Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!zaphod.mps.ohio-state.edu!mips!apple!ames!sgi!shinobu!odin!hellfire.csd.sgi.com!ghelms
From: ghelms@hellfire.csd.sgi.com (Gretchen Helms)
Newsgroups: comp.sys.sgi
Subject: Re: SGI's X
Message-ID: <4060@odin.SGI.COM>
Date: 13 Feb 90 22:08:37 GMT
Sender: news@odin.SGI.COM
Reply-To: ghelms@hellfire.csd.sgi.com (Gretchen Helms)
Organization: Silicon Graphics, Inc., Mountain View, CA
Lines: 47

Jim Helman writes:
>The question that I would REALLY like the answer to (besides the
>obvious, when can I get a better release) is whether the SGI stuff on
>the X11R4 tape runs better or worse than the stuff shipped with IRIX
>3.2.  Any answers from SGI?  Any experience from the real world?

Clients and libraries and such should work fine.
I have been informed by one of our X engineers 
that the server from the X11R4 tape will *not*
build.

>BTW, don't bother trying to get anyplace with the hotline on this.
>I've been going around with them for over a year about X.  The best
>they can do is put you on the "hot list" for the next release.  The
>support "engineer" who answered my latest X call, after making a
>couple pointless digressions, focussed on THE REAL PROBLEM, asking
>whether the final "b" in "exit status: 0x8b" was capitalized or not.

Many times when a call comes in that is not a
request for information, the best policy for 
our customers is to have in hand a list that 
contains the sequence of events that caused 
the problem, the behaviour exhibited by the
machine as it crashes or bombs, and the specific
error codes produced at that time.  This allows
us to search through our database to locate 
the same error and cross-reference a previous
call to see if it could impact the current one.

Unfortunately, many of the error codes are 
somewhat obscure.  Frequently a missing digit
or the difference between uppercase and lowercase
can be crucial in pinpointing the problem.  For
example, the difference between "aborted with"
and "killed by" often lets us know whether a 
program caught an internal error or was terminated
before it had a chance to progress past a given
point.

For this reason, we ask for your patience when
we request more information on the error codes
or symptoms that are described.

G. "Murdock" Helms			A host is a host & from coast to coast
Silicon Graphics			No one will talk to a host that's close
Product Support Engineer		Unless the host (that isn't close)
ghelms@sgi.sgi.com			Is busy, hung or dead.       -D. Lesher