robert delius royar writes:
Stephen, it is the case, then that you only see this with the
macports
libraries?
Yes. I don't see it with the Apple or macosforge binary distributions.
Are your macports libraries multiarchitecture (even if just 32/64 bit
on
the Intel OR PPC)?
No, I think they're built only for current architecture (i386 on the
Leopard system, PPC on the Tiger).
IIRC, there were some headers that were off version in at least one
of
the revisions. Not enough to cause this kind of differentiation but
perhaps emblematic of a cause.
No, this is not that kind of screwup. This is just bad code in X.org,
which ran into some weirdness that's probably specific to Mac OS X.
They didn't bother to check for corner cases (although they knew they
were possible, because they test for them in a different context in
the same function!)
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta