Sun-Spots Digest, v6n34

William LeFebvre Sun-Spots-Request at RICE.EDU
Mon Mar 21 02:37:34 AEST 1988


SUN-SPOTS DIGEST          Sunday, 20 March 1988        Volume 6 : Issue 34

Today's Topics:
                         Re: Sun 4 info requested
                 Re: pr_load() bug, summary of solutions
                        Re: New improved calentool
                        Wrong number from sprintf
                     Dvipage TeX and LaTeX Previewer
                            sun-4 clock patch
                                YP Problem
                        Monthtool hi-res problems
               Problems reading others' TAR tapes on a SUN
                  Info on P-NUT and Nest packages wanted
                I need technical details on Sun SCSI disks
                 Filter for Versatec Color Plotter 2552C?
             Using AMD AMZ8068 DES chip in SUN workstations?
             Connecting to Internet via X.25 ip-line-driver?
                        MX mailer needed for Suns
                           SunOS 3.5 Problems?

Send contributions to:  sun-spots at rice.edu
Send subscription add/delete requests to:  sun-spots-request at rice.edu
Bitnet readers can subscribe directly with the CMS command:
    TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name
Recent backissues are stored on "titan.rice.edu".  For volume X, issue Y,
"get sun-spots/vXnY".  They are also accessible through the archive
server:  mail the word "help" to "archive-server at rice.edu".

----------------------------------------------------------------------

Date:    Thu, 10 Mar 88 13:56:15 PST
From:    lai at cs.ucla.edu (Nick Lai)
Subject: Re: Sun 4 info requested
Reference: v6n25

We currently have three Sun 3/280 machines which are being used as
timesharing machines.  We regularly have 30 to 40 users on each machine,
and the responsiveness is generally very good.  (Note that we offload
troff jobs to 4 MIP IBM RT PCs, which helps a lot.)

The main bottleneck on the 3/280 appears to be context switching.  Our
3/280s start to degrade markedly when there are more than 8 runnable
processes, apparently due to the fact that the MMU only has 8 context
register sets.  I think I heard that the 4/280 has 16.

I would imagine that a 4/280, with fast disks, could support 40 to 50 users.

Nick

------------------------------

Date:    Fri, 11 Mar 88 16:14:29 EST
From:    warsaw at cme-durer.arpa (Barry A. Warsaw)
Subject: Re: pr_load() bug, summary of solutions
Reference: v6n22

Thanks to everyone who submitted work arounds for the pr_load() bug.  The
basic problem is that pr_load() does not allocate memory for the red,
green and blue arrays in the colormap_t structure.  Also, you need to set
the type and length of the colormap.  The following chunk of code is the
simplest workaround for 256 slot maps:

load_colormap_from_file(fp)
FILE	*fp;
	{
	Pixrect 	*pr;
	colormap_t	map;

	map.type = RMT_EQUAL_RGB;
	map.length = 256;

	map.map[0] = (unsigned char *) malloc(256);
	map.map[1] = (unsigned char *) malloc(256);
	map.map[2] = (unsigned char *) malloc(256);

	if ((pr = (Pixrect *) pr_load(fp, &map)) == NULL) {
		/*
		 * something else went wrong with the pr_load()
		 */
		}
	}


I've also been told that this is changed in release 3.4 and up so that
pr_load() allocates these arrays for you.  This was corroborated by
someone who is running 3.5 which executed my original code (without the
malloc's) unchanged.

I can now load my rasterfile complete with colormap, except for one small
discrepancy.  It seems that the image that I see in my application is not
*exactly* the same as what I see using screenload.  Two of the colormap
slots appear to have been drastically changed.  Oh well, its probably my
program...

NAME:  Barry A. Warsaw         ARPA: warsaw at cme-durer.arpa
TELE:  (301) 975-3460          UUCP: {...}!uunet!cme-durer!warsaw

------------------------------

Date:    Sat, 12 Mar 88 10:12:14 MST
From:    roberts%studguppy at lanl.gov (Doug Roberts)
Subject: Re: New improved calentool

I installed the new calentool on a Sun 3/50, & it looks great.  However,
when you try to run it on a high-res monitor (3/260), the image is all
munged up.

Here's a workaround:  Calentool requires the default font in order to run
on a Sun high res monitor. In addition, calentool will not accept the
suntools -Wt font parameter, so you MUST allow your login to occur using
the default SunView font (screen.r.11, I believe).

If you do this, calentool works fine.

--Doug

Douglas Roberts
Los Alamos National Laboratory
(505)667-4569
dzzr at lanl.gov

------------------------------

Date:    Wed, 9 Mar 88 15:03:17 MST
From:    "Gregg Townsend" <gmt at arizona.edu>
Subject: Wrong number from sprintf

Here's an insidious bug that you might want to keep in mind in case you
ever have a problem with intermittently wrong output:

If a timer interrupt occurs during a sprintf() call to format an integer
(%d), the last digit of the formatted string can come out wrong.  It's
usually turned into a '0' but sometimes into a '4'.

I've seen this on a Sun2 running 3.2 and a Sun3 running 3.4.  The Sun
hotline has confirmed that 3.5 also exhibits the bug.

Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721
+1 602 621 4325      gmt at Arizona.EDU       110 57 16 W / 32 13 45 N / +758m

------------------------------

Date:    9 Mar 88 19:35:40 GMT
From:    hunt at spar.slb.com (Neil Hunt)
Subject: Dvipage TeX and LaTeX Previewer

I have a program called dvipage, based upon the dvisun program distributed
with the TeX tape.

It previews dvi files under SunView; it has several advantages over the
old dvisun program:

	Renders the page at high resolution (for example, using the 300dpi
	  fonts) and then performs filtering and sampling down to produce
	  a page at 75 dpi, with grey scale fonts, for display on
	  a colour or grey scale window. This allows a whole page of
	  8.5x11 or A4 to be viewed on a Sun screen at one go,
	  and gives much nicer pictures than the 118 dpi fonts which
	  are traditionally used.

	Nicer user interface, including menu and mouse control, etc.

If there is sufficient interest, I will attempt to get the powers-that-be
to allow me to make a distribution. Note that it is likely to take a few
weeks before I can get through the paperwork. If you would like such a
program, please send me mail.

Neil/.

------------------------------

Date:    Fri, 11 Mar 88 09:50:59 mst
From:    ed arnold <era at scdsw1.ucar.edu>
Subject: sun-4 clock patch

Sun-4s running Sys4-3.2 have the same clock drift problem as Sun-3s.
Since I haven't seen a msg in sun-spots about the patch procedure for 4s,
you may wish to apply the following patch to your clock.o and vmunix.
(This patch was obtained from Sun's Denver office; call your local rep if
you have questions.)

[[ I hate to disappoint you ed, but this patch did appear in Sun-Spots
Volume 6 Issue 1 along with the patches for 3.[2345] and 3.0.  I double
checked, and the patch you sent me was exactly what was sent in that
issue.  But if anyone out there still has questions about the clock drift
problem, you should certainly call your local rep. (I think Sun is still
just a little embarrassed about that bug).  --wnl ]]

------------------------------

Date:    Wed, 9 Mar 88 15:37:09 EST
From:    clapper at nadc.arpa (Brian M. Clapper)
Subject: YP Problem

I've run into a small problem with the Yellow Pages...

We just started using NFS and YP on two Sun-3's.  Originally, the
yppasswdd(8) daemon was set up on the master server to update the yp
database from /etc/passwd.  After I was confident that I understood a
little about yp (no thanks to Sun's documentation), I decided to move the
yp password file to /etc/yp/`domainname` so that only local users would be
in /etc/passwd.  I made the appropriate changes to /etc/passwd and the new
yp passwd file, and respawned the yppasswdd daemon to update the new file,
using the command:

	/usr/etc/rpc.yppasswdd /etc/yp/`domainname`/passwd -m passwd \
	DIR=/etc/yp/`domainname`

All of the sudden, yppasswd(1) responded with "couldn't change passwd".
When I changed the new yp passwd file manually (with vi) and ran "make
passwd DIR=..." in /etc/yp, everything worked fine, however.  Perplexed, I
then restored the /etc/passwd file to its original state, killed the
yppasswdd daemon, and respawned it with the command:

	/usr/etc/rpc.yppasswdd /etc/passwd -m passwd

Yppasswd(1) now had no trouble changing the network password, from either
machine.

Does anyone have a clue?  What am I missing here?  Please respond to me in
person, as I do not regularly receive Sun-Spots.

Thanks.

Brian M. Clapper                         ARPA:  clapper at nadc.ARPA
Code 7031                                UUCP:  ...!harvard!clapper at nadc.ARPA
Naval Air Development Center
Street and Jacksonville Roads            Phone:   (215) 441-2118
Warminster, PA 18974                     AUTOVON: 441-2118

------------------------------

Date:    Thu, 10 Mar 88 20:55:14 MST
From:    roberts%studguppy at lanl.gov (Doug Roberts)
Subject: Monthtool hi-res problems

I've been using your monthtool & "swear by it". However, I recently
swapped the old Sun 3/50 for a 260 w/high res monitor. I was curious if
your new version supported the high res.

Thanks, Doug

Douglas Roberts
Los Alamos National Laboratory
(505)667-4569
dzzr at lanl.gov

------------------------------

Date:    11 Mar 88 14:52:40 GMT
From:    cmcl2!ll-xn!atexrd!mikeh at rutgers.edu (Mike Harris)
Subject: Problems reading others' TAR tapes on a SUN

Forgive me if this question has already been posed to the group, but can
anyone tell me how to extract from a 1/4 inch cartridge, written on a
MASSCOMP, from a SUN??  I had no trouble whatsoever reading and writing
tapes between Masscomp and Intel machines, but am unable to read the tape
on a Sun.  Any ideas??

Mike Harris, Atex Inc., A Kodak Company                    1-617-276-7392
{ll-xn,genrad,munsell,kodak}!atexrd!mikeh

[[ That's peculiar.  You certainly shouldn't have any problems.  I would
suspect a tape hardware problem.  Have the head alignment on all drives
involved checked and adjusted.  I've seen a similar problem on mag tape
drives, and I imagine it exists for cartridge drives:  if the head is
misaligned it will read tapes it had written with no problem, but it will
not be able to read tapes written by other drives and vice versa.  --wnl ]]

------------------------------

Date:    FRI MAR 11, 1988 10.18.14 EST
From:    "Nick Iliev" <NI00 at LEHIGH.BITNET>
Subject: Info on P-NUT and Nest packages wanted

We are in the market for the P-NUT and Nest simulation packages for the
Sun-3/160C (8 Mb) with SunOS 3.4.  To the best of our knowledge these
tools can be used (and will be used by us) for protocol verificaton and
Petri Net analysis (P-NUT) and for simulating computer communications on a
LAN at the message level.

I would appreciate receiving any recommendations you have regarding these
simulators as well as advice on their use for protocol and network
simulation based on actual user experience. Currently we plan to use P-NUT
for the verification of an IEEE 802.4 based token bus protocol for an
automated factory network to be simulated with Nest 2.5.  We plan to run
the Nest user interface on a diskless Sun-3/50 and the actual simulation
on an AT&T 3b2.

Please reply by e-mail.  I will post a summary.

Thank yo very much.

Nick Iliev <ni00 at lehigh.edu>
Lehigh University, CSEE Dept.
Packard Lab #19
Bethlehem, Pennsylvania 18015
(215) 694-8136
INTERNET :   denton at scarecrow.lehigh.csee.edu

------------------------------

Date:    10 Mar 88 00:42:17 GMT
From:    harry at rainy.atmos.washington.edu (Harry Edmon)
Subject: I need technical details on Sun SCSI disks

I am very confused about Sun 3/50s and their SCSI disks.  Can someone
enlighten me on the following:

1) What controllers and interface boards are in the Sun shoebox?
2) What do you need to add an additional disk to a Sun shoebox?
   Can you chain off the controller or do you need another controller?
3) What do you need to add a disk to a diskless Sun?

Harry Edmon                   UUCP:   harry at rainy.atmos.washington.edu or
(206) 543-0547                        uw-beaver!geops!rainy!harry
Department of Atmospheric Sciences
University of Washington        BITNET: HARRY at UWARITA

------------------------------

Date:    Wed, 09 Mar 88 13:19:04 SST
From:    Tat-Leong Chew <NCBITI3 at NUSVM.BITNET>
Subject: Filter for Versatec Color Plotter 2552C?

I have a very urgent problem (aren't they all?). I have a Versatec 2552C
Color Plotter that I need to dump Sun raster files to. Problem is the
documentation is not here and I have no idea when (if?) it will ever get
here. Does anyone out there have or know of a filter that will do the job.
I shall be eternally grateful (or at least until the next urgent problem
crawls along  :-).

------------------------------

Date:    9 Mar 88 15:41:20 GMT
From:    David M. Balenson <balenson at mimsy.umd.edu>
Subject: Using AMD AMZ8068 DES chip in SUN workstations?

Can anyone provide any information concerning the use of the AMD DES chip
within the SUN workstations (e.g., 3/50, 3/140, 3/160, 3/180)?

I understand that the CPU boards have any empty socket which will
accomodate the AMD chip (AmZ8068) and that the kernel has to be
reconfigured to make use of the chip.  Is that all there is to it?  Will
this permit the 'des' command to use the AMD chip, rather than use
software?  I have also heard that two additional chips may or may not be
necessary -- one is a TTL support chip, the other a PAL.  What exactly are
these chips for?  Are they necessary, or not?

Is anyone out there, in particular anyone at SUN, currently using the AMD
chip?

Thanks.

David M. Balenson (balenson at icst-ssi.arpa)
National Bureau of Standards
(301) 975-2910
balenson at mimsy.umd.edu  -or- balenson at icst-ssi.arpa

------------------------------

Date:    8 Mar 88 13:08:29 GMT
From:    unido!fuhainf!edv at uunet.uu.net (E de Vries)
Subject: Connecting to Internet via X.25 ip-line-driver?

we would like to know, if somebody has tried to connect to the internet
temporarily ( just for ftp-ing PD software ) via X.25 ?

Can some kind internet-node tell help us, if it is possible at all!

thanks in advance
enno de vries

...unido!fuhainf!edv  (uucp)
edv at vax1.informatik.fernuni-hagen.dbp.de (X.400)

------------------------------

Date:    Wed, 9 Mar 88 11:39:14 MST
From:    kurt at hi.unm.edu (Kurt Zeilenga)
Subject: MX mailer needed for Suns

Does any have a MX sendmail for Sun workstations?  We have source, so
hacks would be fine.  If not, I guess I get to trash Sun's sendmail.

- Kurt (zeilenga at hc.dspo.gov)

------------------------------

Date:    Fri, 11 Mar 88 11:36 EST
From:    Roy Richter <RICHTER at gmr.com>
Subject: SunOS 3.5 Problems?

In Sun-Spots vol 5, issue 71, I queried whether I should upgrade to SunOS
3.4 or not.  Thanks to all of those who responded.  The general responce
was that there are some things desired in 3.4 (better compiliers, updated
cmdtool) but that the networking bugs made 3.4 as a whole undesireable.
The solution was that I wait for 3.5, to be released generally in
mid-January.

I requested 3.5 from sun-support (it only comes at your request), they
mailed it, and I'm using it.  The compilers are ~ 5% faster, no hard
numbers to stake that claim, but it's what I've found.  The cmdtool is
better in that you can vi in it (finally).

I've encountered a problem, though.  My /tmp directory is filling up with
suntools droppings.  If you reboot once a week, whether you need it or
not, this is not a real problem (you know who you are).  Otherwise I see a
problem analogous to the extra entries given by `w' in utmp produced by
tools killed before they've cleaned up.  In particular, mailtool, cmdtool,
probably textedit, all get killed before they've removed their scratch
files in /tmp.  Some of these can be ~100k or so, depending on what you're
doing.  I've contacted sun-support on this, no response yet other than
`yeah, it happens..'

In general 3.5 works well otherwise.  I'm happy.  The UPGRADE script had
no problems for my configuration (I'm a standalone) but I've heard of
problems for  losing /etc/group on clients (rumor) .

Any other comments on SunOS 3.5?

Roy Richter
Physics Department
GM Research Laboratories
CSNET: richter at gmr.com

[[ The /tmp problem can be alleviated in two different ways:  first, have
a cleanup shell/find script that gets run once a day and removes old files
from /tmp; second, put /tmp (either by mounting or with a soft link) on a
large partition.  --wnl ]]

------------------------------

End of SUN-Spots Digest
***********************



More information about the Comp.sys.sun mailing list