Sun-Spots-Request@RICE.EDU (William LeFebvre) (06/22/88)
SUN-SPOTS DIGEST Tuesday, 21 June 1988 Volume 6 : Issue 118
Today's Topics:
Re: CDC Sabre series drives
Re: Pointer for Clearpoint?
Re: Sun 3/50 eyestrain
Re: Anti-Glare/Polarizing Screen for 19" screens?
Re: bind: address already in use
WARNING: SunOS 4.0 + TrailBlazer problem
Bugs in SunOS 4.0 Export
SunOS 4.0 problem with .rhosts
4.0 SCSI vs. Exabyte
3/280 to PSN connection
lighting, binders
SUN DES chips?
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: Sun, 12 Jun 88 20:47:33 MST
From: ncc!lyndon@uunet.uu.net
Subject: Re: CDC Sabre series drives
We have been running a 9720-850 on our 3/280S for approximately two months
now. We haven't had any trouble with them to date. The drives are
reasonably quite, and produce very little heat. Our configuration is
using a Rimfire 3200 disk controller with a transfer rate of 2.0 MHz.
We did have some initial configuration problems with this drive/
controller combination. The Sabre's support a "sweep mode" that was
causing the controller to hang. Although the drive is equipped with a
jumper to disable this mode, it did not seem to do anything on our drive.
Ciprico has a new set of controller PROMs that solve the problem. You
should probably test the drive with your existing controller to make sure
this won't be a problem for you as well.
------------------------------
Date: 13 Jun 88 12:22 EST
From: CONKLIN JOEL R <CONKLIN@ge-crd.arpa>
Subject: Re: Pointer for Clearpoint?
An excellent source of Clearpoint memory is from a distributor called
Peripheral Devices. The salesman is Tony Gordon (215-640-0446). The
delivery times of all their products (including Clearpoint memory for SUN
3/60's) is very good, and their prices are quite competitive. Tony has
gone way out his way to get and maintain our business. I'm in no way
connected to Peripheral Devices - just a very happy customer.
On a related note, in most cases, it makes no sense to order a SUN 3/60
with anything more than 4 MB of memory. You can get the SIMM modules
cheaper and faster from third party vendors such as Peripheral Devices.
The chips are quite easy to plug in yourself, and come with a lifetime
warranty and 24 hour replacement.
Also, you can not currently order SUN 4/110's with anything other than 8
MB of memory (using the 256k chip). You can order 32 MB of the Clearpoint
DRAM (1 MB) chips, pull out the original 8 MB you got from SUN, and end up
with a 4/110-32MB. I don't yet know what you'd do with the 8 MB of SUN
memory.
If, for some reason, Tony Gordon can't help you out, or you'd rather work
with someone on the West Coast (Peripheral Devices is in Paoli, Pa.),
Helios Systems also sells memory for the entire SUN product line. The
salesman is Mike Brown (408-723-9012). Helios is based in San Jose, Ca,
and also offers good prices, fast delivery, and a lifetime warranty.
Joel Conklin (conklin@ge-crd.arpa)
General Electric Corporate Research & Development
Schenectady, NY
(518) 387-5817
------------------------------
Date: Mon, 13 Jun 88 21:50:46 EST
From: munnari!trlamct.oz.au!andrew@uunet.uu.net (Andrew Jennings)
Subject: Re: Sun 3/50 eyestrain
Reference: v6n105
I get eyestrain too, even though I have excellent vision. I saw an
optometrist about this, and she recommended glasses that apparently let my
eyes default focus at about the right distance, ie. slight magnification,
nothing fancy.
Even though I felt silly doing this, it seems to work. No more problems.
------------------------------
Date: Mon, 13 Jun 88 10:15:47 -0700
From: Joe Kwan <rabbit@psivax.psi.siemens.com>
Subject: Re: Anti-Glare/Polarizing Screen for 19" screens?
Last Friday, my officemate brought in the latest Inmac catalog (June 1988)
and on the cover was a glare filter for 19" workstations. I immediately
sent out an order for two of them (they also offer a filter than cuts
low-freqency radiation which is the type I bought). The price of the
regular filter is $229 (without static and radiation control) and $279 for
the static and radiation cutting filter. They call it the "CAD Glare
Sentry Plus" and have filters for Sun, Apollo, HP, Mitsubishi, Sigma and
NEC 19" monitors.
Sun also sells their workstations now with OCLI filters. Does anyone know
what these filters do? (cut glare, radiation, both?) Sun charges $1000
more for this option. I've also ordered our new workstations with these
filters for comparision. My salesman said he thought the OCLI filters
were probably the filter plus some electronics to brighten the display to
maintain the picture brightness. Does anyone know for sure if this is
true? I suppose we'll find out in a month or so when we get our new
workstations.
Joe Kwan
Systems Analyst
Pacesetter Systems, Inc. A Siemens Company
uucp: {csun, scgvaxd, hoptoad, sdcrdcf, uunet}!psivax!rabbit
domain: rabbit@psivax.psi.siemens.com
------------------------------
Date: Sun, 12 Jun 88 14:47:29 EDT
From: Stephen J. Roznowski <sjr@mimsy.umd.edu>
Subject: Re: bind: address already in use
This error occurs when two daemons (programs) try to use the same port
(address) at the same time. Well known services listen on a well defined
port, and then fork off a copy of themselves and rebind to an open port.
Daemons like SMTP (sendmail) initially listen on port 25; when a connect
request comes in, it forks a copy of itself and rebinds to the first
available port less than 1023 (a priviliged port). [FTP listens on port
21, TELNET on 23; see /etc/services for the complete list]
INETD listens on several ports, and forks off a copy of the correct
program (see /etc/servers for the complete list) to handle the request.
Hope that this answers your question.
Stephen J. Roznowski sjr@mimsy.umd.edu
------------------------------
Date: Sat, 11 Jun 88 22:25:04 CDT
From: texsun!swrinde!maxwell!ksp@sun.com (Keith S. Pickens)
Subject: WARNING: SunOS 4.0 + TrailBlazer problem
There is a bug in the SunOS 4.0 tty driver which results in lost
characters during high speed input. A TrailBlazer based uucp link will
fail when receiving data. This problem was observed on a Sun-3/180
running SunOS 4.0. Sun has observed this bug internally and it is
assigned reference number 1010898.
My understanding is that the fix for this problem is known. However, Sun
is not currently making it available [ :-( ] and has instead scheduled it
for inclusion in release 4.0.1.
If you are considering upgrading to 4.0 be aware of this problem. You may
want to wait until the fix is available. If on the other hand you have
already taken the plunge and your uucp connections are failing, a call to
Sun would be in order.
The above represents my understanding of the situation.
Keith S. Pickens ut-sally!swrinde!ksp
Southwest Research Institute
San Antonio, Texas
(512) 522-3149
------------------------------
Date: Sun, 12 Jun 88 15:06:35 EDT
From: Mark Mendell <mendell@turing.toronto.edu>
Subject: Bugs in SunOS 4.0 Export
I have installed SunOS 4.0 on a couple of our Sun-3 systems here, and
there is one Sun-4 now running 4.0. I have found one major bug, two minor
ones, and a 'feature'
The major bug is that Sun has managed to break 'fputs' in certain
situations. Given the following program:
#include <stdio.h>
main()
{
fputs("\nabcd\n");
}
[[ I hope that isn't the exact program you used, because it's wrong.
"fputs" requires two arguments: the string and the FILE pointer. --wnl ]]
and compiling, output to the terminal will be "\n" only. Output to files
works fine. If you do a printf first, then it works perfectly. This has
been tested on both the Sun-3 and the Sun-4. On the sun-3, I tried
dynamic linking, static linking, and the system V library (dynamic). On
the sun-4, I tried only the dymanic link. They seem to have messed up the
code for line-buffered output in their attempt to optimize it. The code
seems to be vastly different from the 4.3 version of fputs.
This brings up the first minor problem. When the static library for 4.0
export was made, they messed up the following routines:
des_crypt.o
des_soft.o
_crypt.o
auth_des.o
svcauth_des.o
They are obviously completely linked, as the are about 24K in size, and
define _end, _edata, and _etext. This means the -Bstatic won't work until
these routines are removed from the library and it is ranlib-ed.
A related problem is that the 'newkey' program will fail at runtime with a
message that the routine '_cbc_crypt' cannot be found.
I assume that the problems are both due to export restrictions.
The 'feature' that I discovered is the 'fortune' will not work on a
diskless or dataless (/usr mounted from server) workstation, when /usr is
mounted readonly. Doing a strings -1 on fortune gives the impression that
it is doing a fopen(datafile, "r+"). the 'trace' command does show that
it is trying to open the data file read/write. I can't think of any
reason for this! [[ How "un-fortune-ate". --wnl ]]
Mark Mendell
Computer Systems Research Institute University of Toronto
Usenet: {linus, allegra, decvax, floyd}!utcsri!mendell
CSNET: mendell@Toronto
ARPA: mendell%Toronto@CSNet-Relay
------------------------------
Date: Sun, 12 Jun 88 15:13:30 PDT
From: Jim Nisbet <GG.JDN@forsythe.stanford.edu>
Subject: SunOS 4.0 problem with .rhosts
Dear Sun-Spots,
I sent a more detailed description of this problem last week. I'm sure
most people don't want to know all the details of the bug, but I think it
might save someone some time if they knew that this bug is out there...
so here is a shorter version...
In SunOS 4.0 there appears to be a bug in the code which looks up an
equivalenced host/userid in the user's .rhosts file (the ruserok routine
in rcmd gets a seg violation).
The effect is that when you try to 'rlogin' or 'rsh' to a SunOS 4.0
computer you may get an immediate 'connection closed' (because
login/in.rshd got the seg violation and exited).
Sun tech support duplicated the problem last week, and they are looking in
to it. I assume that they will come up with some sort of workaround for
this problem shortly.
------------------------------
Date: Mon, 13 Jun 88 00:46:47 CDT
From: slevy@uf.msc.umn.edu (Stuart Levy)
Subject: 4.0 SCSI vs. Exabyte
Brad Powell's comment (v6#105) which seemed to say that, since the SCSI
driver code had changed between 3.5 -> 4.0beta -> 4.0 release, THEREFORE
it was reasonable to expect that media written under 3.5/4.0beta would be
unreadable under 4.0... was pretty disturbing.
I could understand, sort of, if he were saying that Exabyte drives just
wouldn't work under the 4.0 driver. But no, he expects them to work fine
with tapes written under 4.0.
Can somebody explain how two different drivers for the same SCSI tape
device could be expected to yield mutually incompatible media? Was that
really what you said, Brad?
Stuart Levy
------------------------------
Date: Sun, 12 Jun 88 14:25:51 EDT
From: Stephen J. Roznowski <sjr@mimsy.umd.edu>
Subject: 3/280 to PSN connection
I'm in the process of connecting a 3/160 (but a 3/280 should be no
different) to ARPAnet via a 56K leased line to a PSN. We are using an SCP
[which Sun has replaced by the MCP] and are using Sunlink X.25 DDN.
We have had the node on ARPAnet, but are now having trouble with our
connection to the PSN [Our problem, not Sun's fault].
I would like to report that this connection and installation of software
was painless, but that is not so. When installing X.25 DDN, Sun made the
assumption that an ethernet is connected to the node. Since our only
connection was via an X.25 line, we had problems with the node trying to
do RCP broadcasts to the (non-existant) ethernet interface. Sun has told
me that they did not believe people would be using their product in that
manner since it is marketed as a "connectivity" solution. However, they
since have had other requests (ours was the first) for this type of
connection, and are coming out with a solution.
Of course, there are the little problems with Sun not having domain
support in the sendmail files, a sendmail that does not support MX
records, a resolver that can only be access through yellow pages (and
undocumented at that!!), minimal documentation on in.named, etc.
These would be livable problems, if Sun included documentation with this
package. What little documentation there is, is WORTHLESS in trying to
diagnose problems. [For example, ... if everything is connected ok, you
should see "HDLC is UP. Packet Level is UP." -- we didn't and did not know
where to turn to try to find the problem; the PSN, the line to the Sun,
the SCP, a configuration file, ...]
We would not be as far as we are if it wasn't for the help we have
received from Sun Datacom, especially Stephen Carter and Bill Melohn. The
product as it currently stands is useful, but be ready to pull a lot of
hair out while installing. We have had several discussions with various
people involved in this product, including the product manager, and have
high hopes for many, if not all, of these problems to be fixed in the next
couple of releases.
Stephen J. Roznowski sjr@mimsy.umd.edu
------------------------------
Date: Sun, 12 Jun 88 08:14:18 PDT
From: cgl.ucsf.edu!hoptoad.UUCP!cfcl!rdm@sun.com (Rich Morin)
Subject: lighting, binders
LIGHTING:
> I just got a Sun 3/50 (monochrome display) for my office. While I love
> the thing in general, I've been experiencing marked eyestrain problems.
> ...
> My office has fluorescent lighting...
Hmmm. Actually, there is a very simple workaround: bring in a
supplementary incandescent table or floor lamp. This will (because of the
thermal inertia of the filament) "fill in" between the flashes of the
fluorescent. A typical 100 watt lamp should work just fine...
BINDERS:
> Does anybody know how to map the new sun documentation set to orange DEC
> manual binders?...
No, but I *do* know how to map sun's documentation into the Sun User Group
binder set, since John Gilmore (gnu@hoptoad.com) and I did just that. May
I respectfully suggest that you send your preciousss dollarsss to SUG,
instead? The SUG set uses D-rings, has a nice appearance, and is
generally quite tasty...
Disclaimer: I may be on the SUG BOD, but I sure don't get no $$$ from
their binder (or tape, or SUGtool, or ...) sales.
------------------------------
Date: Mon, 13 Jun 88 00:29:08 CDT
From: slevy@uf.msc.umn.edu (Stuart Levy)
Subject: SUN DES chips?
Is anyone using the AmZ8068 DES chip option for a SUN? I'm interested in
knowing (a) whether anyone has felt a need for it, as opposed to the
software DES routines SUN already supplies, and (b) how well it performs
as seen at the user level -- both single transaction speeds (set key,
en/decrypt 1 block) and bulk data rates.
If anybody has one I'd be happy to send a test program to benchmark it.
I suspect there will be more people lusting after DES hardware now that
4.0 is out, with its (optionally) DES-authenticated RPC and NFS.
I'll summarize and post responses.
Stuart Levy, Minnesota Supercomputer Center
slevy@uf.msc.umn.edu, (612) 626-0211
------------------------------
End of SUN-Spots Digest
***********************