*BSD News Article 23920


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!haven.umd.edu!news.umbc.edu!eff!news.kei.com!bloom-beacon.mit.edu!ai-lab!life.ai.mit.edu!mycroft
From: mycroft@duality.gnu.ai.mit.edu (Charles Hannum)
Newsgroups: comp.os.386bsd.misc
Subject: Re: Status on discussed merge between NetBSD and FreeBSD
Date: 15 Nov 1993 01:23:36 GMT
Organization: MIT Artificial Intelligence Lab
Lines: 35
Message-ID: <MYCROFT.93Nov14202336@duality.gnu.ai.mit.edu>
References: <CGD.93Nov14085627@eden.cs.berkeley.edu> <CGHs3y.Au2@kithrup.com>
	<DERAADT.93Nov14110240@pain.agate> <CGIB9E.7z9@aib.com>
NNTP-Posting-Host: duality.ai.mit.edu
In-reply-to: dwex@aib.com's message of Mon, 15 Nov 1993 00:12:49 GMT


In article <CGIB9E.7z9@aib.com> dwex@aib.com (David E. Wexelblat)
writes:

   Because all the discussion on the XFree86 development list makes it
   pretty clear that libXt-based programs DON'T WORK.

Dave, PLEASE STOP STATING THAT.  IT IS WRONG.  I have personally run X
applications using a shared Xt, and they work just fine.  Admittedly,
the current linker is slightly hacked to make this work, but it does
in fact work.

   First compile with PIC, next link with '-G'.  That's it.  There are
   no .sa files with SVR4 shared libraries.

And if you had bothered to check your facts, you'd find that is
exactly what is being worked on.


This entire subthread is ridiculous.  The point of the original
statement regarding shared libraries was that the FreeBSD people tried
to pick them up without any real understanding of what was going on,
and had enormous troubles.  The hack they've done to ld.so to make it
work at all without random segmentation faults on startup (namely, the
stack pointer comparison) will very likely cause a binary
incompatibility if they ever want to enlarge the kernel's virtual
address space, unless they do further kluging in the kernel.

The fact that they do not think about such issues until it bites them
in the ass is one of the reasons I don't want them having unlimited
write access to the source I use.

Of course, there is always the chance they will read this and go fix
it now.