Sun-Spots Digest, v6n133
William LeFebvre
Sun-Spots-Request at RICE.EDU
Fri Jul 8 02:39:40 AEST 1988
SUN-SPOTS DIGEST Thursday, 7 July 1988 Volume 6 : Issue 133
Today's Topics:
Educational source licenses prohibit commercial development
Poor 4.0 performance
Ciprico Disk Controllers and SunOS 4.0
4.0 Sun logo binder spine labels
tape switch sources
touchup - rasfilter8to1
PC-NFS problems
iebark reset crashes
large-screen/projection monitors?
VITec anyone?
experience with array processors?
MAX Groups Question
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 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 at rice.edu" or mail the word "help" to the same address
for more information.
----------------------------------------------------------------------
Date: Mon, 27 Jun 88 20:56:11 PDT
From: hoptoad.UUCP!gnu at cgl.ucsf.edu (John Gilmore)
Subject: Educational source licenses prohibit commercial development
Tait Cyrus of UNM asked whether he can use a University computer to write
a proprietary software package and then later charge money for it.
If the University computer has an Educational Unix license, the answer is
no. The license requires that any software developed on the machine be
made available to anyone else for a copying cost (subject to license
limitations, e.g. if you used AT&T code in your product, recipients must
also have an AT&T license). This is not a bad deal, considering that
universities get full Unix source for $1K total. The idea is that you are
doing education or research. If you are doing development of commercial
applications, you'd better pay the $80K or whatever it now costs for a
commercial Unix source license.
I would be interested in getting "tips" on any commercially sold software
that was developed on University machines. I'm always looking for good
publicly distributable software to put on SUG tapes :-).
Several universities, including U of British Columbia and UC Berkeley,
have bought commercial Unix licenses for one of their machines, and they
claim to be doing all their commercial development on those machines (X.25
at UBC, and VorTeX and other things at UCB). In the case of Berkeley I
doubt it's true -- the machine is "violet.berkeley.edu" and it's networked
to all the educationally licensed Suns on peoples' desks, where they are
really doing the work. But just having the machine there makes it harder
to prove that they're breaking the license.
You can verify this information either by reading your Unix license (your
company or University promised that you would abide by it if you have
access to the Unix machine, so it would help if you'd read it!) or by
calling AT&T Unix Licensing at +1 800 828 UNIX. (Sorry, I don't know the
number if area code 800 doesn't work from your country.)
John Gilmore
[[ But people doing development work under a standard (binary) Sun license
are okay, because Sun has a binary redistribution license from AT&T. Your
summary only applies to an educational Unix source license from AT&T. At
least, this is my limited understanding of the situation. By the way,
VorTeX's cost is low enough to be considered just a copying fee (under
$300). However, it is not freely redistributable. --wnl ]]
------------------------------
Date: 28 Jun 88 02:08:26 GMT
From: hedrick at athos.rutgers.edu (Charles Hedrick)
Subject: Poor 4.0 performance
We're just now bringing up 4.0 on some test machines. The first person
whose workstation moved to 4.0 threatened to kill me. Suntools is taking
5 to 7 minutes to start for him. (This is on a 2/50 with 4 MB.) So we
brought up a second machine located right next to a machine still running
3.2 (both 4 MB 2/50's - we have bigger machines, but we decided to start
with the machines that were most likely to cause trouble). We found the
following:
starting up suntools, and any tool under suntools - 4.0 is
incredibly slow
starting up Common Lisp or Emacs under suntools - 4.0 is very slow
starting up Common Lisp outside suntools - 4.0 is slightly faster
starting up X with a couple of windows - 4.0 is slightly faster
Common Lisp and X were both built under 3.x, and thus are static linked.
So now the question is: is there something about 4.0 that makes suntools
lose, or is it the new version of suntools itself? So the next experiment
was to NFS mount our 3.2 file server, and chroot to it. Thus I was
running entirely 3.2 software on a 4.0 kernel. This configuration now won
every test again 3.2, including starting suntools, starting Common Lisp
under suntools, and starting emacs under suntools. Thus it appears that
the real lose is not the 4.0 kernel, but either the 4.0 version of
suntools or something about the way it is loaded (e.g. dynamic loading).
More about this later. I wanted to post results before we've had a chance
to look at them very much so as to get other people looking in the same
direction. It appears to be wrong to say that 4.0 in general is slow or a
memory hog, as some have been saying. If necessary, I am perfectly
willing to move to 4.0 but use the old suntools, disable dynamic linking,
or whatever else turns out to be necessary to get reasonable performance
on 4MB systems.
------------------------------
Date: Tue, 28 Jun 88 08:57:06 -0400
From: Dan Grim <grim at udel.edu>
Subject: Ciprico Disk Controllers and SunOS 4.0
I have learned from Aviv Corp., who we have bought Ciprico controllers
from, and confirmed with Ciprico, that Sun has decided not to allow
Ciprico, and presumably other controller vendors, access to boot-level
code necessary to allow booting from Ciprico attached disks. This has
been possible previously up through SunOS 3.5 but is being disallowed by
Sun as of release 4.0 THIS IS A VERY DISTURBING TURN OF EVENTS!
We have stopped buying equipment from Digital Equipment Corp. due to
similar restrictions, and Sun is quite likely to push us into considering
other vendors who may be more accomodating. This is quite unfortunate
coming from a company which promotes so-called "open architecture."
By way of background, we purchased the Ciprico 3220 controller
specifically to attach CDC 9720-1230 disk drives which transfer data at 24
MHz. Sun was not at the time, and may not yet be, selling a controller
capable of supporting drives of that speed. If the Xylogics 7053 had been
available we might well have purchased it, but we were certainly not going
to buy Sun's 900 MB disk for around $14,500 (with discount) when we could
buy the CDC 1.2 GB drives for about 9,500! Besides, the CDC drives are
half the physical size of the Sun-supplied drives.
If Sun is determined to be like DEC, who in turn seems to want to be like
IBM, I guess we'll have to learn to live with that, but it does seem like
a shame!
Dan Grim
Manager, EE/CIS Research Computer Lab
University of Delaware
------------------------------
Date: 27 Jun 88 20:09:55 GMT
From: ulowell!applix!jim at eddie.mit.edu (Jim Morton)
Subject: 4.0 Sun logo binder spine labels
I have taken the previous posting of the Sun logo binder spine labels for
the SunOS documentation and modified them for 4.0, as well as cleaning up
the fonts and logo. This postscript file contains labels for the entire
4.0 "Docubox" documentation set, which comes out to 15 2-inch binder
manuals.
You can mail me a request for this file if it doesn't make it into the
archives.
Jim Morton, APPLiX Inc., Westboro, MA
UUCP: ...harvard!m2c!applix!jim
jim at applix.m2c.org
[[ It is 6107 bytes and it has been placed in the archives as
"sun-source/sun4.0labels.ps". It can be retrieved via anonymous FTP from
the host "titan.rice.edu" or via the archive server with the request
"send sun-source sun4.0labels.ps". For more information about the archive
server, send a mail message containing the word "help" to the address
"archive-server at rice.edu". --wnl ]]
------------------------------
Date: Tue, 28 Jun 88 10:49:57 mdt
From: era at scdpyr.ucar.edu (Ed Arnold)
Subject: tape switch sources
Thanks to the sun-spots readers who responded to my recent inquiry about
sources for a tape switch for Pertec-interface drives. For anyone who may
be interested (i.e. who doesn't like w-a-i-t-i-n-g for dumps over
ethernet, and can't afford to buy a tape drive for every system), the
sources are listed below. Aviv and Iverson appear to be the major
sources; the rest either weren't reachable by phone, or didn't seem
terribly interested in selling a product.
Don't bother to ask your Sun rep about this equipment, because Sun doesn't
sell it (our rep had never heard of such a thing). Should you decide to
buy one, however, be sure to let your Sun rep know that you had to go
elsewhere.
Aviv Inc., Woburn, MA; 617-933-1165
Iverson Inc., San Anselmo, CA; 415-459-5665
Eakins Associates, Mtn View, CA; 415-969-5109
Gafford Technology, CA; 415-499-1673
South Hills Electronics, PA; 800-245-6215 (Bob Karabinos)
Response Computer Technology, Fremont, CA; 415-490-3851
Ed Arnold * NCAR (Nat'l Center for Atmospheric Research) * Mesa Lab
PO Box 3000 * Boulder, CO 80307-3000 * 303-497-1253
era at ncar.ucar.edu (128.117.64.4) * {ames,gatech,noao,...}!ncar!era
------------------------------
Date: Tue, 28 Jun 88 10:06:56 EDT
From: ufnmr!ufnmr_1!gareth at bikini.cis.ufl.edu (Gareth J. Barker)
Subject: touchup - rasfilter8to1
I've just picked up a copy of 'touchup' from the archives and am very
impressed, although I find the '-n' option essential for working on our
color 3/110. I have a problem, however, with printing out the final
results.
We have a non-postscript HP LaserJet II so I can't use pssun or similar on
the rasterfile, so I intended to use 'prraster' (also from the archives).
This requires running the rasterfile through 'rasfilter8to1 (1)' to strip
it down to one bit deep. When run on a 'touchup' rasterfile,
'rasfilter8to1' complains about an "incorrect ras_maptype". An octal dump
of the rasterfile header shows a ras_maptype of 0 (ie RMT_NONE), and a
ras_maplength of 0. (RMT_NONE is described as 'Sun supported' in
'/usr/include/rasterfile.h', but I suspect that 'rasfilter8to1' is wanting
RMT_RAW, which is 'Sun registered'). Has anyone out there overcome this
problem? Would it be easier to modify touchup to generate a different
maptype, or to write something to convert the rasterfile?
Any help would be appreciated,
Gareth J. Barker,
University of Florida,
Department of Radiology.
BITNET : GJBARKER at UFFSC.BITNET
INTERNET : ufnmr!gareth at BIKINI.CIS.UFL.EDU
UUCP : ...gatech!uflorida!ufnmr!gareth
------------------------------
Date: Mon, 27 Jun 88 14:42:01 PDT
From: wolf%ssyx.UCSC.EDU at ucscc.ucsc.edu (Mike Wolf)
Subject: PC-NFS problems
I've been using Sun's PC-NFS package on a PS/2 model 30. I've had a
couple problems with the software.
1) The PC-NFS routines won't let you bind two sockets to the same port
number, even if they are in different domains. i.e. you can't
have a TCP and a UDP socket bound to the same port. BSD sockets
will let you do this. In fact, the port-mapper for RPC requires
both a TCP and a UDP socket bound to port 111.
2) The PC-NFS software intercepts all incoming messages on port 2049.
Port 2049 is the port an NFS server must be listening on. Also,
PCNFS doesn't give an EADDRINUSE error when a socket is bound
to that port. I don't understand why Sun chose to have the NFS
client bind to port 2049. Port 2049 is only necessary for a
NFS server. In addition, if the client uses port 2049, it is
impossible to have both an NFS client and server running.
Has anyone else run into these problems? If so, I'd be interested in
hearing your solutions.
-- Michael Wolf
ARPA: wolf at ssyx.ucsc.edu
UUCP: ...ucbvax!ucscc!ssyx!wolf
------------------------------
Date: 28 Jun 88 00:14:28 GMT
From: mkhaw at teknowledge-vaxc.arpa (Mike Khaw)
Subject: iebark reset crashes
Since about 3 weeks ago, nearly all of our several 3/110s keep crashing with
ie0: iebark reset
ie0: panic: iexmitdone: tbds out of sync
They have boot PROMs at different rev. levels; some are running SunOS 3.4,
some 3.5. Some have 8 Mbytes of Sun memory, others have 4 meg. Sun and 12
meg. Clearpoint. 3 are located near each other on the main ethernet
cable, while others are some distance away and 1 is even on the other side
of an ethernet bridge.
We also have a lone 3/75 that's crashing with the same problem; a 3/110
that *hasn't* had the problem (it's close to the 3 crashing ethernet
neighbors); and a 3/160 server that *doesn't* have this problem. That's
all the Suns with ie0 interfaces. None of our 3/50s or 3/60s are affected
(they have le0 interfaces).
I think we can rule out the Clearpoint memory as a contributing cause.
Sun support doesn't seem to know what the cause might be. They even tried
swapping a CPU board on one of the crashing 110s with no effect. None of
our other machines on our ethernet is experiencing any network related
problems.
Anyone have a clue what's going on?
Thanks,
Mike Khaw
internet: mkhaw at teknowledge.arpa
uucp: {uunet|sun|ucbvax|decwrl|uw-beaver}!mkhaw%teknowledge-vaxc.arpa
hardcopy: Teknowledge Inc, 1850 Embarcadero Rd, POB 10119, Palo Alto, CA 94303
------------------------------
Date: Mon, 27 Jun 88 16:23:51 PDT
From: wrs!jerry at sun.com (Jerry Fiddler)
Subject: large-screen/projection monitors?
I'm looking for a Sun-compatible monochrome large-screen monitor or
projection video system to use at trade shows for demonstrations. The
biggest monitor I've found is a 25" available from Moniterm. Does anyone
know of anything larger, or of a reasonably priced video projector?
Also, will I have any video degradation problem connecting the new monitor
in parallel with the existing monitor? I want them showing exactly the
same image.
Jerry Fiddler
Wind River Systems
sun!wrs!jerry
------------------------------
Date: Tue, 28 Jun 03:57:54 1988
From: ihnp4!sys1!pat (Patrick Dryden)
Subject: VITec anyone?
Just a query to see if you've heard about the VITec immage-processing
system for Sun stations. This north-Dallas company began shipping product
in February after two years designing their parallel image-processor (PIP)
chip and Image Display List-Processing (IDLS) software. The three-board
subsystem fits inside the Sun -- not in an external box like Pixar -- and
delivers 172 MIPS worth of interactive image-processing power. The VITec
runs under UNIX(r) and X Windows with C toolkit and other goodies for
OEMs/developers. I've seen some nifty sample applications in aerial
reconnaisance, color separation for electronic publishing, lots more.
If you would like to get complete information for sun-spots, I'll hunt up
their phone or ask them to mail a packet to you.
Full name = Visual Information Technology in Plano, TX.
------------------------------
Date: Tue, 28 Jun 88 10:58:34 EDT
From: Ned Danieley <ndd at sunbar.mc.duke.edu>
Subject: experience with array processors?
Does anyone have any experience using an array processor with a Sun 3 (or
Sun 4)? Sky Computers is advertising their Warrior II as running at up to
26 MFLOPS and selling for $14900. Would this be a reasonable addition to a
4/110? Any advice or comments would be welcome.
Ned Danieley (ndd at sunbar.mc.duke.edu)
Basic Arrhythmia Laboratory
Box 3140, Duke University Medical Center
Durham, NC 27710
(919) 684-6807 or 684-6942
------------------------------
Date: Tue, 28 Jun 88 08:32:34 PDT
From: Steve Blair <spar!ascway!scb at decwrl.dec.com>
Subject: MAX Groups Question
I have to work within multiple groups and have recently run into a
problem. When I am in more than 6 groups, I get the following: (after
logging-in)
init:too many groups
prompt%
Is there a hack I can install to make it a larger number? I really need to
be in these multiple groups. ADB hacker instructions no problem. Please
send the adb fix to the file(/etc/init(?)) directly to me!
Steve Blair
Unix System Administrator
Austin Systems Center
Schlumberger Well Services
uucp:{backbone}!sun!decwrl!spar!ascway!scb
internet: spar!ascway!scb at decwrl.dec.com
[[ No way. The upper bound is determined by an array in the user
structure that is fixed at 8 entries. Impossible to fix in a binary, hard
to fix in the source. --wnl ]]
------------------------------
End of SUN-Spots Digest
***********************
More information about the Comp.sys.sun
mailing list