Re: scroll-conservatively overflow

[no-spam@cua.dk (Kim F. Storm)]

Romain Francoise <romain@orebokech.com> writes:

> Romain Francoise <romain@orebokech.com> writes:
> 
> > About scrolling: I've been bitten by a bug for some weeks now, I
> > originally thought it was specific to the multi-tty code but I
> > discovered this morning that it's not.  (A fresh checkout from CVS HEAD
> > this morning exhibited the same issues.)
> 
> I have finally found enough time to track it down.  The change
> responsible for all my scrolling problems is Richard's change to
> indent.c on March 2nd:
> 
> 2004-03-02  Richard M. Stallman  <rms@gnu.org>
> 
> 	* indent.c (compute_motion): Save vpos in prev_vpos, like hpos etc.
> 
> Reverting this change (see attached patch) solves all issues in Dired,
> Gnus, etc.
> 
> I have tried to come up with a fix but failed.  I don't really
> understand why this change was necessary, it adds complexity and does
> not fix anything as far as I can see.  If it was just a code cleanup,
> maybe the best would be to revert it for the time being.

The change was made in response to the following bug report:
The fix was obviously not correct.


Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:

Place the following in a file named `/tmp/x'
======cut here=====
Barf

        Whee!

Yadda
======cut here=====

(the above contains three lines, no tabs, and no trailing spaces, for
a total of 27 characters)

Now do the following:

* start `emacs -q --no-site-file'

* find the file: C-x C-f /tmp/x RET

* turn on selective-display: C-u C-x $

* attempt to move point forward: C-n C-n

Note that the cursor did not appear to move after the second C-n.
That's the bug; it should have appeared on the third visible line
(right before the `Y').

* sanity check: C-x =

You should see point 6 of 27.

* again attempt to move point forward: C-n

* again sanity check: C-x =

It's still at 6 of 27.

If it matters, you can "un-stick" point by typing C-f, *then* typing
C-n; in that case, point indeed winds up on the third visible line,
where I'd expect it to.


-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk