Sun-Spots-Request@RICE.EDU (William LeFebvre) (08/10/88)
SUN-SPOTS DIGEST Tuesday, 9 August 1988 Volume 6 : Issue 176
Today's Topics:
Re: Helios RAM
Re: Sun video to VCR
Re: Running screenblank on a Sun 3/60 with color display
Re: Misery loves company...
Re: xt hard errors
dynamic libraries (a suggestion to sun)
problem with xstr
RPC help needed
lucites mailing list
Need sendmail .cf files
4.0 device driver interface bug?
Secure RPC, Lock Manager, Status Monitor info?
Campus mailserver?
Calentool mods: recurring engagements?
Latest 3.x version?
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: 3 Aug 88 12:06:56 GMT
From: steinmetz!olympus!botticel@uunet.uu.net (David J. Botticello)
Subject: Re: Helios RAM
Reference: v6n157
This is a note to the query on helios ram. I ordered a Sun 3/60 with only
4 meg at the end of February. I also ordered 4 meg of additional memory
from Helios on the same date. The helios memory showed up the following
week. The Sun arrived at the end of June. So it seems they can ship!
(Note: the chips on the Helios board ar not the same manufacturer as those
on the Sun boards. If Sun is limiting themselves to a particular
manufacturer, than could explain why Helios has boards to ship and Sun
doesn't)
I had another 3/60 on loan from another department at the time and
installed the memory into it (bringing that machine up to 12 meg) and it
has worked fine. If there are any problems with the Helios memory boards
I haven't seen it yet. And considering the cost at the time was over $600
dollars less for the 4 meg from Helios, seems like I made the right
choice.
Dave Botticello - GE CRD - phone(voice) 518-387-6065 Home 518-393-8671
E-mail - botticel@orion.steinmetz.ge.com -modern domain address or
uunet!steinmetz!orion!botticel -archiac source routing version or
botticel%orion.tcpip@ge-crd.arpa -thru the Arpanet or
botticel%orion@steinmetz.UUCP -You guessed it...thru UUCP
land - 1 River Road KWC-1604
Schenectady, N.Y. 12301
------------------------------
Date: Wed, 3 Aug 88 18:57:35 EDT
From: henry@zoo.toronto.edu
Subject: Re: Sun video to VCR
>Does anyone have any experience capturing the video output from a Sun
>workstation to video tape? ...
There is a non-trivial problem here, because the Suns (at least, the ones
I'm familiar with) use very non-standard high-speed video that your video
tape gear will not understand. No way, no how. Scan conversion between
video formats is possible but outlandishly expensive. Almost certainly
the simplest thing to do is to just point a video camera at the screen.
This works better than you'd think. We did it quite a bit in the CHI+GI
87 demos sessions (which I was co-chair of), and had only one real
problem: you must bear in mind that conventional video gear simply cannot
reproduce the resolution of a Sun screen, meaning that you have a choice
of a sharp closeup of one part of the screen or a fuzzy overall view. If
you want detail, you'll need an alert cameraman who can zoom in on the
right parts of the screen at the right times.
Henry Spencer @ U of Toronto Zoology
uunet!mnetor!utzoo!henry henry@zoo.toronto.edu
------------------------------
Date: 3 Aug 88 16:47:04 GMT
From: chris@spock.arizona.edu (Chris Ott)
Subject: Re: Running screenblank on a Sun 3/60 with color display
mike@ninja.cc.umich.edu (Michael Nowak) writes:
> What is the proper way to run screenblank on a color Sun 3/60? Whenever
> the screenblank program blanks out the screen, we can't get the console
> back. Does suntools need to be run in a special way to allow this?
This isn't just on the color displays. I had the same problem on my
system, too. We have a Sun 3/260 and a few 3/50's. Screenblank worked fine
at first. Then, for no apparent reason, it just stopped working. I never
did figure out what the problem was. Fortunately, they have more Suns in
another lab to which we have Ethernet access. I just copied their
screenblank program to our system and it worked fine again.
Our screenblank broke the same way yours did: the screen would blank and
not come back on again. Being able to copy it from another system to make
it work tells me that something happened to the file, but the chances of
the program still running (without a segmentation fault or something) AND
both of us having exactly the same problem tells me that there must be
some serious weirdness going on.
Anybody have any other ideas?
Chris Ott Internet: chris@spock.ame.arizona.edu
Computational Fluid UUCP: {allegra,cmcl2,hao!noao}!arizona!
Mechanics Lab amethyst!spock!chris
University of Arizona
------------------------------
Date: Wed, 3 Aug 88 15:53:16 edt
From: oxtrap!rich@umix.cc.umich.edu (K. Richard Magill)
Subject: Re: Misery loves company...
Reference: v6n156
In article <1988.07.29.11.44.14.892.27239@rice.edu> you write:
>From: bzs%bu-cs.bu.edu@bu-it.bu.edu (Barry Shein)
>> ...
>> ----- Transcript of session follows -----
>> ld.so: swap space exhausted for mmap data of /usr/lib/libc.so.0.10
>> 554 xxx@xxx... unknown mailer error 127
Stop me if I'm wrong, but doesn't this imply that memory is full yet
nothing else in memory is using /usr/lib/libc.so?
rich.
------------------------------
Date: Thu, 4 Aug 88 4:49:58 EDT
From: scs%lokkur.UUCP@umix.cc.umich.edu (Steve Simmons)
Subject: Re: xt hard errors
Reference: v6n161
Neil Bodick <Bodick@cis.upenn.edu> writes in SunSpots v6n161:
>We have a 3/160C, a Fujitsu M2361A disk, and a Fujitsu M244X tape drive
>with cache, running SunOS 4.0 Release. Whenever I try to write a file
>from the local disk to the local tape I get the following errors:
>[lists various xt hard errors]
We had similar problems at Schlumberger. The cause was worn out hubs on
the tape drives. As the hubs wore out, they didn't hold the tapes quite
as firmly. Eventually during a reposition the tape wouldn't move to the
right place, and whambo -- hard errors.
To check to see if you have the problem, open the front panel of your
drive. On the tape hub(s) you will see 3 small wheels sticking out. If
they're kind of chewed up, you've got the problem. On some drives at
Schlumberger we had worn out 2 sets!
It was hard to get replacements as I recall. For a stopgap, put masking
tape around the little wheels. It'll increase the tension enough to get
you thru until your new ones are delivered.
Steve Simmons, Inland Sea Software, Ltd. scs@lokkur.uucp
9353 Hidden Lake, Dexter, MI. 48130 313-426-8981
UNIX Admin and Training Services
------------------------------
Date: Thu, 4 Aug 88 09:38:29 CDT
From: kimery@helicon.math.purdue.edu (Sam Kimery)
Subject: dynamic libraries (a suggestion to sun)
[This started out on the Sun-Nets mailing list as a discussion on the new
"plug-n-play" daemon shipped with the 386i. It managed to degenerate to a
"Sun has a bad attitude" discussion with named (not named and YP) support
being mentioned as an example]
My feeling is that SUN screwed up an rather easy problem. (:-)) The
strong points for dynamic linking are:
1. smaller binaries
2. easier maintaince (just build a new library)
3. There must be others, just not on my mind :-)
So far SUN has 1. 2 would have been SO easy. Currently, SUN ships a
libc.a for things compiled without dynamic linking. This is easily ripped
apart with ar(1) resulting in a collection of .o's. Why not ship a
libc.a.pic? The stuff could have just been compiled with the -pic option
to cc, and stuffed in a library. This allows ME to pull the sucker apart,
compile my OWN replacement modules (ie. gethostby* that uses named ONLY)
with the -pic option and BUILD A NEW dynamic library. A rather simple way
to cause things to use named without running YP.
Of course, the long term solution is for a tool than can manipulate
dynamic libraries, much like ar(1).
Sam Kimery - System Admin. - Math Department Purdue University
UUCP: !{decwrl,gatech,ucbvax,ihnp4}!purdue!kimery
ARPA: kimery@purdue.edu BELL: 317-494-6055
------------------------------
Date: Wed, 3 Aug 88 18:29:07
From: Geoffrey D. Alexander <alexande@cs.unc.edu>
Subject: problem with xstr
I am using xstr as follow:
cc prog.c -E | xstr -v -
cc x.c -c -O -o x.o
cc xs.c -c -O -o xs.o -R
cc x.o xs.o -o prog
rm x.c x.o xs.c xs.o
strip prog
In some cases the C preprossor can generate fairly large lines, especially
when processing nested macros. I am running to a problem with xstr
producing a syntatically incorrect translation on these large lines. It
appears that xstr is partially substituting for a string. For example,
the string "test it" is substituted with (&xstr[0])t it in x.c where
&xstr[0] references the string "tes" in xs.c. Does anyone have a
suggestion on how to get around this problem with xstr?
Note that I am running on SUN BSD UNIX 4.2. Please e-mail responses
directly to me.
[[ What is "SUN BSD UNIX 4.2"? Which version of SunOS are you running?
Sun has its own version numbering scheme. --wnl ]]
Thanks,
Geoff
------------------------------
Date: Wed, 03 Aug 88 16:33:30 -0400
From: bromley@pizza.bbn.com
Subject: RPC help needed
To SUN RPC wizards:
The problem:
I am using the client broadcast facility (clnt_broadcast()) to access some
servers. Frequently, no server is running on any machine. Normally, I
detect this when a TIMEOUT occurs. The problem is that it takes about 45
seconds for this to occur. 45 seconds is just too long to wait.
1) Can I pre-check to see if the clnt_broadcast request will timeout?
2) How do I modify the default timeout value for a clnt_broadcast request?
(I think I know how to modify the timeout interval for a point-to-point
request.)
Slightly unrelated question:
3) Is there a way to run multiple instantiations of a server program on a
SINGLE cpu such that a clnt_broadcast request will reach BOTH
instantiations of the server program?
Please send your wisdom to me personally at
bromley@bbn.com
or if you prefer actually talking to someone (617) 873-3785
Thanks in advance
Gregg Bromley
------------------------------
Date: 2 Aug 88 00:40:13 GMT
From: munnari!trlamct.oz.au!andrew@uunet.uu.net (Andrew Jennings)
Subject: lucites mailing list
I seem to have lost contact with the lucites mailing list. (For users of
Lucid Common Lisp) : does this list still exist ? How do I get back on ?
Also a while ago I saw somewhere a utility for compacting Lisp run-time
images : can anybody help me with the wherabouts of this ?
Thanks.
UUCP: ...!uunet!munnari!trlamct.oz!andrew
ARPA: andrew%trlamct.trl.oz@uunet.uu.net
Andrew Jennings AI Technology Telecom Australia Research Labs
------------------------------
Date: Wed, 3 Aug 88 20:43:37 PDT
From: John Bossert <bossert@thalatta.com>
Subject: Need sendmail .cf files
I need sendmail.cf files for a Sun network, everyone running SunOS 3.5
One machine serving as a main machine, with the smail routing database,
with multiple subsidiary machines feeding it. It is not desired to
replicate the database on the subsidiary machines.
We're a registered second level domain.
Please send mail directly to me and I'll summarize to sun-spots if there's
interest. Thanks in advance.
John Bossert
Thalatta Corporation
bossert@Thalatta.COM
------------------------------
Date: 3 Aug 88 22:04:37 GMT
From: dcdwest!phb@ucsd.edu (Peter H. Berens)
Subject: 4.0 device driver interface bug?
Is there a bug or an undocumented change in the 4.0 version of the kernel
routine "mbsetup"?
The 4.0 documentation says mbsetup can be used to get the VME address
(mbcookie) of the virtual address contained in the buffer header passed to
it. This seems to work fine for PHSIO transfers to/from user space and
real kernel buffers from the buffer pool. However, when one contructs
one's own buffer header by using the prescription in the device driver
manual:
bp->b_flags = B_BUSY;
bp->b.un.b_addr = buf;
bp->b_bcount = sizeof(buf);
where buf is a small (10k) array in kernel (device-driver), and give this
to "mbsetup", the value returned is NOT the VME address of this kernal
address. Running the returned value through "MBI_ADDR" does nothing for
it as well nor does using the "mballoc" interface to "mbsetup". This
works in all versions of 3.X!!!
Does anyone else have any experience with this? How is one supposed to
get a kernel address mapped onto the DVMA space? What is the "mbpresetup"
that is talked about in buf.h? Who uses and sets the b_mbinfo field of
the buffer header?
All these things appear to be undocumented in 4.0. Can anyone help?
Thanks in advance.
Pete Berens
ITT
------------------------------
Date: Wed, 3 Aug 88 15:30:48 EDT
From: wang!ellison@decvax.dec.com
Subject: Secure RPC, Lock Manager, Status Monitor info?
Can anyone provide a pointer to where I can locate the following
specifications within the Public Domain (or at some nominal charge):
Secure RPC protocol
Lock Manager protocol
Status Monitor protocol
Thanks for your help. I will submit responses to SUN-SPOTS.
Mark Ellison
Wang Labs 014-590 ..!decvax!wang!ellison
One Industrial Av (508) 967-2583
Lowell, MA 01851
------------------------------
Date: Thu, 4 Aug 88 08:38:55 EDT
From: <dennis@williams.edu>
Subject: Campus mailserver?
We are interested in having a campus mailserver. We will initially
include faculty and staff (~500 users) and increase to 3,000 when we add
students. Does anyone have experience with this size clientele on a
single SUN server?
Dennis Aebersold
dennis@cs.williams.edu
aebersol@bitnet
------------------------------
Date: 3 Aug 88 20:15:44 GMT
From: Jerry Nelson <jerryn@tekigm2.tek.com>
Subject: Calentool mods: recurring engagements?
Reference: v6n159
Has anyone modified calentool to handle recurring engagements ( eg.
weekly, monthly, and annual events )? I'd love to get rid of appt, but
for now, that's the only way to handle these kinds of dates.
------------------------------
Date: 3 Aug 88 16:28:05 GMT
From: enea!cad.luth.se!sow@uunet.uu.net (Sven-Ove Westberg)
Subject: Latest 3.x version?
I want to know which version is of 3.x is the latest. Is 3.6 released? Why
do I stay with 3. when all now is talking about the "beta public 4.0". I
have two reasons:
1) When Sun Microsystems Sweden visited us last spring they said that we
as a university should not have to pay any extra for the compilers if we
ordered them together with the operating system. They also said that they
where forced by some crazy laws in the states to unboundle the software.
But now they say something quite different.
2) 4.0 is a beta public release. I want to know how they test their
software before a release when fixed 3. bugs appear in 4. ? Why didn't the
merge the old bugfixes in too this release or at least test if the bug is
there? With a lot of bugs you have to use a lot of your time to localize
what is wrong. (Should I send my 1000000 old Fortran code to them and say
gives the wrong result works vax and ibm etc). From all other computer
companys we got list of known bugs, so you can get an idea what to look
for. Should we start a comp.bugs.sun where we can archive all the bugs?
It is our bug reports and not Suns. How many man weeks do we spend to
track down known bugs??
Sven-Ove Westberg, CAD, University of Lulea, S-951 87 Lulea, Sweden.
ARPA: sow%cad.luth.se@ucbvax.berkeley.edu (only dumb ARPA mailers)
Internet: sow@cad.luth.se
Bitnet: sow%cad.luth.se@sekth.bitnet
------------------------------
End of SUN-Spots Digest
***********************