Sun-Spots-Request@RICE.EDU (William LeFebvre) (01/29/88)
SUN-SPOTS DIGEST Wednesday, 27 January 1988 Volume 6 : Issue 8
Today's Topics:
Re: Removable media
If it's lessons you want...
Sun User Group
Sun/2 Computers for Sale
A weird NFS bug
Problems with in.routed and the yp
Is there a lprm command? (2)
Question about mt and xt devices
Reading mouse using SunCGI from f77?
Sunlink/DNA bug?
Sun-3/110 and Ikon DR11-W Emulator?
High Speed Serial Communications HW/SW?
VersaCAD for Sun?
Re: Spreadsheets for Sun?
----------------------------------------------------------------------
Date: 18 Jan 88 21:36:10 GMT
From: potomac!potomac.ads.com!jtn%sundc@uunet.uu.net (John T. Nelson)
Subject: Re: Removable media
> From: km@emory.UUCP (Ken Mandelberg)
>
> Sun doesn't sell removable media disk drives, but I'm sure there must be
> third party vendors that do. Can anyone tell me what is available?...
At ADS on the East Coast we have daisy-chained a Century Data Systems T306
disk drive to our Fujitsu Eagle running off of a Xylogics 451 controller.
The only problem we have encountered is an overall slowness for disk
accesses. I attribute this to the SMD terminator being on the first drive
rather than the second (we left the terminator on the Eagle which is the
root drive).
The T306 is basically an RM05 in another cabinet although I believe it has
a faster overall access time. We have found the T306 to be quite reliable
and economical to maintain (just change filters every now and then).
I choose this drive because it was inexpensive (about $8,500) and takes
inexpensive ($500 each) readily available disk packs.
John T. Nelson UUCP: sun!sundc!potomac!jtn
Advanced Decision Systems Internet: jtn@potomac.ads.com
1500 Wilson Blvd #512; Arlington, VA 22209-2401 (703) 243-1611
------------------------------
Date: Mon, 18 Jan 88 13:34:15 PST
From: ho@tis-w.arpa (Hilarie K. Orman)
Subject: If it's lessons you want...
...then try one in editing. It is not necessary or useful to insult
your readers for asking questions, as in:
> [[ It's a lesson in reading ability, and it is amazing how many people
> fail. The correct host is "titan.rice.edu", as every editor's comment on
> the subject has clearly stated. --wnl ]]
It wouldn't hurt you to include the archive machine name in the digest
header for those of your readers who take the digest casually. How about:
Submissions: sun-spots@rice.edu
Administrivia: sun-spots-request@rice.edu
Archives: anonymous, titan.rice.edu
[[ Every time I insert a message regarding source that has been placed in
the archives, I include a brief but complete description of how one may go
about retrieving it, including the phrase `...via anonymous FTP from the
host "titan.rice.edu".' This comment always appears immediately after the
poster's original message that describes the software. I can't think of a
better place to put it, and I don't know how much clearer I can be. I
have gotten many messages sent to sun-spots, the -request address, and my
personal mailbox, informing me that the sender tried to access the
archives on "rice.edu" and it did not let him/her log in. I did not mean
to be insulting. I was merely stating a fact: many people seem to miss
the correct host name. I have been considering a list of addresses as you
suggested, but do you really think it would help that much? --wnl ]]
------------------------------
Date: 15 Jan 88 15:50:41 EST (Fri)
From: keithw@lehi3b15.csee.lehigh.edu
Subject: Sun User Group
What info can you send me about the Sun User's Group in Mountain View CA.
Is Sun-Spots associated with this organization? Is it worth joining the
Sun User's Group? Is their PD tape contain what is on your Rice.edu host?
[[ SUG can be contacted at the address "users@sun.com" or "sun!users".
The Sun User Group is not associated with Sun-Spots in any direct way.
Any relation between the SUG software tape and the contents of the
Sun-Spots archives is coincidental. I am personally not a member of SUG,
but I understand that they are a useful organization. They do
periodically publish a newsletter called "README". --wnl ]]
How can I get a listing of what files exist on Rice.edu and Titan.Rice.Edu
for the Sun Workstations?
[[ No publically accessible files are available on "rice.edu"---that is
strictly our mail relay. "titan.rice.edu" accepts anonymous FTP access.
There are four interesting directories: public, sun-spots, sun-source,
and sun-icons. Each contains a file called "00index" which (hopefully)
describes the contents of the directory. Non-ARPA users can access the
archives via the archive server. Send a message containing the word
"help" to the address "archive-server@rice.edu". If you never receive a
response, it is almost certainly because the server's reply, which is sent
to the "From" or "Reply-to" address in your message, failed. This happens
all too often. If you suspect that this is a problem, you may include a
separate line with your request that contains "path address" where
`address' is a legal mail address *relative to rice.edu*! --wnl ]]
Thanks, Keith
------------------------------
Date: Mon, 18 Jan 88 09:56 EST
From: <STELLABO@CSHLAB.BITNET>
Subject: Sun/2 Computers for Sale
**** Sun/2 COMPUTERS FOR SALE ******
The following Sun 2 Equipment is available for sale from Cold
Spring Habor Lab:
QUAN ITEM ASKING PRICE
---- ---- ------------
1 Sun 2/50 2mb Memory
19" Bitmapped Display $1,200.00
1 Sun 2/120 Deskside
Workstation
4mb Menory
71mb SCSI Disk
1/4" Tape
19" Bitmapped Display $2,000.00
1 Sun 2/170 Rackmounted
File Server Workstation
6 mb Memory
XyLogics 450 Controller card
1/4" Sun SCSI tape
6 RS-232 Serial Ports
19" Bitmapped Display $2,000.00
A the equipment is clean and in good working condition.
Inquires/offers should be Directed to:
Fred J. Stellabotte
Computer Systems Manager - Cold Spring Harbor Laboratory
Bungtown Road, Cold Spring Harbor, NY 11724
STELLABO@CSHLAB
(516) 367 - 8420
------------------------------
Date: Tue, 19 Jan 88 08:33:04 +0200
From: Jyrki Kuoppala <t33876w%puukko.hut.fi@fingate.bitnet>
Subject: A weird NFS bug
I generally use rdist(1) to update my .rhosts, .cshrc and .login files on
our BSD machines. When I got an account on a Sun on which users' home
directories were NFS-mounted, added that host to my distfile and run rdist
as usual, I noticed strange behavior. When I logged in to that host, I
couldn't get in without a password. After getting a root shell and doing
some testing I reduced the problem to following :
(This must be done on a NFS-mounted file system)
Script started on Wed Jan 13 22:40:04 1988
% mkdir foo
% chmod 777 foo
% umask 077
% cp /etc/group foo/bar
% cat foo/bar
wheel:*:0:
...
% su root
Password:
# cat foo/bar
cat: read error: Permission denied
# ^D
% cat foo/bar
cat: read error: Permission denied
% ls -la foo
total 3
drwxrwxrwx 2 jkp 512 Jan 13 1988 ./
drwxr-xr-x 6 jkp 512 Jan 13 1988 ../
-rw-r----- 1 jkp 111 Jan 13 1988 bar
% cat foo/bar
cat: read error: Permission denied
% cp /etc/group foo/fii
% cat foo/bar
wheel:*:0:
...
script done on Wed Jan 13 22:42:17 1988
In short, if you create a file readable by owner only and then try to read
it the first time as root it becomes inaccessible for anybody. After a
while it however can be read as normal, and also if you do anything in the
directory where it resides. The same behaviour occurs on Suns and BSD 4.3
NFS.
By the way, what is the true logic of root privileges and NFS? I thought
root is always dropped to -2 (nobody), but if chmod my .rhosts file to 0,
root can still read it when I log in. Anybody care to enlighten? The
NFS on that host is not locally hacked.
Jyrki Kuoppala Helsinki University of Technology, Finland.
Internet : t33876w@puukko.hut.fi / kuoppala@hupu.hut.fi
BITNET : t33876w%puukko@fingate.bitnet
------------------------------
Date: 19 Jan 88 21:29:21 GMT
From: deke@galaxy.ee.rochester.edu (Deke Kassabian)
Subject: Problems with in.routed and the yp
My apologies if this has been discussed before, or if this problem is of
my own creation but.....
I run the yellow pages to keep information like host tables, passwd files,
netgroups, etc, current on my 30+ machines. I also run the route daemon
to dynamically pick up routing information from the far side of my gateway
(anybody else out there running yp AND the routed??).
Recently I updated to SUNIX 3.4 after which my server had incredible yp
problems. The domain was no longer bound (and couldn't be forced with
ypset). Through a few quick changes in rc.local it was obvious that
in.routed was the culprit. If I started the daemon, yp would fail.... if
I didn't, everthing worked normally. I know that there is a patch for 3.4
(my Sun rep is shipping it to me), so if this is a known problem and the
patch is the fix, drop me some mail (full of flames if you like) so that I
know. If not, I'm looking for input from anyone with a similar
experience.
Deke Kassabian, University of Rochester
Deptartment of Electrical Engineering
deke@ee.rochester.edu
or ...!rochester!ur-valhalla!deke
------------------------------
Date: Mon, 18 Jan 88 08:23:41 PST
From: dean@devvax.jpl.nasa.gov (Dean Okamura)
Subject: Is there a lprm command? (1)
I have tried to use the lprm command to remove jobs from the LaserWriter
queue without success when I spooled output from a computer which was not
the server for the printer.
For example, I printed a file from devvax to lw2 which is served by cdb0.
I tried to remove the printer job from the queue with the following
results:
% lpq -Plw2
lw2 is ready and printing
Rank Owner Job Files Total Size
active charng 17 standard input 27077 bytes
1st charng 18 standard input 587 bytes
2nd dean 534 (standard input) 7410 bytes
3rd dean 535 (standard input) 28700 bytes
4th russk 374 /etc/hosts 212534 bytes
5th tj 536 (standard input) 2065 bytes
6th tj 537 (standard input) 17230 bytes
7th cm 208 sfoctape2.lst 19474 bytes
% lprm -Plw2 534 535
cfA534devvax: Permission denied
cfA535devvax: Permission denied
-- I login to cdb0 and check the lw2 spool directory to find:
% ls -l /usr/spool/lw2
total 331
-rw-rw---- 1 daemon 75 Jan 8 11:01 cfA017cdb0
-rw-rw---- 1 daemon 75 Jan 8 11:01 cfA018cdb0
-rw-rw---- 1 root 74 Jan 8 12:11 cfA208cdb2
-rw-rw---- 1 root 69 Jan 8 11:32 cfA374tis0
-rw-rw---- 1 root 66 Jan 8 11:05 cfA534devvax
-rw-rw---- 1 root 66 Jan 8 11:13 cfA535devvax
-rw-rw---- 1 root 62 Jan 8 11:41 cfA536devvax
-rw-rw---- 1 root 62 Jan 8 11:42 cfA537devvax
-rw-rw---- 1 charng 27077 Jan 8 11:01 dfA017cdb0
-rw-rw---- 1 charng 587 Jan 8 11:01 dfA018cdb0
-rw-rw---- 1 root 19474 Jan 8 12:11 dfA208cdb2
-rw-rw---- 1 root 212534 Jan 8 11:32 dfA374tis0
-rw-rw---- 1 root 7410 Jan 8 11:05 dfA534devvax
-rw-rw---- 1 root 28700 Jan 8 11:13 dfA535devvax
-rw-rw---- 1 root 2065 Jan 8 11:41 dfA536devvax
-rw-rw---- 1 root 17230 Jan 8 11:42 dfA537devvax
-rw-r--r-- 1 root 16 Jan 8 12:33 lock
-rw-rw-r-- 1 root 26 Jan 8 12:33 status
Since the files are owned by root for all files that were spooled from
other computers, does this mean that only charng can remove his files from
the lw2 printer queue?
--- Dean
[[ My reply was too long to be an editorial comment, so it's the next
message... --wnl]]
------------------------------
Date: Wed, 27 Jan 88 18:11:35 CST
From: William LeFebvre <phil@Rice.edu>
Subject: Re: Is there a lprm command? (2)
I have seen this problem before. It is not isolated to Suns, but
exists on all versions of Berkeley's lpr spooling system. A remote lprm
request must always come from the same username on the same host that
queued the job. That's why the hostname is contained as part of the file
name.
Why then doesn't it work? Well, when the job is queued, the file name is
actually built by the originating end (devvax in this case) and sent to
the server machine along with the job's data. The originator gets its
hostname with a "gethostname" call, and the server does no validity
checking on this name. But when the lprm request is sent, the server
checks the hostname found in the file name against the *primary* host name
associated with the address of the socket that made the request. So, if
the primary name-server name (or host-table name) is not the same as the
hostname set on the originating host at boot time, the names will not
match and no remote lprm can be done from that host. The hostname is
set at boot time in one of the rc scripts with the "hostname" command.
There are two fixes: either set devvax's hostname to be identical to it's
primary name-server name ("devvax.jpl.nasa.gov") or get Sun to fix lpd so
that it checks the file name against *all* the aliases as well as the
primary name. This bug was "fixed" in 4.3BSD---as distributed, the
hostname is set to the primary name at boot time.
By the way, the first fix tends to annoy people who set their prompt in
their .login with the command: set prompt="`hostname`% " You can get
around long and ugly prompts by using the following teo lines instead
(assuming your local domain is "jpl.nasa.gov"):
set host=`hostname`
set prompt="`basename ${host} .jpl.nasa.gov`% "
William LeFebvre
------------------------------
Date: Mon, 18 Jan 88 12:23 EST
From: SYSRUTH@UTORPHYS.BITNET
Subject: Question about mt and xt devices
I have a bit of a poser to ask.
Among other Suns here, we have a 3/280 and a 3/180, each of which has a
Xylogics 472 interface with a Pertec tape drive on it (made by Fujitsu and
converted to Pertec from the usual STC). The kernel configures these as
xt0 on boot (as it should), and when you are booting the miniroot off
magtape, the device is also referred to as xt0. But on the running system
it appears that when these were connected, no-one ever did a MAKEDEV xt.
As a result, when you try to use xt0 in normal multi-user mode, it
complains. *However*, they can be accessed just fine by calling them mt0
(or any permutation thereof), and they work beautifully.
There are no links of any kind in the /dev area that might cause *mt* to
point to *xt*. In fact there are no xt entries at all. I can understand
why the standalone system wants to call it xt, but why does the running
system accept mt devices and work? Even if the TapeMaster and Xylogics
drivers are similar, I wouldn't have thought that one would run the other
when the drive configures as xt0.
Any ideas? The technical people at the local office were stumped, at least
during the 5 minutes I talked to them.
Thanks.
Ruth Milner
Systems Manager
University of Toronto Physics
SYSRUTH@UTORPHYS (BITNET - preferred) sysruth@helios.toronto.edu (InterNet)
{ihnp4,decvax}!utgpu!helios!sysruth (UUCP)
------------------------------
Date: Mon, 18 Jan 88 15:35:30 EST
From: mckay@courageous.ecn.purdue.edu (Dwight D McKay)
Subject: Reading mouse using SunCGI from f77?
Has anyone tried to read the mouse using SunCGI from Fortran (f77)?? I
have asked my student programmer to try and he has been unsuccessful. I'm
not a big Fortran hacker myself, so I ask you if you can tell me what's
wrong with the following program:
--- cut here ---
program try
parameter (ibignum=256)
include '/usr/include/f77/cgidefs77.h'
integer valid
c parameters to cfopenvws
integer name
character screenname* (ibignum)
integer screenlen
character windowname* (ibignum)
integer windowlen
integer windowfd
integer retained
integer dd
integer cmapsize
character cmpaname* (ibignum)
integer cmaplen
integer flags
character ptr* (ibignum)
integer noargs
c parameters to cfinitlid
c cfreqinp also uses these
integer x
integer y
integer xlist (ibignum)
integer ylist (ibignum)
integer n
real val
integer choice
character string* (ibignum)
integer segid
integer pickid
integer echotype
integer exup
integer eyup
integer exlow
integer eylow
integer trigger
integer timeout
c
c open cgi
call cfopencgi()
c
c open a pixwin
c
dd = 4
call cfopenvws(name, screenname, screenlen, windowname,
+ windowlen, windowfd, retained, dd, cmapsize,
+ cmpaname, cmaplen, flags, ptr, noargs)
c prepare logical input device to be read by
c turning it on via cfinitlid
c initialize_lid
choice = 2
call cfinitlid(0, 1, x, y, xlist, ylist, n,
+ val, choice, string, segid, pickid)
c associate
trigger = 2
call cfassoc(2, 0, 1)
c track_on
call cftrackon(0, 1, echotype, exlow, eylow,
+ exup, eyup, x, y, xlist, ylist, n, val, choice,
+ string, setid, pickid)
c
c request_input
c valid = stat
timeout = 10*1000*1000
call cfreqinp(0, 1, 10000000, valid, x, y, xlist,
+ ylist, n, val, 1, string, segid, pickid, 2)
print *, x
print *, y
c
call cfclosevws(name)
call cfclosecgi()
call exit()
end
--- end of program ---
--Dwight Mckay, ECN Workstation Software Support
[arpanet: mckay@ee.ecn.purdue.edu, usenet: ...ihnp4!pur-ee!mckay]
[Compu-serve: 75776,1521, office: EE 348B, phone: (317) 494-3561]
------------------------------
Date: 19 Jan 88 03:30:59 GMT
From: sun!laic!root@decwrl.dec.com (Operator)
Subject: Sunlink/DNA bug?
I cannot seem to be able to send or receive a 0 length record in
sunlink/dna (the sun seems to just ignore or compress those records).
Since I cannot rewrite the application on the other side of decnet
(vaxmail), does anyone know if there is a way to do this?
Darin Johnson
sun!sunncal!leadsv!laic!darin
------------------------------
Date: Tue, 19 Jan 88 15:45:37 PST
From: nomdenet@venera.isi.edu
Subject: Sun-3/110 and Ikon DR11-W Emulator?
We have a project involving DMA communication via an Ikon 10084 VMEbus
DR11-W emulator. We've gotten the 10084 functioning in a 3/160, but we've
had little luck with a 3/110 -- even taking a functioning board out of our
3/160 and plugging it into the 3/110. Programmed-I/O functions in loop-
back mode work fine, but DMA functions hang; the UNIX process is waiting
for I/O completion, and we can regain access to the 11084 only by
rebooting.
We know about back-plane jumpering, we know about changing the jumper for
loop-back vs. normal-operation modes (W19-20-21), and we've been in touch
with Ikon (Bill Waddington); they're stumped, and have told us (politely)
that we're on our own.
Does anyone in Sun-land have such a 3/110 - Ikon 10084 combination
working? Please?!??!
Thanks,
A. R. White
USC/Information Sciences Institute
4676 Admiralty Way
Marina Del Rey, California
90292-6695
(213) 822-1511, x162
ARPA: nomdenet @ Venera.ISI.edu
------------------------------
Date: 18 Jan 88 22:01:53 GMT
From: benson%ksr@uunet.uu.net (Benson Margulies)
Subject: High Speed Serial Communications HW/SW?
I am interested in information/experience/product plugs for high speed
serial interfaces -- e.g. 1Mbaud RS422/423. DMA would be a real win. It
would be particularly rewarding to hear of such a thing with a SunOS
device driver that came with.
thanks, benson
Benson I. Margulies Kendall Square Research Corp.
harvard!ksr!benson ksr!benson@harvard.harvard.edu
------------------------------
Date: Mon, 18 Jan 88 12:43 EST
From: <PFLUGRAT@CSHLAB.BITNET>
Subject: VersaCAD for Sun?
Does anyone out there in Sun land have VersaCAD running on a color
workstation? If so, how does it compare to VersaCAD on a PC or MacIntosh?
How do you get hardcopy output (plotter, color printer, slide maker)? I
would enjoy reading comments on this from everyone.
Jim Pflugrath
PFLUGRAT@CSHLAB bitnet
roberts.pflugrath@bionet-20 arpanet
------------------------------
Date: Tue, 19 Jan 88 13:30:34 GMT
From: Eric Ole Barber <mcvax!nw.stl.stc.co.uk!sizex@uunet.uu.net>
Subject: Re: Spreadsheets for Sun?
Bob Sutterfield refers to a spreadsheet calculator that went around the
net a while back in Volume 6 : Issue 3. It's called sc. Can anyone send me
a copy please? It's not in sun-source or public, and I cannot ftp
anywhere. Thanks, Eric Barber
[[ It was distributed over the Usenet as part of comp.unix.sources and
should be available in any archives for that newsgroup. Rich Salz (who
frequently spells his last name "$alz") moderates that group, but I do not
have a net address for him handy. --wnl ]]
------------------------------
End of SUN-Spots Digest
***********************