On Mon, Mar 26, 2007 at 12:49:47AM +0900, Stephen J. Turnbull wrote:
COMMIT 21.5
Per Olivier's request.
I'm curious: Why does this fix the problem? It's "obviously correct,"
but I would think this would be a lazy equivalent of an explicit copy.
Was the wrong thing getting copied, and then left in place because the
dirty flag wasn't set?
To be honest, I haven't checked. But I suspect you have it right, the
dirty flag wasn't being copied which was a bug in any case. I first
found out that unconditionally setting the dirty flag on copy killed
the bug. Given that copying the reference to the mirror table in some
cases looked extremely dubious in the first place, I decided to just
simplify it. The complexity came in part from using cct to copy the
mirror table itself, so I nixed that and the result was much nicer.
OG.
_______________________________________________
XEmacs-Patches mailing list
XEmacs-Patches(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-patches