[mod.computers.sun] SUN-Spots Digest, v4n17

Sun-Spots-Request@RICE.EDU (Vicky Riffle) (06/12/86)

SUN-SPOTS DIGEST           Thursday, 12 June 1986	 Volume 4 : Issue 17

Today's Topics:
				Ports for SUN
			       Squealing noise
		   Suntools under 2.0 and SunView under 3.0
				RLL encoding?
			 Smalltalk-80 for Sun 3/50?
			      Sun-3 rlogin hangs?
		Interfacing franz with c graphics functions?
			      Backup strategies?
			       2.0 and 3.0 ld?
			 SUN manuals development?
			       CAD packages?
			   Incomplete NFS mount?
			    LCF international?
		       Concurrent Programming in SUN?
------------------------------------------------------------------------

Date: Tue, 3 Jun 86 12:25:49 PDT
From: keppel%pavepaws@BERKELEY.EDU (David Keppel)
Subject: ports for SUN

    I don't know prices or practicality, but here is at least one possible
    way to get more ports for your Sun:  Get a VME-bus to Multibus adapter,
    buy a Multibus multiple-port board, and plug the whole thing in to your
    system.  My impression is that the adapter board is cheap and that boards
    for the Multibus tend to be both cheaper and slower than VME boards.
    Consult your local guru.

    I have seen (been on) a Sun-1 with Sun-2 Multibus CPU board (*lots*
    slower than a Sun-3) and performance was acceptable to me, but certainly
    no great shakes.

-- 
----
    :-D avid  K eppel				    ..!ucbvax!pavepaws!keppel
		"Learning by Osmosis: Gospel in, Gospel out"

-----------------------------

Date: Tue, 3 Jun 86 12:19:40 PDT
From: keppel%pavepaws@BERKELEY.EDU (David Keppel)
subject: squealing noise

	I'm told that the squealing noise is made by the power supply
-- 
----
    :-D avid  K eppel				    ..!ucbvax!pavepaws!keppel
		"Learning by Osmosis: Gospel in, Gospel out"

-----------------------------

Date:     Mon,  9 Jun 86 11:37:57 CDT
From: William LeFebvre <phil@proserpina>
Subject:  Suntools under 2.0 and SunView under 3.0

From the SunView Programmer's Guide, page 3:

    "SunView is an extension and refinement of SunWindows 2.0,
     containing many enhancements, bug fixes and new facilities not
     present in SunWindows.  SunView is upward compatible with
     SunWindows---applications originally written under 2.0 can be
     recompiled and run under SunView."

Bull!  One of the SunView enhancements is the "panel" subwindow, which
is a replacement for option subwindows.  Unfortunately (and this is
even documented in the release notes) they decided to completely drop
option subwindows.  The include files aren't even around anymore.  So,
any unfortunate souls who wrote a tool that has an optionsw won't have
any luck compiling their programs under 3.0.  So much for upward
compatibility!

Who, besides me, got stuck in this trap?  Well, the VorTeX people,
among others.  The VorTeX dvi previewing tool will not even compile
under 3.0 because option subwindows are no longer supported.  Granted,
this is usually not that hard to fix, but it really annoys me that this
Sun publication contains such a blatant lie.  Don't get me wrong, I
much prefer panels to option subwindows, but c'mon Sun---how hard could
it have been to continue to support option subwindows so that old
programs would continue to compile?

			William LeFebvre
			Department of Computer Science
			Rice University
			<phil@Rice.edu>

-----------------------------

Date: Fri, 6 Jun 86 09:45:06 edt
From: Ken Mandelberg <km%emory.csnet@CSNET-RELAY.ARPA>
Subject: RLL encoding?

I have been reading articles about manufacturers of disk controllers
switching to RLL encoding to increase capacity and performance by
50%, without changing the drives. Adaptec is one of the manufacturers
mentioned.

Does anyone know if this relevant to the Adaptec SCSI<->ST506 controller
that Sun uses? 

Ken Mandelberg
Emory University
Dept of Math and CS
Atlanta, Ga 30322

{akgua,sb1,gatech,decvax}!emory!km   USENET
km@emory                      CSNET
km.emory@csnet-relay          ARPANET

-----------------------------

Date: Sat, 7 Jun 86 06:05:13 pdt
From: Herb Barad <barad%brand@usc-oberon.ARPA>
Subject: Smalltalk-80 for Sun 3/50?

I would like to know how to get a hold of a working version of
Smalltalk-80 for a Sun (I have a 3/50).  Has anybody had any
experience with it?  Thanks in advance.

-- 

Herb Barad	[USC - Signal and Image Processing Institute]

USENET:		...!sdcrdcf!usc-oberon!brand!barad			or
		...!mcvax!seismo!sdcsvax!sdcrdcf!usc-oberon!brand!barad

ARPANET:	barad%brand@USC-ECL.ARPA

USMail:		Univ. of Southern California
		Powell Hall 306, MC-0272
		Los Angeles, CA 90089-0272
		phone: (213) 743-0911

-----------------------------

Date: Mon, 9 Jun 86 10:40:28 edt
From: mike%bambi@mouton.bellcore.com
Subject: Sun-3 rlogin hangs?

Can someone say what the problem actually is?  Is it likely to show
up in general usage of TCP streams, or is it an rlogin-specific
difficulty?

If I had to guess, I'd guess they messed up the old URGENT data again...
Oh well, it wouldn't be Sun if it worked from release to release...

	- Mike

-----------------------------

Date: 09 Jun 86 10:18:02 PDT (Mon)
From: Dave McArthur <dave%mercury@rand-unix.ARPA>
Subject: Interfacing franz with c graphics functions?

We are interested in developing a simple windows package in Franz lisp for
our suns.  The package should allow users to construct windows and menus,
draw lines and bitmaps, and get mouse input, all in lisp. 

We are aware that external function calls (to c from lisp) are supported by
Franz lisp, but has anyone out there actually developed such a package,
using any of suncgi, sunwindows, suntools, suncore, etc., as a basis?  

Thanks.

dave

-----------------------------

Date: Mon, 9 Jun 86 17:37:49 EDT
From: Steve M. Burinsky <smb@mimsy.umd.edu>
Subject: backup strategies

It was bad enough when I had one host to worry about backups  on.
Now  I  have a network of 5 disked Sun's and two 785's to backup.
Has anybody come up with a good strategy for doing this?  I  want
backups  to  painless.   To me that's mounting one tape on a 6250
bpi drive before I go home every night.  If I have the luxury  of
extra  disk  space, how far do I dare go in backing up to disk on
my 785's before I dump these to tape?  I'd like  to  protect   my
users  from  themselves (by providing a history of disk snapshots
for when they delete their last years work and  don't  detect  it
for  a  month) as well as protect myself from a disk catastrophe.
I prefer the 1/2" drives on my 785 to the  (slow)  1/4"  cartidge
drives  on  the  Sun's.   Any suggestions based on how you handle
this problem?

Steve

-----------------------------

Date: Fri, 6 Jun 86 10:16:35 EDT
From: dbo%textset%umich.csnet@CSNET-RELAY.ARPA
Subject: 2.0 and 3.0 ld?

We have a problem.  We would like to be able to generate both 2.x 
binaries from a 3.0 system.  We are a third party software vendor
and we anticipate that some of our customers may be running 2.x for
a while.  We, ourselves, may be running 2.2 for a while because
of the delays Sun has had in shipping out upgrade kits
and because we have a 100u that we have to find a 4 megabyte memory 
board for. 

You may think that you can just ferret copies of the 2.x ld and libraries
somewhere and use them instead of the 3.0 stuff--wrong.  It doesn't work.
Sun tex support informs me that the problem is that ld dynamically figures
out what the page size is of the machine it is working on and the difference
in page sizes between the two systems is responsible for the problem.
The object file it produces under these circumstances looks like garbage
to me, so I am not sure exactly what they bought themselves by automatically
recognizing the different page size.  Not portability.

I asked Sun about a conversion program to translate 3.0 binaries into
2.0 format.  Their solution was to ask us to pay their consulting group to 
write a conversion program.  It seems to me that this is not the sort of 
thing that it is unreasonable to expect Sun to provide as a matter of course, 
since they are making yet another incompatible change.

So...  has anyone written a Sun 3.0 -> 2.0 object format conversion
routine?  If so, please send me some mail.  If not, let me know if you
are interested since it looks like I may be looking into what's involved.

Thanks in advance,
Doug

doug@umix.cc.umich.edu
doug@umich.csnet
ihnp4!umich!doug

-----------------------------

Date: Tue, 10 Jun 86 08:59:55 pdt
From: fluke!kelpie@uw-beaver.arpa
Subject: SUN manuals development

Can anyone provide information about how SUN goes about producing their
manual sets.  In particular, I am interested in who the writers were, what
automated tools were used, and the relative amount of effort it took to produce
them (cost, person-months).  Any piece of information would be helpful.
Also, pointers to those that may know would be helpful too.

I realize that this posting is not entirely in the flavor of 
mod.computers.sun, but it seemed like a good place to get a start on this 
subject.

Thanks in advance,

	Tony Garland, N7DX             	decvax\
	John Fluke Mfg Co.		ihnp4  >!uw-beaver\
	PO Box C9090				   allegra >!fluke!kelpie
	Everett, WA 98206 USA		ucbvax >!lbl-csam /
	(206) 356-5268			hplabs/

-----------------------------

Date: Mon, 9 Jun 86 13:50:26 PDT
From: sdcarl!tre@ucbvax.Berkeley.EDU (Tom Erbe)
Subject: CAD packages

	I have been using CADroid for about a year now and I am about ready
to try something new (The Droid Works gives almost no support).  Are there
any schematic entry packages out there that work well?  They must have good
customer support and keep up with the Sun software releases.

	Tom Erbe
	Computer Audio Research Lab, UCSD
	Q-037, La Jolla, CA  92093
	619-452-4384
	{ucbvax,ihnp4,akgua,hplabs,sdcsvax}!sdcarl!tre

-----------------------------

Date: Thu, 5 Jun 86 13:40:15 PDT
From: fluke!jeff@uw-beaver.arpa (Jeff Stearns)
Subject: Incomplete NFS mounts

I'm puzzled by the behavior of NFS mounts, and I'm wondering if anyone
can explain why things work the way they do (rather than the way I want...).

Say there are two hosts, A and B.  Each mounts a filesystem provided by the
other, as in:

    host A:	mount /dev/xy1c /usr.A ....
		mount  B:/usr.B /usr.B ....

    host B:	mount /dev/xy1c /usr.B ....
		mount  A:/usr.A /usr.A ....

Now everything works fine in the steady state once both machines are up and
running.  It's the boot procedure which bothers me.

There are clear opportunities for deadlock here, since neither host will
boot until its companion is up.  I have this impression that the "bg" option
should be applied here to background the mount so that booting can proceed.
Once each host is up, the backgrounded mount will complete and everything
should be cool.

BUT!  What about the time period while the mount is backgrounded and hasn't
yet completed?  Now you have this directory which will have a filesystem
attached to it *at some time in the future*.  In the meantime, it's a sitting
duck for filesystem activity, apparently TOTALLY UNAWARE that it is in peril
of being overlaid by a newly-mounted filesystem.

A program on host A may try to reference /usr.B/foo before the mount 
completes.  Or it may create it, only to have it covered up once the mount 
succeeds.

I'm bothered by the fact that the entire process is so nondeterministic.
What if host B is down for an hour?  Consider what happens when you generalize
this to a network comprising dozens of hosts, some of which boot in minutes,
others which may require hours of operator intervention.

The bottom line is this:  Shouldn't the directory be locked somehow while
the mount is in progress?  What can I (as a lowly system administrator) do
to bring this about?

    Jeff Stearns	John Fluke Mfg. Co, Inc.	(206) 356-5064
    {uw-beaver, decvax!microsoft, ucbvax!lbl-csam, allegra, sun}!fluke!jeff

-----------------------------

Date: Mon, 9 Jun 86 12:50:37 EDT
From: Martin Schoffstall <schoff%rpics.csnet@CSNET-RELAY.ARPA>
Subject: LCF international

Does this place really exist?  I noted in a sun-spots posting
of a couple of months ago, someone had bought LCF memory
for their sun2.  I haved called the number in "Catalyst"
about 10 times over the last couple of months and left messages
but to no avail...

marty schoffstall

-----------------------------

Date: Wed, 11 Jun 86 09:25:33 est
From: munnari!csadfa.cs.adfa.oz!gyp@seismo.CSS.GOV (Patrick Tang)
Subject: Concurrent Programming in SUN

Does anyone done any concurrent programming in C language under Unix 4.2 BSD.
Note that the C language is 'normal' C language NOT concurrent C.

I am trying to develop a graphics package under SUN3 or SUN2 to simulate 
real-time events.  Any advice to literature and packages written are most 
appreciated.

Thanks in advance.

-- 

Tang Guan Yaw/PatricK		 ISD:	+61 62 68 8170
Dept. Computer Science		 STD:	(062) 68 8170
University College	      ACSNET:	gyp@csadfa.oz
Uni. New South Wales		UUCP:	...!seismo!munnari!csadfa.oz!gyp    or
Aust. Defence Force Academy	...!{decvax,pesnta,vax135}!mulga!csadfa.oz!gyp
Canberra. ACT. 2600.		ARPA:	gyp%csadfa.oz@SEISMO.ARPA
AUSTRALIA		       CSNET:	gyp@csadfa.oz
Telex: ADFADM AA62030

-----------------------------

End of SUN-Spots Digest
***********************