[SPARCbook] Netscape 4.73 under Solaris 2.6 final results (sortof...) --FINAL--

[Ken Hansen]

Altavista.com...

I get four hits for the patch 105284-31 - one that is pointless (server
stats page),
one that points to a patch library in the UK (which has replaced -31
with -34 as
a service), a company with the patch listed as compatible with their
hardware and
one that leads to a dead link.

I did insist that my results be in English, but that should have picked up a
site in
VA? Shouldn't it?

Until you rnote I never heard of topclick.com...

Ken
(I must stay in more - I'm spending too much time outside these days... ;^)

----- Original Message -----
From: "David Hansen" <dhansen at salug.org>
To: <sparcbook at sunhelp.org>
Sent: Thursday, June 15, 2000 3:03 PM
Subject: [SPARCbook] Netscape 4.73 under Solaris 2.6 final results
(sortof...) --FINAL--


>
>
> On Thu, 15 Jun 2000 sparcbook-request at sunhelp.org wrote:
> > Thanks for the pointer - when I tried "conventional" search engines,
every
> > link I found pointed to a "dynamic" patch library, where older revs are
> > replaced!
>
>
>  Not sure what you claim to be "conventional" but upon my first attempt at
> a "conventional" web search engine for the string "105284-31", the first
> return is a match.
> www.topclick.com #"105284-31" for a search field.
>
>
http://solaris.license.virginia.edu/os_product_patches/patches/5.6/105284-31
/
> #this is the very first match
>
> >
> >
> > As for the idea that this may be a problem with the Weitek 9100
framebuffer,
> > I don't think so, but I am willing to be found wrong - it is just a
feeling.
> > The driver
> > for the Weitek came from Tadpole, and they were selling it to clients
that
> > wanted
> > their machines to work "right." The patch from Sun has a history of
breaking
> > and
> > then fixing things (that's how you get up to Rev 34!). If it was Rev 3
or 4
> > that
> > caused the problem, and the patch didn't have a histroy of failures, I
be
> > more
> > open minded about this I suppose. Besides, it is not that scrolling
didn't
> > work,
> > it was that using the arrow keys that failed - PageUp and PageDown,
scroll
> > arrows all worked fine.
>
>
>  You don't seem very openminded. You have been going on about a
> function within a precompiled program not working when you not only change
> the operating system it was compiled against, but you also change the
> libraries for its GUI environment (i.e.- libXm.so.3) as well as changing
> the version of the GUI environment.
>
>  I'm assuming that you are the one who installed Netscape Communicator
> v4.73 onto your machine. This application has the following filename...
> communicator-v473-us.sparc-sun-solaris2.5.1.tar.gz. This file is also,
> precompiled (you don't have the source, you didn't compile it). This means
> that it was compiled under Solaris 2.5.1 and Motif 1.2.3 (you've chnaged
> quite a few dependencies here). (Also, this clarification was made in
> response to an earlier email which asked if anyone knew which
> OS version this program was compiled against.)
>
>  If you cared to read the history of the 105284 patch (or any patch), then
> you would be able to realize that it did not take 34 times to get the
> patch right. But rather, these are mainly responses to bug reports related
> to the Motif environment as well as a correction for one bad patch and a
> handful of tweaks for prior patches that had side effects along the range
> of memory leaks and resizing problems. A far cry from your slanders that
> Sun first breaks the libraries and then fixes them in some form of
> incredibly insane loop.
>
>  Let's think about this, 3 main possibilities of regions to blame:
> Tadpole, Sun, Netscape. (according to you)
>
> 1: Could it be Tadpole? Doubtful, this behaviour has been seen
> outside of the sparcbook area. (try a simple newsgroup
> search if you do not have a sparc handy to field-test)
> 2: Could it be Netscape? Doubtful, Netscape supplies a precompiled
> version of their platform for Solaris Sparc 2.5.1 and it
> works just fine on that platform.
> 3: Could it be Sun? According to you it must be. Because if you
> muck around with the OS,X,Motif, versions and all of the
> relevant libraries then Netscape breaks...so that must be
> Sun's fault. Nice logic.
>
>  Let's add some to this. Solaris 2.6, in itself probably isn't at fault
> based on reports of this behaviour not being present in earlier revisions
> of the 105284 patch. That leaves X, Motif and the patchset under
> suspicion. I'll save you some time and tell you that X isn't at fault in
> this case. So let's go actually _read_ those pesky little files that come
> with patches. (namely pkginfo)
>
> 105284-31 pkginfo contents:
> PKG=SUNWmfrun
> SUNW_PKGTYPE=usr
> NAME=Motif RunTime Kit
> PRODNAME=Motif RunTime Kit
> VERSION=1.2.7,REV=10.97.07.02
> PRODVERS=1.2.7
> CATEGORY=system
> DESC=Motif 1.2.3 libraries, headers, xmbind and bindings
> BASEDIR=/usr
> VENDOR=SunSoft, Inc.
> HOTLINE=Please contact your local service provider
> EMAIL=
> MAXINST=1000
> CLASSES=none
> RESPONSE=Y
> MOTIFRUNTIME=0
> PSTAMP=pasta970702140402
> ARCH=sparc
> SUNW_PATCHID=105284-31
>
> and the pkginfo contents from 105284-34:
> PKG=SUNWmfrun
> SUNW_PKGTYPE=usr
> NAME=Motif RunTime Kit
> PRODNAME=Motif RunTime Kit
> VERSION=1.2.7,REV=10.97.07.02
> PRODVERS=1.2.7
> CATEGORY=system
> DESC=Motif 1.2.5 libraries, headers, xmbind and bindings
> BASEDIR=/usr
> VENDOR=SunSoft, Inc.
> HOTLINE=Please contact your local service provider
> EMAIL=
> MAXINST=1000
> CLASSES=none
> RESPONSE=Y
> MOTIFRUNTIME=0
> PSTAMP=pasta970716173826
> ARCH=sparc
> SUNW_PATCHID=105284-34
>
>
> Hmm, look at that... There seem to have been new versions of the libraries
> added in this patch for The OpenGroup's Motif. 105284-31 put v1.2.3 of the
> Motif libraries etc... whereas 105284-34 was using v1.2.5. This is most
> likely your culprit. Just for kicks let's see if 105284-34 was the patch
> to introduce the newer version of the libraries for the Motif Runtime Kit
> v1.2.7.
>
> pkginfo contents for 105284-33:
> PKG=SUNWmfrun
> SUNW_PKGTYPE=usr
> NAME=Motif RunTime Kit
> PRODNAME=Motif RunTime Kit
> VERSION=1.2.7,REV=10.97.07.02
> PRODVERS=1.2.7
> CATEGORY=system
> DESC=Motif 1.2.5 libraries, headers, xmbind and bindings
> BASEDIR=/usr
> VENDOR=SunSoft, Inc.
> HOTLINE=Please contact your local service provider
> EMAIL=
> MAXINST=1000
> CLASSES=none
> RESPONSE=Y
> MOTIFRUNTIME=0
> PSTAMP=pasta970716173826
> ARCH=sparc
> SUNW_PATCHID=105284-33
>
>
>  Welp, looks like 105284-33 was the first introduction of the v1.2.5
> libraries for the v1.2.7 Motif Runtime Kit.
>
>  But you want to blame someone for breaking your arrow keys while using
> Netscape Communicator. I would suggest blaming yourself for using a
> precompiled application in an environment that it wasn't built in and
> against libraries that differ from those that the application was
> dynamically linked against and therefore has dependencies for.
>
>  I don't think it would be fair to blame The OpenGroup for updating their
> programs, Sun for updating theirs _and_ for providing patches to help out
> with 3rd party applications that Sun didn't write, nor Netscape for not
> providing you with a precompile of every version of their browser
> under every possible version of every possible combination of every
> possible environment.
>
>  But if you want to know who to go to for a fix then I would either
> suggest that you downgrade your Motif libraries to v1.2.3 (ala 105284-31)
> and be mindful of what other programs you may break in the process of
> doing this. Or go to Netscape and ask for a precompile for your new
> environment. Or grab Mozilla and compile it yourself, or switch to another
> browser that does work with your environment.
>
>
>  >
> > I'll try the back-rev patch and report back, but things look OK without
the
> > patch
>
>  I would suggest that you try doing some research first, before you start
> changing your system environment around without understanding what you are
> even looking for. One day it may not result in a harmless change.
>
>
> > Thanks for all your help,
>
>
>   you are welcome.
>
>  -david
>
> p.s.- it took _much_ longer to compose this email than the five minutes of
> research that it took to solve your problem that has persisted for 2
> weeks.
>
>
> _______________________________________________
> Sparcbook mailing list  -  Sparcbook at sunhelp.org
> http://www.sunhelp.org/mailman/listinfo/sparcbook