GNU bug report logs - #4981
C-l during query-replace

Previous Next

Full log

Message #41 received at 4981 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Juri Linkov <juri <at> jurta.org>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: Dan Nicolaescu <dann <at> ics.uci.edu>, 4981 <at> debbugs.gnu.org
Subject: Re: bug#4981: C-l during query-replace
Date: Mon, 30 Nov 2009 18:07:45 +0200

>> Thanks for fixing this.  Are you sure that the new `recenter-positions'
>> is needed?  Given that there are 3 choices, it's easy to cycle through
>> them, so adding yet another defcustom that would be use by a very small
>> number of users does not seem justified (IMHO).
>
> I agree that it's overengineering.

I think what is overengineering is adding recenter-top-bottom
in the first place.  It imposes the arbitrary fixed cycling order
on users with no hope to customize such fundamental feature as
recentering.  `recenter-positions' mitigates this problem in the true
Emacs way as the *customizable* editor.

Please also note that even a minor feature `next-error' allows a similar
customization with `next-error-recenter'.  Perhaps we should try to merge
them, or at least provide a new option in `next-error-recenter' to use the
first value of `recenter-positions' as the primary position the user
prefers to put point after recentering.

> This patch is only acceptable if (to compensate) it unifies the two
> duplicate code paths of move-to-window-line-top-bottom and
> recenter-top-bottom.

Do you mean we should merge move-to-window-line-top-bottom and
recenter-top-bottom into one function?

-- 
Juri Linkov
http://www.jurta.org/emacs/

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.