Sun-Spots-Request@RICE.EDU (William LeFebvre) (04/03/88)
SUN-SPOTS DIGEST Friday, 1 April 1988 Volume 6 : Issue 41
Today's Topics:
Re: nfsd problems
Re: screendump from SUN 3/110 to laserprinter; greyscale
Re: Suns for general research computing
Re: troff or Postscript previewer
Re: colored buttons in sunview
SUN3 & 4 for time sharing system (or TSS)
Looking for robot simulators for Sun 3 or 4
laserwriter print limits
Problem moving display's mapped memory on Sun-3/160C
TCPDUMP for SunOS 4.0?
Does anyone have Calctool V2.1? Other calculators? (2)
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 stored on "titan.rice.edu". For volume X, issue Y,
"get sun-spots/vXnY". They are also accessible through the archive
server: mail the word "help" to "archive-server@rice.edu".
----------------------------------------------------------------------
Date: Sat, 19 Mar 88 16:06:07 EST
From: nesheim@think.com
Subject: Re: nfsd problems
Reference: v6n20
Steve D. Miller <steve@brillig.umd.edu>:
> We've seen similar behavior (strange hangings of nfsd instances in disk
> wait) here....we finally tracked the problem down to a very subtly
> corrupted inode on the disk. Fsck didn't see anything wrong with it, but
> there was something about it that caused nfsd to endless loop whenever it
> touched that inode.
I've seen this on a sun (3.4) when one of our users hard linked to a
directory. I havn't tried duplicating the problem, but zapping the
extraniously linked directory inode made the hung nfsd problem go away.
-- Bill Nesheim; Thinking Machines Corporation, Cambridge, MA +1 617-876-1111
nesheim@think.com, {mit-eddie,ihnp4}!think!nesheim
------------------------------
Date: Mon, 21 Mar 88 09:06:18 EST
From: Chuck Musciano <chuck@trantor.harris-atd.com>
Subject: Re: screendump from SUN 3/110 to laserprinter; greyscale
>> Are there laserprinters which can do greyscale?
>
>The Apple Laserwriter does greyscale. The is a Postscript command called
>`setgray' which takes a floating point number between zero and one, and
>interprets it in the obvious way. This is the SUN's standard laser
>printer, so I would imagine there is a way of producing screen dumps on
>it. Good luck.
Users should be aware that the LaserWriter is not a "gray" device, but is
monochrome device which emulates gray levels using halftoning techniques.
The setgray command in PostScript dictates which dither pattern will be
used in an effort to produce a pattern of dots which appear gray to the
human eye when viewed at some distance. For example, displaying a black
box at 50% gray does not result in every pixel being "turned on" 50%, it
means that every other pixel will be on and the others will be off. When
viewed at a distance greater than 18" or so (depending on your eyes and
visual acuity), it will appear "gray". Various dot patterns yield
different gray densities.
I believe that the LaserWriter really has only 20 builtin dither patterns,
meaning that, for example, gray levels 90% through 94% are all really the
same thing.
Finally, back to the original problem. Converting a true gray scale image
to something suitable for rendering by LaserWriter requires a tradeoff
between resolution and gray scale range. Each pixel in the original image
will require some larger number of pixels in the rendered image in order
to be displayed. For example, using a 3x3 pattern for each original pixel
allows 512 gray levels (since there 9 bits in the pattern). However, many
of these patterns are rotations and reflections of each other, and yield
the same gray density. In any case, the resulting image will be three
times bigger than the original, which may not be possible. An alternative
is to consider the original as a matrix of 3x3 patches, and replace each
patch with a pattern representing the average gray level of the patch.
This yields an equal size image with a corresponding loss of resolution.
The bottom line is that there is no real good answer, until true gray
scale and color printers become affordable.
Chuck Musciano
Advanced Technology Department
Harris Corporation
(305) 727-6131
ARPA: chuck@trantor.harris-atd.com
------------------------------
Date: Mon, 21 Mar 88 09:27:00 EST
From: bob@allosaur.cis.ohio-state.edu (Bob Sutterfield)
Subject: Re: Suns for general research computing
Reference: v6n32
bwiner@carleton.bitnet writes:
>We at Carleton University are investigating the potential for use of
>Suns in support of general research computing...
>
>(1) a variety of Sun hardware, including server(s) and diskless
>workstations.
With careful selection and configuration, it should all plug and play. Be
sure you get sales config specs that are up-to-date with the current
software versions. If you don't have anyone handy with networking and Sun
configuration expertise, hire someone fast, before unrealistic
expectations are established and the whole campus is grumbling about you.
>(2) provision of "timesharing" access for dumb terminals.
Get TCP/IP terminal servers from someone like Encore or Bridge or any of
several other vendors (reference the periodic round of questions on the
tcp-ip list). With some interesting tweaks to a nameserver, on can say
"telnet any-sun.carleton.edu" and be connected to the least-loaded
workstation in the namespace. With careful management of the namespace,
any Sun can look exactly like any other, so far as mail and news and all
that, so there's no need to maintain any concept of a home machine.
>(3) interconnections to Macs, PCs, non-Sun LANS.
For the Macs: Get Kinetics AppleTalk-Ethernet gateway boxes, and run
AppleShare and/or NCSA Telnet on the Macs, and the Columbia AppleTalk
library file and print server software on the Sun servers (note: Avoid
Centram's TOPS). You'll pay for the boxes, but all the above software is
free - you can get it via anonymous UUCP from our department.
For PCs: Run PC-NFS. For non-Sun networks, run TCP/IP software on those
machines, or run SNA or BITNET or DECnet on the Suns, or get any of
various protocol gateway boxes.
>(4) a mixture of user types, from naive to experienced. Some would
>want to have as little to do with Unix as possible.
Plop a Mac on their desk, don't tell them that the second disk icon on
their screen is really part of a double Eagle in a machine room across
campus somewhere, and they'll never know the difference. Similarly with a
PC and the E: drive (or whatever).
>(5) definite interest in statistical software: SPSS-X, BMDP, TSP, etc.
Get a copy of Catalyst, Sun's reference guide to other vendors' software
that runs on Suns. Of the above, I don't see TSP in the table of
contents, but it may be listed by vendor name.
Our department doesn't support the variety of users or specific
applications you describe, but we believe the general networking solutions
we're working out will extend into your environment pretty happily.
Bob Sutterfield, Department of Computer and Information Science
The Ohio State University; 2036 Neil Ave. Columbus OH USA 43210-1277
bob@cis.ohio-state.edu or ...!cbosgd!osu-cis!bob
------------------------------
Date: 21 Mar 88 20:20:53 GMT
From: elan!jlo@hplabs.hp.com (Jeff Lo)
Subject: Re: troff or Postscript previewer
Reference: v6n32
Elan Computer Group, Inc. sells troff based software including laser
printer drivers and screen previewers that run under SunView and X. Call
or write for details.
Jeff Lo
Elan Computer Group, Inc.
410 Cambridge Avenue, Suite A
Palo Alto, CA 94306
(415) 322-2450
..!{ames,hplabs}!elan!jlo
------------------------------
Date: 18 Mar 88 20:34:26 GMT
From: bdrc!jcl@mcnc.org (John C. Lusth)
Subject: Re: colored buttons in sunview
Reference: v6n30
Thanks Dick Keene (Sun-Spots Digest v6n30)! I've always been disappointed
with the look of SUNVIEW buttons on a color display, but I never figured
out how to do colored buttons before. What follows is a variation of
Dick's "color_button" routine. While "color_button" gives you a button
with a new foreground color, "color_bordered_image" yields a more
substantial looking button by filling the background within the outline of
a button with a second color. This routine can be used for all images
with borders, such as the dial image shown on page 155 of the SUNVIEW
Programmer's Guide.
/* example calls */
/* this creates a colored button */
panel_create_item(control, PANEL_BUTTON,
PANEL_LABEL_IMAGE,
color_bordered_image (
panel_button_image(control,"QUIT",6,Bold),
COLOR1, COLOR2),
PANEL_NOTIFY_PROC, quit,
0);
/* this creates a correctly colored dial */
panel_create_item(control, PANEL_CHOICE,
PANEL_CHOICE_STRINGS,
"0 ", "1 ", "2 ", "3 ", "4 ", 0,
PANEL_CHOICES_BOLD,
TRUE,
PANEL_MARK_IMAGES,
color_bordered_image (&dial0, YELLOW, BLACK),
color_bordered_image (&dial1, YELLOW, BLACK),
color_bordered_image (&dial2, YELLOW, BLACK),
color_bordered_image (&dial3, YELLOW, BLACK),
color_bordered_image (&dial4, YELLOW, BLACK),
0,
PANEL_NOMARK_IMAGES,
0,
PANEL_CHOICE_XS,
CX - 50, CX - 40, CX + 0, CX + 40, CX + 50, 0,
PANEL_CHOICE_YS,
CY - 0, CY - 40, CY - 50, CY - 40, CY - 0, 0,
PANEL_MARK_XS,
CX - 28, 0,
PANEL_MARK_YS,
CY - 28, 0,
PANEL_NOTIFY_PROC, speed_control,
0);
/* This is the routine that returns the colored image, with the background
within the border colored "bg", and all set pixels colored "fg" */
struct pixrect *
color_bordered_image (pr, bg, fg)
struct pixrect
*pr;
char
bg, fg;
{
int
x, y,
start, finish;
struct pixrect
*color_pr;
if(pr == NULL)
return(NULL);
/* make new pixrect */
color_pr = mem_create(pr->pr_size.x, pr->pr_size.y, 8);
/* copy with the foreground color added */
for (y = 0; y < pr->pr_size.y; ++y) { /* for each row ... */
/* give each pixel up to the left hand border a color of zero */
for (start = 0; start < pr->pr_size.x; ++start)
if (pr_get (pr, start, y) == 0)
pr_put (color_pr, start, y, 0);
else
break;
/* give each pixel down to the right hand border a color of zero */
for (finish = pr->pr_size.x - 1; finish > start; --finish)
if (pr_get (pr, finish, y) == 0)
pr_put (color_pr, finish, y, 0);
else
break;
/* give all clear pixels in between the new background color */
/* give all set pixels in between the new foreground color */
for (x = start; x <= finish; ++x)
if (pr_get (pr, x, y) == 0)
pr_put (color_pr, x, y, bg);
else
pr_put (color_pr, x, y, fg);
}
return(color_pr);
}
/* end of code
John C. Lusth
Becton Dickinson Research Center
Research Triangle Park, NC
...!mcnc!bdrc!jcl
*/
------------------------------
Date: Sat, 19 Mar 88 22:12:25 JST
From: Kusumoto Hiroyuki <kusumoto@etl.jp>
Subject: SUN3 & 4 for time sharing system (or TSS)
Reference: v6n25, v6n28
In v6n25:
>From: wucs1!wucs2!krf@uunet.uu.net (Kevin Fenster)
>Here's the question: How many users can we support on a Sun 4/280?
>....
> Is there anyone
>running 3/280's as timesharing machines (i.e. non-workstation)? I realize
>that a 4/280 has more cpu kick, but it uses the same I/O; a bottleneck?
In v6n28
>From: arnold@emory.UUCP (Arnold D. Robbins {EUCC})
>We are planning on replacing the two 780's with a single Sun 4. Is there
>anyone out there who has already done something like this?
We have used 3/280S as TSS machine and file server since April'87. And we
got two SUN4/280S for centeral system at last week.
Our 3/280S configuration was:
32MB main memory
32 serial line. (and 2 ports on CPU board.)
two 575MB disks. (recently, we have added 3rd and 4th 575MB disks.)
6250bpi 1/2inch MT.
Usually, 30 users (approx. 20 via serial line, 10 via rlogin) log in this
machine. there are 5-6 news reader using rn, 1-2 lisp users, many text
editing users, and a few latex users.
Further, this machine provides executable images of many PDS and freewares
(emacs,tex,latex,X window systems etc...) for about 30 SUN3s(diskfull) on
the LAN through the NFS.
Finally, we use it to serve /usr/spool/news.
We get resonable system response under this condition. (Of course, 3/60
with single user shows better performance.) The amount of main memory is
important. I think 24MB or more memory is needed for 30 users
[[ Not to mention lots of swap space! --wnl ]]
Last week, we got two SUN4/280S systems. It is very fast, comparable to
our lab's VAX8800. (note: VAX8800 is dual CPU machine. but one process is
not running under both of CPU simultaneously.)
SUN4 is two or three times as fast as SUN3/280. So, I think it can
support 30 or more users for educational and text processing purpose under
TSS configuration. (But it can not support 30 lisp users :-)
If you want to get more imformation, mail to me freely.
KUSUMOTO Hiroyuki (KUSUMOTO is my family name)
kusumoto%etl.jp@relay.cs.net
# ETL is the ElectroTechnical Laboratory (MITI's Lab.)
# MITI is the Ministor of International Trading and Industry
------------------------------
Date: 17 Mar 88 22:32:04 GMT
From: masticol@clash.rutgers.edu (Steve Masticola)
Subject: Looking for robot simulators for Sun 3 or 4
Hi,
The CS department at Rutgers is exploring the possibility of buying a
robot simulator for use on Sun 3 or Sun 4 workstations. This will be
primarily for student use in a graduate robotics class.
If you supply, use, or have knowledge of a commercially produced and
supported robot-motion simulator, please reply by email. Thanks for your
help!
- Steve Masticola
masticol@paul.rutgers.edu
------------------------------
Date: Mon, 21 Mar 88 08:57:42 PST
From: Stuart Cracraft <cracraft@hyper-sun7.jpl.nasa.gov>
Subject: laserwriter print limits
Has anyone managed to get the laserwriter (when connected to a SUN 3/110,
running SUN OS 3.4) to limit the length of its print jobs to an arbitrary
number of pages?
Stuart
------------------------------
Date: Mon, 21 Mar 88 09:32:51 -0500
From: Dan Grim <grim@udel.edu>
Subject: Problem moving display's mapped memory on Sun-3/160C
We have a Sun-3/160C to which we have added some memory boards. That
meant that we had to move the address for the color board. Now the PROM
monitor doesn't write on the display although suntools does. Any ideas?
------------------------------
Date: Sun, 20 Mar 88 1:59:58 EST
From: Doug Kingston (MS|mike) <dpk@brl.arpa>
Subject: TCPDUMP for SunOS 4.0?
Is there a version of Van Jacobson's tcpdump program available for
SunOS 4.0 (beta)?
-Doug-
aka <dpk@morgan.com>
------------------------------
Date: Mon, 21 Mar 88 17:44:31 EST
From: bernhold@qtp.ufl.edu
Subject: Does anyone have Calctool V2.1? Other calculators? (1)
A few weeks ago comp.sources.misc "published" the sources for Calctool
V2.1, however the copy we received here was messed up. I have had no luck
in trying the moderator or the archive sites I can reach to get a good
copy. Does anyone out there have the complete and correct sources? I
would appreciate hearing from you (via e-mail) in order to work out the
logistics of getting a copy.
In a more general light, can anyone recommend a good *scientific*
calculator program (not necessarily just for SUN's)? Something with good
precision (i.e. > 6-7 decimal digits), the major mathematical functions,
and probably some memory registers? Once again, please use e-mail. If I
get any response, I'll summarize.
Thanks muchly...
Dave Bernholdt
bernhold@qtp.ufl.edu
bernhold.ufpine.bitnet
[[ See my reply....next. --wnl ]]
------------------------------
Date: Fri, 1 Apr 88 18:02:56 CST
From: William LeFebvre <phil@Rice.edu>
Subject: Re: Does anyone have Calctool V2.1? Other calculators? (2)
Hans Boehm and Vernon Lee here at Rice put a calculator in the source
archives. I haven't really used it, so I don't know what features it has.
I do know that it uses "constructive real arithmetic" (One of Dr. Boehm's
research efforts) and will accurately produce the result to whatever
precision you request. Unfortunately, it's written in Russell. Not that
Russell is a bad language---it's just not very widely available.
I'll enclose some information from the distribution files. A binary
distribution is available in the archives under "sun-source" as
"calc.shar.01", "calc.shar.02", "calc.shar.03", and "calc.shar.04".
____________________
This is a desk calculator utility that operates on constructive real
numbers. The underlying idea is to represent a real number as a function
from a precisions specification to an approximation accurate to the
indicated presision. For efficiency reasons, these functions "remember"
the best known approximation to a number. More details on the ideas
underlying the implementation can be found in:
Hans-J. Boehm, "Constructive Real Interpretation of Numerical Programs",
Proceedings of the SIGPLAN '87 Symposium on Interepreters and Interpretive
Techniques, SIGPLAN Notices 22, 7 (July 1987), pp. 241-221.
Hans-J. Boehm, R. Cartwright, Michael J. O'Donnell, and Mark Riggle,
"Exact Real Arithmetic: A Case Study in Higher Order Programming",
Proceedings of the 1986 Lisp and Functional Programming Conference,
pp. 162-173.
The underlying constructive real package, as well as the desk calculator
interface to it, are written in the programming language Russell. A
Russell compiler along with source for the constructive reals package is
available either over the net (ftp required) or on tape for a distribution
fee. If you are interested, please send mail to boehm@rice.edu.
Authors: Hans-J. Boehm & Vernon Lee
boehm@rice.edu scorpion@rice.edu
__________
William LeFebvre
Department of Computer Science
Rice University
<phil@Rice.edu>
------------------------------
End of SUN-Spots Digest
***********************