Path: utzoo!attcan!uunet!snorkelwacker!bloom-beacon!EXPO.LCS.MIT.EDU!keith
From: keith@EXPO.LCS.MIT.EDU (Keith Packard)
Newsgroups: comp.windows.x
Subject: Re: XDMCP questions
Message-ID: <9003100541.AA21850@expo.lcs.mit.edu>
Date: 9 Mar 90 23:54:10 GMT
References: <7663@b11.ingr.com>
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 26



> Why is the call to WaitForSomething ifdef'ed out for SYSV in the R4 xdm?
> Is their some fundamental problem with supporting XDMCP on System 5 that
> I should be aware of?

Not really; I didn't spend the time to port it.  SYSV isn't exactly a fun
debugging environment.  I don't envision any real problems except those
with SYSV signals; xdm does some crufty things which may get botched with
unreliable signal handling.  Perhaps someone with experience using SYSV
reliable signals could perform the necessary changes.

> There appears to be no code to perform byte-swapping for XDMCP in the server
> or xdm.

The XDMCP implementation provided with R4 is byte-order independent; it
will work correctly on both big-endian and little-endian machines

> Is there any reason why the execute() function was not used in place of
> execve() in xdm/session.c?

No there isn't -- in fact the execute function was written to be used in
exactly those places, somehow it was never set up correctly (whoops!).

Keith Packard
MIT X Consortium