[12544] in Athena Bugs
globalmessage on SunOS 4.1.3_U1 (if anyone at MIT cares)
daemon@ATHENA.MIT.EDU (yandros@MIT.EDU)
Thu Sep 8 04:16:14 1994
From: yandros@MIT.EDU
Date: Thu, 8 Sep 1994 04:15:48 +0500
To: haynes@cats.ucsc.edu (Jim Haynes), bugs@MIT.EDU
Cc: yandros@MIT.EDU, mhbraun@MIT.EDU, tom@MIT.EDU
Hello there...
I had a couple people try this out here (we're only running SunOS 4.x on a
very few machines here; public clusters are Solaris 2.3, currently) on
Sparc 5's running 4.1.3_U1, and it seemed to work fine. Perhaps the source
used differs somehow? I had thought that the GMS had remained basically
unchanged for a while (``since the dawn of time'' was my first thought. :-).
The `athena port' work is done by a member the subgroup of DCNS that uses
the SunOS machines (they're only used, right now, in certain server positions,
with a few client machines allocated for people to beat on), and he may have
made some changes to get around this problem somehow...
In any case, thanks for letting us know, and do tell us if you find the
source of the problem.
Thanks,
chad
>The get_message client quit working for us when we went from SunOS 4.1.3
>to SunOS 4.1.3_U1. The locus of the problem is the select() call
>in get_message_from_server.c right after the client has sent the GMS
>version number to the server. It times out. Right now I'm running
>identical static-linked binary on 4.1.3 and 4.1.3_U1; the former works,
>the latter does not. So we're looking into this with the Sun bugs data
>base and that sort of thing. I'm passing this along just in case
>anybody at MIT might want to run get_message on a pre-Solaris 2.x Sun
>and have the same trouble.
