Carl Worth <cworth@...> writes:

> 
> On Wed, 22 Jun 2011 08:50:23 +0200, Sebastien Binet
> <seb.binet@...> wrote:
> > On Tue, 21 Jun 2011 15:09:20 -0700, Carl Worth <cworth@...> wrote:
> > > Perhaps this is an emacs bug?
> > could be, but,
> > 
> > > I'm using Debian's emacs 23.3.1 and have not encountered the problem
> > > described.
> > I don't see anything related to 'point', 'point-invisible-p' nor
'forward-char' in there:
> > http://patch-tracker.debian.org/package/emacs23/23.3+1-1
> 
> Perhaps I was confused. I thought 23.3.1 was a distinct upstream release
> From 23.3, but maybe not, (emacs version numbers use 1-based indexing
> perhaps?).
> 
> > anyways... for the moment I'll just live with it...
> 
> Sorry about that.
> 
> > except if somebody has some insights on how to "catch" the 'End of
> > buffer' error and add some more protective programming around the
> > 'scroll-up' function ?
> 
> My emacs lisp skills are primarily restricted to actual
> experimentation. So if I can replicate the bug, I might be able to find
> a fix. I'm not very good at doing theoretical fixes with emacs lisp
> code.

I bisected the problem to commit 95ef8da29439f2e79115c36ab4d2a80aef1a1462 ("Fix
hiding a message while some citations are shown in notmuch-show view").
Reverting that fixes the problem for me. (If you don't want to resolve
conflicts, also revert 432e091924c1d1d8950a44ca78bc5b9c5ade47e4 first.) I don't
have the time just now to try to figure out the problem in that, but perhaps
this will help others go forward.

Jani