[12643] in Athena Bugs
[Theodore Ts'o: Re: Reqid: s15915 [Re: purify-solaris/5264]]
daemon@ATHENA.MIT.EDU (Theodore Ts'o)
Thu Sep 22 11:59:45 1994
Date: Thu, 22 Sep 1994 11:59:34 +0500
From: Theodore Ts'o <tytso@MIT.EDU>
To: bugs@MIT.EDU, alexp@MIT.EDU
The bug I reported to Purify has been resolved (they have an impressive
turnaround! :-).
It turns out to be a problem with the Cygnus ar and ranlib commands. If
you use the FSF ar and ranlib commands, the resulting library is not
compatible with Purify. You have to use the ar and ranlib in
/usr/ccs/bin. I've asked Purify to let me know what the FSF utilities
are doing "wrong", so we can forward this off to Cygnus.
- Ted
------- Forwarded Message
Date: Tue, 20 Sep 1994 17:37:11 +0500
From: Theodore Ts'o <tytso@MIT.EDU>
To: support@pure.com
Cc: pure-gnats@pure.com
In-Reply-To: Helen Bragg's message of Tue, 20 Sep 1994 14:22:42 -0700,
<199409202122.AA00781@pure>
Subject: Re: Reqid: s15915 [Re: purify-solaris/5264]
Address: 1 Amherst St., Cambridge, MA 02139
Phone: (617) 253-8091
Content-Length: 924
Date: Tue, 20 Sep 1994 14:22:42 -0700
From: bragg@pure.com (Helen Bragg)
Theodore,
Development looked at libkrb5.a and has this question:
How was the archive created? The format does not appear to comply with the
SVR4 spec. Did you use the normal Solaris ar utility, or did you perhaps
use a version of ar from GNU or some other source?
Thanks in advance for the info!
Ah, OK; that must be it. It just so happens that the first "ar" that's
in my path is from a GNU ar, from Cygnus Support's 1994 Q2 release. I
hadn't realized this was a problem, since SunSoft's cc and ld have no
problem dealing with its output.
If you can send me the nature of how the format does not comply with the
SVR4 release, I will be glad to pass that information on to Cygnus. (If
Purify could be made to be more tolerant of GNU ar produced archive
files, that would be quite nice as well.)
- Ted
------- End Forwarded Message