Sun-Spots-Request@Rice.edu (William LeFebvre) (10/03/88)
SUN-SPOTS DIGEST Sunday, 2 October 1988 Volume 6 : Issue 246
Today's Topics:
Re: How do I tell what version of SunOS is running
Re: Sun 4.0 string problem
Re: RPC failure in OS4.0 client boot
Sun Domain Name Server & YP
Rudeness of SunOS 4.0 getwd
bad user stack error message
Error in -f argument processing for netstat
rsh & rcp with Domain Name Server
Problem: SLOW Boot on diskless Suns
Sun 3/60c questions
Questions about extension cables
Graphics conversions?
Wanted: DEC LAT daemon for the SUN
Drawing system for Sun?
Mathematica for Sun?
loading clusters on 386i?
CASE and Tom Gilb?
Send contributions to: sun-spots@rice.edu
Send subscription add/delete requests to: sun-spots-request@rice.edu
Bitnet readers can subscribe directly with the CMS command:
TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name
Recent backissues are available via anonymous FTP from "titan.rice.edu".
For volume X, issue Y, "get sun-spots/vXnY". They are also accessible
through the archive server: mail the request "send sun-spots vXnY" to
"archive-server@rice.edu" or mail the word "help" to the same address
for more information.
----------------------------------------------------------------------
Date: Tue, 27 Sep 88 18:29:17 PDT
From: Paul O'Neill <pvo1478@neptune.oce.orst.edu>
Subject: Re: How do I tell what version of SunOS is running
Reference: v6n235
This isn't as elegant as you'd like--but here's what I use in my .login:
set j = `cat /etc/motd`
if ($j[1] == "SunOS") then
....................
.... 4.0 stuff ..... ( $j[3] = 4.0 )
....................
else
....................
..... 3.x stuff .... ( in 3.4, $j[5] = 3.4 )
....................
endif
Paul O'Neill pvo@oce.orst.edu
Coastal Imaging Lab
OSU--Oceanography
Corvallis, OR 97331 503-754-3251
[[ I think the original poster was looking for a way to tell the
difference at compile time, like you do with an "#ifdef sun". I know
that's what I want. --wnl ]]
------------------------------
Date: 27 Sep 88 14:13:00 EDT
From: Walter Roberson <WCSWR@CARLETON.BITNET>
Subject: Re: Sun 4.0 string problem
Reference: v6n237
I don't have any vax or sun fortran manuals handy, but I did run across
the following in an IBM fortran manual that might perhaps explain the
problem:
Under the heading of "character assignment statement":
"None of the character positions being defined in 'a' must be referenced
in 'b' directly or indirectly or through associations of variables
(that is, using COMMON, EQUIVALENCE, or argument passing)."
As you define string in terms of itself, perhaps this has suddenly become
a problem. A difference in the underlying instruction set could easily
cause the problem to show up now where it did not before.
Hope this helps,
Walter Roberson <WCSWR@CARLETON.BITNET>
------------------------------
Date: 28 Sep 88 18:59:36 GMT
From: eos!jbm@eos.arc.nasa.gov (Jeffrey Mulligan)
Subject: Re: RPC failure in OS4.0 client boot
Reference: v6n240
A few days ago I posted this plea for help:
> Under OS4.0 the clients can't reboot when we are connected to the outside
> world, although if we disconnect from the outside world and connect our
> lab machines together with a multiport transceiver they will boot.
...
> We have reported this problem to SUN but are still waiting for someone to
> get back to us ...
Yesterday, we received the following message from our man at SUN (good one
SUN!):
__________
This is from our internal newsgroup. Some other customers had the same
problem that we were seeing at your place - it looks like there is some
other machine on the network (as described below) that causes 4.0 diskless
clients failure to boot.
After a little bit of sleuthing with etherfind followed by conversations
with some of the PAL people in Mtn View, I found a solution to the above
problem. There is a bug in a DEC product that causes these symptoms. So,
if you have a VMS VAX on your Ethernet AND you are running 4.0 AND you
experience the booting problems described below, go find the VAX system
manager and ask if he is running something called the "ULTRIX BRIDGE".
ULTRIX BRIDGE is DEC's own version of the Wollongong package that allows
VMS machines to use REAL communications protocols. At any rate, DEC has a
patch for this software. You might warn the DEC person that he might have
to be a little persistent -- our first inquiries brought comments on the
order of "never heard of such a thing!".
__________
So, it looks like we have to identify the VMS system upstream from us and
ask them to install the patch. I anticipate that the burden will be on us
to get the patch from DEC, so if anyone has a part number (wishful
thinking perhaps ... )
Jeff Mulligan (jbm@aurora.arc.nasa.gov)
NASA/Ames Research Ctr., Mail Stop 239-3, Moffet Field CA, 94035
(415) 694-6290
------------------------------
Date: Wed, 28 Sep 88 06:38:49 EDT
From: Alexander Dupuy <dupuy@columbia.edu>
Subject: Sun Domain Name Server & YP
[ Doug Peterson (doug@icase.edu) complains that users are unable to rcp or
rsh in one direction between YP/named driven hosts and non-YP hosts. ]
It's not clear from your description, but are you _sure_ that the .rhosts
files are set up correctly? The hostnames in the .rhosts must be the
"official" hostnames. In a table-driven (or YP) environment, this is the
first name in the /etc/hosts file (or the equivalent YP database). In a
named (or YP/named) environment, this is the fully qualified domain name.
There is no assurance that these are the same, unless you always use the
fully qualified domain name as the first name in the /etc/hosts table. As
a quick test, try putting an entry in the .rhosts file for _every_ valid
alias for the machine in question. If things suddenly start working, then
that was your problem...
@alex
------------------------------
Date: Wed, 28 Sep 88 06:51:30 EDT
From: Alexander Dupuy <dupuy@columbia.edu>
Subject: Rudeness of SunOS 4.0 getwd
Under SunOS 4.0, getwd(3) keeps a cache of mounted filesystems in
/tmp/.getwd. This file is not protected, and every process which calls
getwd(3) will check to see whether it is out of date w.r.t. /etc/mtab, and
attempt to update it. You can see this behavior with "trace /bin/pwd".
If you are sensible, and security-minded, one of the first things you will
try to do is "chmod 1777 /tmp /usr/tmp" to make the tmp directories
sticky, so that users can't delete each other's temp files. However, this
prevents getwd(3) from renaming its temporary copy of the cache to be
/tmp/.getwd (unless the user is root, or the person who originally created
/tmp/.getwd).
There are two solutions, the easy but less secure, "chmod 777 /tmp", and
the more complicated, where every time you run mount or unmount (or the
automounter does it for you) you run /bin/pwd as root. That way, the
/tmp/.getwd file is almost always up to date, and user processes don't try
to update it (much). We chose the easy one - since /usr/tmp is sticky,
people can put their files there (as they should, anyhow).
@alex
------------------------------
Date: 27 Sep 88 19:06:58 GMT
From: phri!roy@philabs.philips.com (Roy Smith)
Subject: bad user stack error message
I was scanning one of our /usr/adm/messages files and found the following:
sendsig: bad user stack pid=19957, sig=11
usp is ef7fe30, action is 89a18, upc is 89bf4
Anybody know what that means, and what might have caused it? If a user
process trashed its stack pointer, presumably it would die with some sort
of signal (like the SEGV apparently delivered above) but why should that
show up in a system error log file?
--
Roy Smith, System Administrator
Public Health Research Institute
{allegra,philabs,cmcl2,rutgers}!phri!roy -or- phri!roy@uunet.uu.net
------------------------------
Date: Wed, 28 Sep 88 11:21:15 EDT
From: ehrlich@shire.cs.psu.edu (Dan Ehrlich)
Subject: Error in -f argument processing for netstat
Machine Type: Sun 4/260S
O/S Version: SunOS 4.0
Organization: Computer Science Department, The Pennsylvalia State University
333 Whitmore Laboratory, University Park, PA 16802
Phone Number: +1 814 865 9723
Description:
On a sparc and 68020 machines 'netstat -f' will die with a
segmentation fault if the address family is not specified.
Repeat-By:
% netstat -f
Fix:
Apply the following patch to ucb/netstat/main.c and recompile.
*** /tmp/geta2880 Wed Sep 28 11:03:42 1988
--- /tmp/getb2880 Wed Sep 28 11:03:42 1988
***************
*** 192,197 ****
--- 192,201 ----
case 'f':
argv++;
argc--;
+ if (argc < 1) {
+ fprintf(stderr, "address family not specified\n");
+ exit(10);
+ }
if (strcmp(*argv, "inet") == 0)
af = AF_INET;
#ifdef ENABLE_XNS
------------------------------
Date: Wed, 28 Sep 88 16:40:51 EDT
From: doug@work1.icase.edu (Doug Peterson)
Subject: rsh & rcp with Domain Name Server
The problem I posted earlier about rsh & rcp not working with Sun's
nameserver patch kit under OS3.5 has a fix. Namely, we were 'doing it
wrong'.
If a user attempts an rsh to a machine within my YP domain from a machine
which is not (assume a non-Sun), the nameserver returns a host name in
UPPER CASE. Consequently, the lower case entry in the user's .rhosts file
ON THE SUN MACHINE didn't match. Changing the host names in each entry to
upper case fixed everything.
Someone had responded to me about checking the .rhosts file for
correctness, and making an entry for each alias, just to be sure the
Official Internet Name was covered. Well, all the entries were the
official name, but weren't in upper case.
However, the requirement for upper case names only existed for rsh to the
YP master. When the rsh was to a client, no problem. The entries in my
namd.hosts file (for the name server) are all in lower case, as they are
for /etc/hosts (for the YP database).
Some programs are smart enough to ignore case (e.g. sendmail), but not so
with rsh, et. al.
Doug Peterson
Systems Manager
ICASE
Mail Stop 132C
NASA Langley Research Center
Hampton, VA 23665-5225
(804) 865-4090
FTS: 928-4090
------------------------------
Date: Wed, 28 Sep 88 11:24:31 pdt
From: rusty@cartan.berkeley.edu
Subject: Problem: SLOW Boot on diskless Suns
That happens all the time here. When I notice the client is booting
slowly I just L1-A and type b again. Usually it boots normally the 2nd
time.
------------------------------
Date: Wed, 28 Sep 88 09:44:33 EDT
From: palmer@ncifcrf.gov (Thomas Palmer)
Subject: Sun 3/60c questions
I just received a 3/60c with the Sony 16 inch monitor. Two questions:
- If I run suntools with the "-i" option (invert foreground and
background) all menus come up completely black. I've seen this
before on a 3/110c but never on a 3/260c. Does anyone have a
workaround (without going back to black on white)?
- I seem to remember hearing nasty things about some Sun monitors a
short while back. Were they Sonys?
-Tom
Thomas C. Palmer National Cancer Institute Supercomputer Facility
c/o PRI, Inc. Phone: (301) 698-5797
PO Box B, Bldg. 430 Uucp: ...!uunet!ncifcrf.gov!palmer
Frederick, MD 21701 Arpanet: palmer@ncifcrf.gov
------------------------------
Date: Wed, 28 Sep 88 13:41:37 EDT
From: chuck@wooglin.scc.com (Charles Williams)
Subject: Questions about extension cables
I know I have seen this information before, but I didn't need it then, and now
I do.
Where/who/how can I get my hands on extension cables for the video and
keyboard for my workstations?? I though I saw Belden, but they are giving
me the royal run around and don't appear to know what I'm talking about.
Thanks,
Chuck Williams
(Send replies direct to cwilliams@bluto.scc.com or chuck@wooglin.scc.com)
------------------------------
Date: 28 Sep 88 15:35:01 GMT
From: steinmetz!germany!shapiro@uunet.uu.net (neil r. shapiro)
Subject: Graphics conversions?
Reply-To: shapiro@germany.steinmetz.ge.com
Does anyone have a program(s) to translate from:
o .PCX (Paintbrush) to Sun Raster format
o TIFF (Tagged Image File Format) to Sun Raster format
o Autocad to Framemaker-compatible vector graphics
Neil R. Shapiro ARPANET: shapiro@germany.steinmetz.ge.com
GE Corp. R&D USENET: steinmetz!germany!shapiro
Schenectady, NY
(518) 387-5192
------------------------------
Date: Wed, 28 Sep 88 16:52:28 EDT
From: khai%amara.uucp@umix.cc.umich.edu (S. Khai Mong)
Subject: Wanted: DEC LAT daemon for the SUN
Does anybody know about the availability of software which would allow SUN
access to DEC's LAT via ethernet? I am primarily interested in having the
capability to login to the SUN via DEC's LAT. Thank you.
------------------------------
Date: 27 Sep 88 12:47 +0100
From: Kjell Eidem <eidem%regtek.unit.uninett%NORUNIX.BITNET@cunyvm.cuny.edu>
Subject: Drawing system for Sun?
We are looking for a graphic system for Sun workstations that may be used
to produce simple drawings on the screen, colors are not necessary,
really. It should have the ability to put the screen image on to a
printer, preferably a laser printer with Postscript. An application would
be to make slides of the drawings. Similar systems are McDraft/McDraw for
Macintosh. We have been looking through the Catalyst (January 1988
edition) without finding a product like this. Any comments would be
appreciated.
[[ There is a public domain M*cDraw-like program called "Fig" available in
the Sun-Spots archives under "sun-source". It is contained in 9 shar
files: "fig.shar.01" through "fig.shar.09". It can be retrieved via
anonymous FTP from the host "titan.rice.edu" or via the archive server.
For more information about the archive server, send a mail message
containing the word "help" to the address "archive-server@rice.edu".
--wnl ]]
Kjell Eidem <eidem%regtek.unit.uninett@norunix.bitnet>
The Foundation for Scientific and Industrial Research at the
Norwegian Institute of Technology
------------------------------
Date: 28 Sep 88 7:33 +0100
From: Kjell Eidem <eidem%regtek.unit.uninett%NORUNIX.BITNET@cunyvm.cuny.edu>
Subject: Mathematica for Sun?
We are searching for information about the program packet Mathematica,
developed by Stephen Wolfram, preferably running on Sun 3/50 and 3/60
workstations. If anyone have an idea where to get some information please
let us know.
Kjell Eidem <eidem%regtek.unit.uninett@norunix.bitnet>
The Foundation for Scientific and Industrial Research at
The Norwegian Institute of Technology
------------------------------
Date: Mon, 29 Aug 88 14:32:43 EDT
From: ted@braggvax.arpa
Subject: loading clusters on 386i?
Is it just me, or does everyone find that loading optional software
clusters on a 386i is extremely braindamaged? We have the small harddrive
configuration, and the clusters on 3.5" disks. Even though we have plenty
of room (well, enough room anyway) to load what we want, the load script
insists of building and uncompressing HUGE "bar" files in /tmp. We do not
have enough room to hold both the bar archive and all the files extracted
from it. The only way I've been able to load is mount /tmp on a remote
machine (lucky we've got a network!). Also, loadc does not seem to delete
these humongous files when it is done with them. Why they didn't use a
pipeline I'll never know. I haven't had time to read all the docs I'm
supposed to, am I missing something?
Ted Nolan
ted@braggvax.arpa
------------------------------
Date: 27 Sep 88 13:00:07 GMT
From: mcvax!pe!fk@uunet.uu.net (Flemming Korslund)
Subject: CASE and Tom Gilb?
Does anybody know of a CASE product which support the Tom Gilb method of
software engineering? (Described in: Tom Gilb, "Software Engineering
Management", Addison-Wesley 1988)
Or does anybody have expirence in building it youself with 4GL database or
scripts?
We use Sun-OS (rel 3.4) today but will use AT&T Unix V R4 in the future.
Thanks for any answers.
Purup Electronics a/s Email: fk@pe.dk
Att: Flemming Korslund, R&D Uucp.: ...!mcvax!dkuug!pe!fk
Sonderskovvej 5 Phone: +45 622 2522
8520 Lystrup Fax..: +45 622 2511
DENMARK Telex: 68242 purel dk
------------------------------
End of SUN-Spots Digest
***********************