Sun-Spots-Request@RICE.EDU (William LeFebvre) (08/02/88)
SUN-SPOTS DIGEST Sunday, 31 July 1988 Volume 6 : Issue 159
Today's Topics:
Re: NCAR graphics package for the SUN (3)
Re: 4.0 vs kill -HUP 1
Re: Strange Network Traffic (Booting a Sun)
more pwd gotchas
Contest: dump(8) parameters for DC300XL 1/4" cartridge tape
Sun OS and 4.3 networking
Bitmap conversions
thinet-info?
Running screenblank on a Sun 3/60 with color display?
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: Tue, 19 Jul 88 08:10:08 edt
From: mlijews@nswc-wo.arpa
Subject: Re: NCAR graphics package for the SUN (1)
Reference: v6n143
NCAR has been ported to the Suns. I have it running on ours, though we
haven't yet made much use of it. You can order it from:
MINEsoft, Ltd.
2345 S. Federal Bldg.
Lincoln Center, Suite 190
Denver, CO 80219
At least that is the last address I have for them.
------------------------------
Date: Tue, 19 Jul 88 11:17:18 MDT
From: woods@handies.ucar.edu (Greg Woods)
Subject: Re: NCAR graphics package for the SUN (2)
There is indeed a Sun version of NCAR graphics. It costs $500 for academic
sites and $2000 for commercial. You can order it from Adrianne Link
(adrianne@bierstadt.ucar.edu, (303) 497-1201)
--Greg (woods@ncar.ucar.edu)
------------------------------
Date: Tue, 19 Jul 88 16:15:50 CDT
From: tuchman%uicsrd.csrd.uiuc.edu%uxc.cso.uiuc.edu@uxc.cso.uiuc.edu (Allan Tuchman)
Subject: Re: NCAR graphics package for SUN (3)
I got a copy of the NCAR Graphics Software distribution late last fall.
(I was quite familiar with their earlier, pre-GKS version, and slightly
familiar with their first GKS attempt. Following their installation
instructions, I installed the software in just 2 days. This included
doing it "right", with Makefiles and a several page man page describing
its "local" use (but no routine descriptions).
The installation was pretty straightforward. They supply sample low-level
machine dependent routines in C. Most of these worked without change. I
chose to add one bell (or was it a whistle?) of my own.
The only disadvantage is that there is no supplied support for any SunView
graphics. You have two choices for screen viewing: Use tektool on the
Sun and tell NCAR to use the Tektronix 4014 graphcap, or if you can run
NeWS, have NCAR use the PostScript graphcap entry. I tested both of these
as well as hardcopy output to both the Apple Laserwriter and Imagen laser
printers.
As a nice side-effect, the EXACT code that I finally compiled and
installed on the Sun ported without change to our Alliant FX-8's. Just
recompiled it. Now I keep a copy of the source only on the Sun's, but
binaries and libraries on both machines.
-Allan Tuchman University of Illinois
Center for Supercomputer Research and Development
ARPANET: tuchman%uicsrd@a.cs.uiuc.edu
UUCP: {ihnp4,pur-ee,convex}!uiucdcs!uicsrd!tuchman
CSNET: tuchman%uicsrd@uiuc.csnet
BITNET: tuchman@uicsrd.csrd.uiuc.edu
AT&T: (217) 244-0048
------------------------------
Date: Tue, 19 Jul 88 16:49:17 EDT
From: mnetor!utzoo!henry@uunet.uu.net
Subject: Re: 4.0 vs kill -HUP 1
>... it is arguable that logged in users should never
>be kicked off, that init should merely remember the change for the next
>getty, but that's how init works, and always has.
Nope, it's not the way V7 init worked. Another wonderful "Berkeley
enhancement"...
Henry Spencer @ U of Toronto Zoology
{ihnp4,decvax,uunet!mnetor}!utzoo!henry
------------------------------
Date: Tue, 19 Jul 88 19:49:37 PDT
From: Craig Leres <leres@helios.ee.lbl.gov>
Subject: Re: Strange Network Traffic (Booting a Sun)
Paul Lutt writes:
> 3) Vmunix starts up. One of the first things it does is send out
> a very strange looking ARP request:
>
> Packet #1172: client -> ether-broadcast
> ARP request (0.0.38.80 -> 0.0.38.80)
>
> The source and destination internet addresses are totally bogus.
> As expected, no ARP reply is ever received for this bogus request.
This is just the result of a kludge in the reverse arp setup. Before you
can output packets on the ethernet (to do the rarp), you have to
initialize the interface. But to do this, you must supply the ip address
of the interface. Since this isn't know, one is invented using the machine
id and serial number. The comment in revarp_myaddr() says "Hopefully, the
address won't get used" but it does.
> 5) Now another funny packet. The client sends out a strange ARP
> request:
>
> Packet #1175 client -> ether-broadcast
> ARP request (client -> client)
>
> The client has just asked what its own ethernet address is! This
> same request appears at least twice during the booting process.
I recently discussed this in comp.protocols.tcp-ip:
I think the best reason for gratuitously broadcasting an arp reply
for yourself is to force hosts you were talking to (say, before you
shutdown, swapped your ethernet interface, and rebooted) to learn your
new hardware ethernet address.
Almost all Berkeley network derived machines arp themselves on bootup.
Craig
------------------------------
Date: Tue, 19 Jul 88 10:32:34 PDT
From: frame!troy!drf@sun.com (David Fuchs)
Subject: more pwd gotchas
Another old problem with NFS vs. pwd that hasn't gone away in 4.0 is shown
by:
cd /tmp A safe place to play
mkdir mydir Make a new mount point
chmod 000 mydir Give it a strange protection
su Need to be root to do mount
mount foo:/tmp /tmp/mydir Mount some random foreign directory
^d Gets out of su mode
cd mydir This works fine
ls So does this
pwd But this fails!
The pwd command says "pwd: getwd: can't open ..".
The moral to the story is that, while you may think that the mode of a
directory that serves as a mount point is completely inconsequential
(since the mode of the directory that is mounted there is what controls
actual file access), it ain't necessarily 100% so. Of course, the
documentation on mount, pwd, chmod, etc. has nothing to say at all about
anything close to this issue...
-David Fuchs (sun!frame!drf)
------------------------------
Date: Tue, 19 Jul 88 08:28 EDT
From: Roy Richter <RICHTER@gmr.com>
Subject: Contest: dump(8) parameters for DC300XL 1/4" cartridge tape
I've had it with 1/4" cartridge tape. Everybody I ask has a different
opinion on the parameters to use for making a dump(8). I would like to
run a contest to see who can send me parameters that maximize the number
of bytes one can put on such a tape. Assume you are dumping to QIC-24
format (/dev/nrst8 on the Sun). The command line is thus:
dump 0ucbsf $block $size /dev/nrst8 /dec/rxy0g
and I need the values for $block and $size.
[[ How about minimizing the time, as well? You could use that for a
tie-breaker (although where cartridge tape is concernced, I think speed is
more important than compactness). --wnl ]]
Yes, I know Sun recently went to rst8. Some sources say this reduces
total tape usage, others say the tape usage dropped due to shared
libraries in SunOS 4.0. My expirience is that one can use the same
parameters for rst0 and rst8, with similar tape usage. Both can use 9
tracks.
For your information, I use $block=1750 and $size=3825 for a 450 ft. tape
and get 43.5 Meg per tape.
I will post *all* unique sets of parameters after a reasonable time.
(Oh, contest....for a prize, how about <to be negotiated> but not
less than a free tape?)
Roy Richter UUCP: {teemc,edsews,mcf}!rphroy!roy
Physics Dept, GM Research CSNet: richter@gmr.com
------------------------------
Date: Wed, 20 Jul 88 08:59:30 EDT
From: doug@icase.arpa (Doug Peterson)
Subject: Sun OS and 4.3 networking
I've seen a lot of negative comments about Sun's unwillingness to port
(incorporate) 4.3BSD networking capabilities into their operating system.
Those that are incorporated seem to also include the 4.2/3 bugs. These
unhappy people are also the mainstay of the network knowledge base in the
community.
It strikes me that the time is near for such individuals to converge and
produce the Berkeley equivalent of an OS for a distributed computing
system such as a collection of diskless Sun workstations connected to a
file server, which is in turn a gateway to another network. It seems that
such a group could incorporate all the improvements which have been
learned about both in networking and client/server situations, and produce
a more performance oriented operating system, and make it available to the
academic/research community.
Too often (I think), I've seen postings to this list complaining of a Sun
OS (problem/feature/bug), with a response from Sun to the effect that it's
not an issue (ehternet connectors? name resolver?). I'm beginning to sense
a slight overtone of '...If our system doesn't do what it's supposed to
for you, well...' [[ Is it beginning to sound like "big blue"? --wnl ]]
Doug Peterson
Systems Manager
ICASE
Mail Stop 132C
NASA Langley Research Center
Hampton, VA 23665-5225
(804) 865-4090
------------------------------
Date: Tue, 19 Jul 88 15:37:33 PDT
From: gandalf@csli.stanford.edu (Juergen Wagner)
Subject: Bitmap conversions
I have a program in a beta-test version which converts bitmaps between
different formats.
Currently, I am modifying everything such that the conversion utilities
will go into a library where you can just pull them from (for use in
programs which have to read a particular bitmap format). A set of
operations then operates on the resulting internal representation which is
opaque to the user.
I am including the man page for the beta-version of the bitmap conversion
program. It is similar in its claims to Jeff Poskanzer's PBM package, but
is more powerful (more formats, faster, still portable, and occupies only
about 110kB (executable) on disk. The man page gives an impression of
what's supported but is incomplete. The synopsis changed quite a bit in
the new version...
I am also in the course of adding the G4 FAX format to it. When that's
running, I'll post the whole beast to comp.sources.unix. If anybody out
there feels like needing another bitmap format (*NOT* graphics language!),
let me know.
Juergen "Gandalf" Wagner, gandalf@csli.stanford.edu
Center for the Study of Language and Information (CSLI), Stanford CA
[[ The name "bm" is already taken. It is used for a grep-like program
that uses the "Boyer-Moore" searching argorithm. --wnl ]]
-------------------------------------------------------------------------
NAME
bm - convert bitmaps between formats
SYNOPSIS
bm [-fmt] [ops] [file] [options] > newfile
bm -help
DESCRIPTION
Bm provides a flexible tool to convert bitmaps between dif-
ferent formats. A few unary operations are also available.
For more complex bitmap operations, please use the program
'bmop' which also comes with this package. Whereas 'bm'
mostly serves as a conversion program, 'bmop' allows to
manipulate bitmaps in various ways.
Examples:
bm -ri mask.r > mask.icon
converts a Sun rasterfile 'mask.r' to a Sun Iconedit
file 'mask.icon'.
bm -rp mask.r same | lpr -Plw
prints a Sun raster file 'mask.r' on a PostScript
printer lw, keeping the proportions of the image.
bm -rp mask.r | lpr -Plw
does the same thing but extends the image over the
whole page.
bm -rp +invert - | lpr -Plw
reads a Sun rasterfile form standard input, inverts the
image, and prints it on the PostScript printer lw,
extending the image over the whole page.
`bm' takes a format description consisting of two charac-
ters, which describes the source and target format. Conver-
sion also pads the bitmap to width and height becoming mul-
tiples of 16. In addition to the format specification, a
number of operators can be specified. These operators will
be applied to the bitmap after it has been read in, and
before it is output again. Input is taken from a file (stan-
dard input is denoted by '-'), output is written to standard
output.
FORMATS
Formats suitable for both, input and output:
1 (ASCII)
a bitmap of zeroes and ones, quite flexible.
a (ASCII)
a bitmap of blanks and non-blanks.
g (Gandalf's favourite format)
a somewhat dense format.
i (Iconedit)
a bitmap in Sun iconedit format.
p (PostScript)
a PostScript file. For input, only PostScript files of
a special format are recognized. On output, options are
"same" (keep ratio of width to height), "half" (shrink
by factor two), "quart" (shrink by factor four). The
default is to resize the bitmap such that it exactly
fits the output page.
r (Rasterfile)
a Sun rasterfile (RT_STANDARD).
x (Xerox)
a bitmap in XEROX 1108 format.
X (X Window)
a bitmap in X10 or X11 format. On output, options "x10"
and "x11" are available. If byte swapping is needed,
the operator +swap can be used.
Formats suitable for input only:
m (MacPaint)
a MacPaint document.
u (Weird ucbvax format)
The format of those bitmaps retrievable from ucbvax via
anonymous ftp. They are located in the directory
pub/xbackgrounds.
w (Xwd format)
Xwd window dumper format is not supported yet.
Formats suitable for output only:
s (LN03)
DEC Sixel format.
h (HP LaserJet)
a HP LaserJet file. Options specify the resolution: one
of 75, 150, 300.
l (Imagen)
an imPRESS format file suitable for Imagen printers.
P (Printronix)
a file suitable for Printronix printers.
OPERATORS
+invert
Invert image.
+frame
Add a black frame to the image.
+swap
Swaps the bytes (word-wise).
AUTHOR
Juergen Wagner (gandalf@csli.stanford.edu), Center for the
Study of Language and Information (CSLI), Stanford Univer-
sity (formerly at Fraunhofer-Institut fuer Arbeitswirtschaft
und Organisation, Stuttgart).
------------------------------
Date: Wed, 20 Jul 88 09:04:03 EDT
From: andover!nexus!ane@swan.ulowell.edu (Anestis Panidis (617) 470-0555x324)
Subject: thinet-info?
First i would like to thank every one who responded to my request for
information and advise on the MR9000 multiport box, in helping with my
over extended thin-ethernet problem. It seems like a very good repeater
as i hear from everyone.
I still make another request, since i have to compare for management sake.
Q: Has anyone been using and has some input on an expandable repeater from
3Com , the MultiConnect Box, or API's ethernet repeater RL6000 ?
I thank everyone, and would be glad to forward the responses to any one
who is in a similar situation.
..ulowell!andover!ane ( uucp )
------------------------------
Date: Wed, 20 Jul 88 12:27:31 EDT
From: mike@ninja.cc.umich.edu (Michael Nowak)
Subject: Running screenblank on a Sun 3/60 with color display?
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?
Michael Nowak
University of Michigan Computing Center
mike@ninja.cc.umich.edu or ...!uunet!umix!ninja!mike
------------------------------
End of SUN-Spots Digest
***********************