogin 138.9.100.11 138.9.100.10 > /dev/tty04 < /dev/tty04 &
quack% ping 138.9.100.10
138.9.100.10 is alive
... etc
slip has been used in the past at uop, so I did not perform a loopback
test there.
I am trying to hook the two machines up via slip on a non-permanent
basis. I.e. If I want to do some ftping from the internet (when
they get it up), it's seemingly more convenient to "slip"
into uop, then ftp directly into quack (over the phone) than
to ftp to uop, then kermit to quack, or some such nonesense.
I am familiar with unix, but just starting TCP/IP. I plan to hook quack
up to Amateur Packet Radio TCP/IP, but that's going more slowly.
I have tried this:
uop% sliplogin 138.9.100.4 138.9.200.4 255.255.255.0
~Clocal command: sliplogin 138.9.200.4 138.9.100.4 255.255.255.0
If I try to ping ..200.4, the packets go out, but they never
come back.
I'm obviously a confused soul. The manuals I have read have explained
the theory of TCP/IP, and I understand things, but how these things
are implemented, and how to make them work is obviously more difficult
than theorizing about them.
Thanks in advance if anyone can help me out.
-----------------------------
From: Paul Gillingwater <paul at actrix.co.nz>
Subject: Displaying 8-bit characters in vi...
Date: 26 Mar 90 10:21:25 GMT
To: info-unix at sem.brl.mil
Hi there,
I've got ISC 386/ix 2.02, and have been using vi with some files
that contain extended characters (8th bit set). These appear on
vi as e.g. \272. How can I force vi to use the extended characters?
Can I have a snappy answer to this one please, i.e. same day, as I
have to show this to a customer tomorrow morning! Note my local time
is GMT-12 hrs (New Zealand).
--
Paul Gillingwater, paul at actrix.co.nz
-----------------------------
From: Paul Gillingwater <paul at actrix.co.nz>
Subject: UNIX front end for DOS -- remote!
Date: 26 Mar 90 10:45:14 GMT
To: info-unix at sem.brl.mil
I've been planning a program to hone my skills using my new Zortech
C++ 2.0 compiler (MS-DOS). I use UNIX quite a lot, but mostly from
remote modems, and usually with terminal emulators on PC's. I also
run a public access UNIX system, and was thinking of ways to add value
for my user base, when I thought of the following idea for a program...
Hit 'n' now if dreams bore you...
My program would run on an MS-DOS machine, as well as OS/2. It may
also be ported to UNIX, although this is less likely. Amiga DOS is
another possibility, depending upon compilers.
It would use lots of C++ features -- the best way to learn is by doing,
in my experience. Extensive use will be made of windows and color,
in a way that tries to assist rather than obscure.
Essentially, the program will insulate the novice user from a UNIX
shell running on a remote machine. The connection can be made via
a comms package such as Procomm Plus, Kermit or Telix, and then the
user can shell out to run my front end. A windowing environment would
then be available, which would use simple WIMPS to provide much of the
interactive shell functionality. One window would show the UNIX
commands which are then generated and sent to the remote system to
achieve the objective. A history of these commands will be maintained,
which can be edited and executed at will, thus adding command history
even to older Bourne shells. A secondary benefit is that the user
would see the commands forming as they go, thus learning UNIX syntax.
When started, the shell (running under DOS) would try to learn what
it can about the UNIX environment which the user is connected to.
If not on-line, a simple dialling directory and login script can
be executed. Useful information would be requested from the remote
system, such as pwd, id, logname, $HOME, $SHELL, $EDITOR, date,
ls *, files *, etc. This would be used to build a friendly
front end, with files sorted into functional groups, e.g. shells
scripts, data, binaries, etc.
The F1 key would provide help at all times. In particular, it will
be highly contextual, although keyword indexing of the help information
will be supported. In addition, it will recognise events, e.g. if
another user executes a 'write' to the user's screen, the shell
will pick this up, and notify the user. Because many events will
run asynchronously, (e.g. notification of job completion for background
jobs, chat requests, mail notification, etc.) the shell will continuously
monitor streams of output from the UNIX host, looking for keywords
(of course this is configurable via macros, allowing the advanced
user to modify the behaviour).
The stty command will be well emulated, e.g. for things like baud rate,
parity, echoing, interrupt and other signals, etc. A later version
will support shell layers, by using windows (locally).
File transfer will be supported (Kermit, [XYZ]modem), using such
excellent products as DSZ and rzsz to implement them. ASCII
file transfers will be supported too, a la cu.
-----------------------
I have lots more ideas, and have started planning the screens, as
well as the objects for representing the data. What i'd like to do is
start some dialogue -- maybe these ideas have already been developed,
so i don't want to re-invent the wheel. In case this is new work,
I hereby Copright 1990 this message. You may comment on it, but
don't rip off the ideas and sell them as your own.
Copyright (C) 1990 -- Paul Gillingwater
--
Paul Gillingwater, paul at actrix.co.nz
-----------------------------
From: "A.V. Raman" <ARaman at massey.ac.nz>
Subject: pipe() doesn't seem to work. (Pyramid OSx 4.4) (Summary)
Date: 26 Mar 90 23:41:47 GMT
To: info-unix at sem.brl.mil
For the sake of other readers who may be interested, I'm restating the
problem here and posting a summary of the replies I received.
---- Faulty Code begins ----
#include <stdio.h>
setup()
{
int fd[2];
if (pipe(fd) < 0) { /* write to fd[1] and read fd[0] */
perror("pipe");
exit(1);
}
if (fork() == 0) { /* set up child to read pipe as stdin */
close(0);
dup(fd[0]);
close(fd[1]);
execl("/bin/cat","cat","-n",NULL); /* ucb cat -n to number lines */
perror("exec");
} else { /* set up parent to write to pipe as stdout */
/**/ printf (" "); /* program fails if this line is removed */
close(1);
dup(fd[1]);
close(fd[0]);
}
}
main(argc,argv)
char **argv;
{
FILE *fp;
char buf [BUFSIZ];
if (argc == 1)
exit(1);
setup(); /* setup pipe with cat reading other end */
if ((fp = fopen(argv[1],"r")) == NULL) {
perror(argv[1]);
exit(1);
}
while (fgets(buf,BUFSIZ,fp) != NULL)
printf("%s", buf); /* write into pipe */
fclose(fp);
close(1);
return 0;
}
---- Code ends ----
Thanks for all the prompt solutions that I received for the problem.
The fault in my program narrows down to my mixing buffered and unbuffered
IO incorrectly.
Reproduced with the permission of Cuong T. Nguyen of Center for Integrated
Systems, Stanford University:
>The problem is with stdio buffering, which can be none, line,
>or block. When the program is started, stdio is typically fully
>(block) buffered. When your first printf() is called (actually,
>_flsbuf()), IF stdout is a tty, buffering is set to line. Otherwise
>it is left alone. That's why the null print(" ") insertion helps,
>and also why write() doesn't, since it is not part of stdio.
>
>When your printf's are fully buffered, anything left unflushed
>in stdout before the pipes close is lost. The simplest thing
>to do is an fflush(stdout) at the end. Else you can do something
>a la setlinebuf() to force the buffering type you like.
- Anand
-----------------------------
From: Jim Harkins <jharkins at sagpd1.uucp>
Subject: Re: Awk-compiler
Date: 27 Mar 90 02:08:24 GMT
To: info-unix at sem.brl.mil
In article <338774 at neabbs.UUCP> aerts at neabbs.UUCP (AD AERTS) writes:
>
>Has anyone ever seen an AWK-compiler? AWK-to-C would be fine too.
Mail bounced so...
Look into the document "A supplemental document for awk" subtitled "things
Al, Pete, and Brian didn't mention much". I snarfed it off the net a few
years ago and it's probably archived somewhere. Anyway, it says
-- begin direct quoting --
2.5 Script "Compilation"
[It is likely that this does not work at most sites. If it does not, the
following will probably not be of interest to the casual user.]
The command
awk -S -f script.awk
produces a file named awk.out. This is a core image of awk after parsing
the file script.awk. The command
awk -Rawk.out datafile
causes awk.out to be applied to datafile (or the standard input if no file
is given). Unfortunately, the way this is implemented requires some special
action on the part of the person installing awk.
-- end quoting --
It's a good document, and I suspect I'm going to get bazillions of requests to
post it :-)
--
jim jharkins at sagpd1
We are all aware of the high cost of alcohol abuse. To help solve this problem
take this signature to your local liquor store for $1.00 off your next purchase.
-----------------------------
From: Paul Gillingwater <paul at actrix.co.nz>
Subject: LU6.2 wanted for 386 system
Date: 27 Mar 90 02:59:46 GMT
To: info-unix at sem.brl.mil
Hi there,
I work for HP in New Zealand. We're offering a deal to a customer
that wishes to buy 14 UNIX boxes (386 and 486). They have a choice
between SCO Xenix, SCO UNIX and ISC 386/ix 2.02.
The customer wants to connect to their IBM mainframe, which requires
LU 6.2 for SNA using SDLC. They wish to be able to operate a number
of host sessions as well from the UNIX, via 3270 emulation, as well
as running RDBMS applications that use an API to access remote
data on the mainframe.
I know that there is a proprietary HP solution that runs under HP/UX,
however this customer wants to go with 386 boxes. Does anyone know
of an LU 6.2 API solution for 386/486 UNIX?
Please e-mail and I will summarise.
--
Paul Gillingwater, paul at actrix.co.nz
-----------------------------
From: Terry Lyons <terry at pride386.uucp>
Subject: Re: Default switches for lp?
Date: 27 Mar 90 07:31:53 GMT
To: info-unix at sem.brl.mil
In article <113 at dynasys.UUCP>, jessea at dynasys.UUCP (Jesse W. Asher) writes:
> I'm tired of
> typing in -onobanner everytime I want to print something because I
> don't want banner pages. I don't need them. So, is there a way
> to set something that it will not do this by default (both system
> and/or user)?
look for /etc/default/lp or whatever looks close on your system
then change banners=1 to 0
or edit the lp/interface/(printername) file and force banners to be 0
OR
make sure your PATH looks at $home/bin first and put something like this in
your bin:
1. make an executable file named lp
2. put the commands in that your tired of typing e.g.: /bin/lp -obanners
-ocompressed -dlazer
and away you go...
--
**************************************************************************
* UUNET ...!pride386!terry * FAX (714) 739 - 2203 *
* Pern is a dragons best freind *
**************************************************************************
-----------------------------
From: "-----> Dave" <komatsu at alcor.usc.edu>
Subject: C-Shell/rusers problem
Date: 27 Mar 90 08:12:20 GMT
Sender: news at chaph.usc.edu
To: info-unix at sem.brl.mil
This should be a fairly simple syntactical question. I am writing a
C-Shell and I want to find a user and keep searching off and on until
I find that person.
---
# C-shell program
while (`rusers alcor | grep -w $1`)
echo $1
sleep 100
end
echo "User has logged in"
---
Usage is: find user
and this should hunt the system alcor but somehow it doesn't work
any ideas.
---------------------
komatsu at alcor.usc.edu
---------------------
-----------------------------
From: Andy Clews <andy at syma.sussex.ac.uk>
Subject: Allowing users to remove files in /usr/spool/at
Date: 27 Mar 90 11:10:20 GMT
To: info-unix at sem.brl.mil
I work on a Sequent Symmetry S81 running DYNIX 3.0.15. My question
could, however, apply to any UN*X box.
I recently set up a command (atq) which lists, in a `friendly' format, the
jobs waiting in /usr/spool/at (i.e. files produced by the "at" command).
I would like to expand the command to enable users to cancel their
own jobs - i.e. remove their own files in /usr/spool/at.
Files in this directory are owned and writeable by the people who gave
the corresponding "at" command, but because /usr/spool/at can only at
present be written by root, users cannot delete their own "at" files.
Can anyone suggest a way in which I can give this sort of access to
/usr/spool/at without compromising system security by allowing the world
to write to it? I'm a bit wary of creating a setuid'd program to do it.
--
Andy Clews, Computing Service, Univ. of Sussex, Brighton BN1 9QN, England
JANET: andy at syma.sussex.ac.uk BITNET: andy%syma.sussex.ac.uk at uk.ac
-----------------------------
From: Steve Hayman <sahayman at iuvax.cs.indiana.edu>
Subject: Re: How can I get the Frequently Asked Questions?
Date: 27 Mar 90 14:37:30 GMT
To: info-unix at sem.brl.mil
>I receive the newsgroup, in digest form, from SEM.BRL.MIL, and have yet
>to see the frequently asked questions posting (I suppose it gets editted
>out??). Should I be seeing it?
Various people have made this same complaint from time to time.
I have checked with the maintainers of the info-unix digest gateway,
and nobody is editing out the FAQ posting from the digest. It is merely
a mail reflector that sends out every posting that comes in.
It could be that the size of the FAQ posting makes that particular issue of the
digest too large for your mailer to handle. The current posting is
over 40K in size. (I am trying my best to keep it from growing much larger.)
Steve
--
Steve Hayman Workstation Manager Computer Science Department Indiana U.
sahayman at iuvax.cs.indiana.edu (812) 855-6984
-----------------------------
From: "ryerson.schwark" <ry at cbnewsl.att.com>
Subject: Re: AT&T source license
Date: 27 Mar 90 14:24:53 GMT
To: info-unix at sem.brl.mil
meissner at osf.org (Michael Meissner) writes:
>jstewart at madhaus.utcs (John Stewart) writes:
>
>| We'd like to get source code for Mach from CMU. The bottleneck is that we
>| need to get an AT&T Unix source license so we can get a BSD source license
>| so we can a Mach license. Confused? Join the crowd.
>|
>| I've tried talking to AT&T but with little success so far. The people
>| I've spoken to seem to think any type of source license would cost around
>| $100,000. Maybe that would be reasonable if we were a commercial company
>| doing Unix development but we're not. We don't need AT&T source code at
>| all; all we need is a piece of paper that will make Berkeley happy.
>
>BSD code contains source code from the original UNIX'es. Mach 2.5
>contains BSD code (which contains AT&T code). Thus you need to get
>AT&T source code. If AT&T no longers offers a reduced license
>agreement for Universities, and does not offer earlier versions of
>Unix (I believe you need at least a 32V license, you may need a System
>V.2 license, which went for ~40,000), you are indeed between a rock
>and a hard place. Since they own the source rights, there is not much
>you can do....
AT&T remains committed to educational pricing. A University-wide license
for Release 3.0 costs $1200, and would grant you the rights that
you need. Release 4.0 is also available if you are interested.
The Price of $100,000 is the commercial price for a Release 4.0
license.
I'm sorry that the people you have spoken to have been less than helpful.
Pleae try the 1-800-828-UNIX number and identify yourself as from a
university(I'm assuming from your header that you are), and let them
know that you need a Release 3 license. They can explain the information
that they need.
If you have any further problems, please feel free to call me at
(201) 644-3770, and I'd be glad to help.
Ry Schwark
ryerson.schwark at att.com
att!uso!ry
UNIX Software Operation
-----------------------------
From: Guy Harris <guy at auspex.auspex.com>
Subject: Re: AT&T source license
Date: 27 Mar 90 18:22:11 GMT
Followup-To: comp.os.mach
To: info-unix at sem.brl.mil
>If AT&T no longers offers a reduced license agreement for Universities,
Which, of course, they used to, and at a relatively low price (< $1K,
when last I kept track of that stuff; I don't know what the most recent
price is). I'd suggest that the original poster beat up their AT&T
office harder to see whether academic licenses are still offered.
-----------------------------
From: Mark Harrison <harrison at necssd.nec.com>
Subject: Re: Programmers Prefer 8.5 x 11
Date: 27 Mar 90 15:49:29 GMT
To: info-unix at sem.brl.mil
My 2 cents:
1. I like 8.5 x 11 because it is easy to copy, and because it is easy
to punch holes in magazine articles, etc, and stick them in.
2. I like the boxes many 5.5 x 8.5 binders come in. One problem with
having a lot of looseleaf binders on a shelf is that the spine is
wider than the other end, and they don't stay on the shelf so well.
The boxes fix this.
3. For general reading, I like bound books. I don't like them for
reference stuff because they don't lie flat and they are too hard
to copy. (I heard a Borland guy give a talk, and being hard to
copy was one of the reasons they went with bound docs... the other
was the much lower price.)
the reasons
Note: the above references to copying should not be interpreted as
a cavalier approach to other's copyrights... I copy within
the restrictions of the US copyright code.
--
Mark Harrison harrison at necssd.NEC.COM
(214)518-5050 {necntc, cs.utexas.edu}!necssd!harrison
standard disclaimers apply...
-----------------------------
From: Daniel McKeel <dm76+ at andrew.cmu.edu>
Subject: PC/IX
Date: 27 Mar 90 16:33:06 GMT
To: info-unix at sem.brl.mil
Could someone advise me if the old unix PC/IX for the
IBM PC will run on a NON TRUE BLUE system.I've tried but
no luck.I have all manuals for the beast.Any would be great
Please mail any info to dtm at me.ri.cmu.edu.
Thanks Dan McKeel
-----------------------------
From: Cookson <cookson at helios.mitre.org>
Subject: Archive Question
Date: 27 Mar 90 17:18:03 GMT
Sender: news at linus.uucp
To: info-unix at sem.brl.mil
Is there a way to get access to things like the comp.sources.unix
archives on uunet (or anywhere for that matter) when one doesn't
have ftp??
Thanks in advance,
Dean
% Dean Cookson $ Anyone can be taught to sculpt %
% cookson%helios at mbunix.mitre.org $ Michaelangleo would have had to %
% {devax,et al..}!linus!mbunix!helios!cookson $ have been taught how not to. %
% Disclaimer: My opinions are my own, leave $ The same is true of great %
% employer out of it $ programmers %
-----------------------------
From: "Brando W. Brown" <brando at uiucme2.me.uiuc.edu>
Subject: Re: Archive to tape software
Date: 27 Mar 90 17:31:48 GMT
Sender: News <usenet at ux1.cso.uiuc.edu>
To: info-unix at sem.brl.mil
In article <22860 at adm.BRL.MIL> ZZASSGL%cms.manchester-computing-centre.ac.uk at nsfnet-relay.ac.uk writes:
>
>
>I am looking for a file archive system for a large Sys V Unix (UTS) system.
>
Try REELbackup by Sceptre Corporation. It supporsts tape recycling, mailing
to administration personnel, paging administrators, etc...
+============================================================================+
| Brandon Brown Internet: brando at uiucme.me.uiuc.edu |
| Addamax Corporation UUCP: uunet!uiucuxc!addamax!brown |
| 2009 Fox Drive GEnie: xmg23356 macbrando |
| Champaign, IL 61820 CompuServe: 73040,447 |
+============================================================================+
-----------------------------
From: "Brando W. Brown" <brando at uiucme2.me.uiuc.edu>
Subject: Re: 4.3 System Administration Training Courses
Date: 27 Mar 90 17:36:38 GMT
Sender: News <usenet at ux1.cso.uiuc.edu>
To: info-unix at sem.brl.mil
In article <5587 at bgsuvax.UUCP> denbeste at bgsuvax.UUCP (William C. DenBesten) writes:
>I am looking for information on a course that I could take that will teach
>me how to become a system administrator on a 4.3BSD machine.
>
>I am looking in particular for things like setting up ethernet
>interfaces, compiling kernels, managing mail/printers/news, and
>building boot tapes
>
If you are set on taking classes, skip this message, but an excellent system
administration book was just published by Prentice Hall called "Unix System
Administration Handbook" written by Nemeth Snyder and Seebass. It provides
good, useful information for the system administrator. It is more of a begin-
ners book, however.
+============================================================================+
| Brandon Brown Internet: brando at uiucme.me.uiuc.edu |
| Addamax Corporation UUCP: uunet!uiucuxc!addamax!brown |
| 2009 Fox Drive GEnie: xmg23356 macbrando |
| Champaign, IL 61820 CompuServe: 73040,447 |
+============================================================================+
-----------------------------
From: Mark Runyan <runyan at hpcuhc.hp.com>
Subject: Re: RCS: how do you freeze a configuration?
Date: 27 Mar 90 17:43:51 GMT
To: info-unix at sem.brl.mil
>/ meyer at gorgo.ifi.unizh.ch (Urs Meyer) / 7:34 am Mar 26, 1990 /
>
>One question remains open in
>my opinion. Say, you have a collection of files, each with a (possibly)
>different revision number. In order to create a configuration ready to
>be released, I'd like to assign the same symbolic name to the latest
>revision of each file, s.t. the symbol is fixed to the latest revision
>forever.
A possible solution using existing RCS commands.
rcs -nSymbolName: *,v
Where "SymbolName" is a symbolic name that you wish to use to indicate
a particular revision, the ":" implies set "SymbolName" to top revision
of the file, and "*,v" means all the RCS files.
For instance, if I had 3 different directories, with RCS subdirectories,
and each directory had twelve files, whose top revision number varies
between 1.1 and 20.6, and I want to mark the top revision with one
symbol, I could do:
rcs -nRel_A/B1/00: */RCS/*,v
and Rel_A/B1/00 would point (fixed) to the top revision of all those
files. So Rel_A/B1/00 might point to 1.1 in fileA of the first directory,
but it could also point to 20.6 in fileN of the third directory.
The ":" without a number means point to the top most revision of the file.
Mark Runyan
-----------------------------
From: Dion Hollenbeck <hollen at megatek.uucp>
Subject: Re: RCS: how do you freeze a configuration?
Date: 27 Mar 90 20:53:51 GMT
To: info-unix at sem.brl.mil
>From article <1990Mar26.153424.16557 at gorgo.ifi.unizh.ch>, by meyer at gorgo.ifi.unizh.ch (Urs Meyer):
> Hello,
>
> I followed the recent discussion about rcs (and sccs), specifically the
> symbolic revision numbers are of interest. One question remains open in
> my opinion. Say, you have a collection of files, each with a (possibly)
> different revision number. In order to create a configuration ready to
> be released, I'd like to assign the same symbolic name to the latest
> revision of each file, s.t. the symbol is fixed to the latest revision
> forever.
>
There are three flags which you need to know about. "-m" will enter the
quoted log message of your choice. "-n" will attach a symbolic name
to the highest existing revision of an RCS file, but will refuse with
an error message if this symbolic name has already been given to
another revision. "-N" is the same, but it will force the highest
rev to get the symbolic name even if it was given before. We use
"-N" so that after you have marked a release snapshot and then
find out that you need to change one file and re-mark it with the
same name, the override will apply.
Here is a sample shell script which we use.
--------------------------- cut here -------------------------------
#!/bin/sh
#
# NAME
# rel - Check in for a new release
#
# DESCRIPTION
# This script checks out all the source and include files, then
# checks them in again with the new release name.
#
for i in \
gc.c gc.h gc_misc.c \
gc_rb.c gc_rbovr.c gd.c gd_aaa.c gd_all.h gd_cb2.c gd_cb2.h \
gm.c gm.h os.c os.h
do
echo $i:
co -l $i
ci -f -u -m"-- Release 2.0 --" -NRelease2_0 -sRel $i
done
-------------------- cut here -------------------
Alternately, the "for i in file1 file2 file3...filen" could be
replaced with a wildcard expression to get all files:
for i in RCS/*,v or for i in *,v or for i in RCS/*--
Dion Hollenbeck (619) 455-5590 x2814
Megatek Corporation, 9645 Scranton Road, San Diego, CA 92121
uunet!megatek!hollen or hollen at megatek.uucp
-----------------------------
From: Paul Houtz <gph at hpcc01.hp.com>
Subject: Problems with diff -e and ed?
Date: 27 Mar 90 18:12:29 GMT
To: info-unix at sem.brl.mil
Does anyone know what types of problems are associated with using
diff -e and then storing the result and the original and using it as
backup file, using ed to restore the original.
I know that a file with a line with a single . in it will confuse ed, and
I have taken care of that. Is there anything else?
Thanks!
gph at hpsdesis.HP.COM
-----------------------------
From: "D'Arcy J.M. Cain" <darcy at druid.uucp>
Subject: Re: ANSI header files
Date: 27 Mar 90 18:17:13 GMT
To: info-unix at sem.brl.mil
In article <2105 at kiwi.mpr.ca> parker at zaphod.mpr.ca (Ross Parker) writes:
>Does anyone know if ANSI header files are available for UNIX
>(specifically Ultrix 3.0 & 3.1)? I.e. header files to replace
>the standard /usr/include/... but complete with function prototypes?
>
I have been modifying the standard files for SV3.2 (ESIX) and would
not mind posting the results but I am not sure what the copyright
situation would be. I don't think diffs would be as universally
useful as the full files. I suppose I could simply extract the function
prototypes and post them but that seems too much like work. Any ideas?
--
D'Arcy J.M. Cain (darcy at druid) | Thank goodness we don't get all
D'Arcy Cain Consulting | the government we pay for.
West Hill, Ontario, Canada |
(416) 281-6094 |
-----------------------------
From: daisy basto <daisy at mavrick.uucp>
Subject: Re: How to set TERM for Televideo?
Keywords: Televideo
Date: 27 Mar 90 19:06:14 GMT
To: info-unix at sem.brl.mil
Thanks for everyone who emailed. It's working now (more or less). I'm not
sure whether it's using termcap or terminfo. There is one LARGE /etc/termcap
file with lots of terminals defined and about 30 files in /usr/lib/terminfo/t.
I think it's using terminfo though. When I tried compiling a C program with
-ltermcap it couldn't find that library. And, by the way, I'm using SysV.
Thanks, Daisy
-----------------------------
From: "Malcolm L. Carlock" <malc at unrvax.unr.edu>
Subject: ``Office Automation'' under Unix?
Date: 27 Mar 90 19:18:04 GMT
Sender: news at tahoe.unr.edu
To: info-unix at sem.brl.mil
Has anyone out there had experience with any "office automation" systems
running on Unix machines? I'm thinking along the lines of something like
DEC's All-In-One, perhaps a bit less ostentatious. The two main facilities
that are being sought are email (piece o' cake) and some sort of calendar/
scheduling system, not unlike what I'm told appears in WordPerfect's
office automation system.
I suspect we could write our own "office automation system" without too
much trouble, using existing lower-level facilities, but I'd be interested
in the opinions and experiences of others in this area. Again, we'd be wanting
to run this on a Unix machine. The machine we have in mind is running 4.3 BSD,
though we could also run this on a medium-size Sun that's here. We're
currently awaiting a call back from Sun on this (we're also waiting to hear
back from DEC on an Ultrix version of All-In-One, though we don't currently
have a large Ultrix machine we'd be running it on.) Be nice if we found a
decent PD one though...
Please respond via email and I will summarize to the net.
Thanks in advance,
Malcolm L. Carlock Internet: malc at unrvax.unr.edu
UUCP: uunet!unrvax!malc
-----------------------------
From: Jim Reid <jim at computer-science.strathclyde.ac.uk>
Subject: Re: File server security
Date: 27 Mar 90 19:22:01 GMT
Sender: news at cs.strath.ac.uk
To: info-unix at sem.brl.mil
In article <Mar90.193112.9456 at ixi.co.uk> paul at ixi.co.uk (Paul Davey) writes:
}In article <9926 at batcomputer.tn.cornell.edu>
}rogerj at tcgould.tn.cornell.edu (Roger Jagoda) writes:
}>Is there a way to limit rlogin, telnet, ftp access to just
}>a few users (the net administrators). ...
}
}Set up a netgroup named say "admin" under yp with your
}administrators listed as a set of users, leave the domain
}and machine fields empty (wild).
}In your server passwd files use + at admin instead of + to
}include just the members of the admin group.
}
}The only problem I have found with this is that the yp master
}passwd file is usually on a server...
Of course, if you're using YP, security has long gone out the window.....
Jim
-----------------------------
From: rkc <rkc at xn.ll.mit.edu>
Subject: gas
Date: 27 Mar 90 21:46:27 GMT
FCC: ~/SMAIL
To: info-unix at sem.brl.mil
(This may be a repost. Apologies for our mailer.)
Awhile back we were using GCC with Sun's assembler. At that time I had
written the following copy code:
asm volatile("tstl %4;jle 1$;subqw #1,%4; 0$: movel %1 at +,%0 at +; dbf %4,0$;1$:"
: "=a" (p_to), "=a" (p_from)
: "0" (p_to), "1" (p_from), "d" (n_longs) );
Now we have converted to Gas (Gnu's assembler), which doesn't seem to
understand the 0$ construct. Can someone point me to some documentation on
gas and how to create these local variables, or perhaps point me to a correct
newsgroup to mail this to.
Thanks,
-Rob
-----------------------------
From: Doug Gwyn <gwyn at smoke.brl.mil>
Subject: Re: termcap or terminfo?
Date: 27 Mar 90 22:49:59 GMT
To: info-unix at sem.brl.mil
In article <255 at mavrick.UUCP> daisy at mavrick.UUCP (daisy basto) writes:
>I think it's using terminfo though. When I tried compiling a C program with
>-ltermcap it couldn't find that library. And, by the way, I'm using SysV.
Try -ltermlib, which is the proper name for the termcap support library.
Or, try -lcurses; it it satisfies references to the tgetent() etc. calls
then it's probably terminfo-based, otherwise you would have needed both
-lcurses and -ltermlib for curses-using applications.
-----------------------------
From: John Nall <nall at sun8.scri.fsu.edu>
Subject: Availability of Korn Shell
Date: 28 Mar 90 00:30:58 GMT
Sender: news at vsserv.scri.fsu.edu
To: info-unix at sem.brl.mil
Can some knowledgeable gentleperson tell me how to go about getting
the source of the Korn Shell (assuming such a thing is possible)?
My understanding - which may be incorrect - is that it is available as
a part of the AT&T Toolchest, so the question might should have been
better phrased. However, hopefully it is adequate.
We (Florida State University) are properly licensed for AT&T Unix, for
BSD 4.x (sufficient for Mach, anyway), and the proper CPU's are covered.
Any suggestions or helpful comments (including potential vendors - we do
not particularly expect things to be free, although it would be nice) by
e-mail would be greatly appreciated.
John Nall
nall at nu.cs.fsu.edu
--
John W. Nall | Supercomputation Computations Research Institute
nall at nu.cs.fsu.edu | Florida State University
#include <clever.h> | Tallahassee, FL 32306
-----------------------------
End of INFO-UNIX Digest
***********************
More information about the Comp.unix.questions
mailing list