GNU bug report logs - #69733
[PATCH] Flyspell (flyspell-word): do not force 'save-excursion' on timers

Previous Next

Full log

View this message in rfc822 format

From: Eli Zaretskii <eliz <at> gnu.org>
To: Daniel Pettersson <daniel <at> dpettersson.net>
Cc: 69733 <at> debbugs.gnu.org
Subject: bug#69733: [PATCH] Flyspell (flyspell-word): do not force 'save-excursion' on timers
Date: Tue, 12 Mar 2024 15:20:10 +0200

> From: Daniel Pettersson <daniel <at> dpettersson.net>
> Date: Mon, 11 Mar 2024 15:37:55 +0100
> 
> As I understand the manual there is nothing that states that one should
> assume that you cant move the point in buffer with timers.  This is of
> course impossible to guarantee as any code can be calling
> accept-process-output.  But it seams like an good idea to have code that
> is triggered by post-command-hook to not impose such conditions. 

Thanks, but I don't think I understand the relation between what you
say above (and in the comments in your patch below) and the code
change you propose:

>                    (while (progn
> -                           (accept-process-output ispell-process)
> +                           ;; don't force save-excursion to timers.
> +                           ;; only accept output from ispell-process.
> +                           (accept-process-output ispell-process nil nil t)

Where's save-excursion and point moving to which you allude here?

More importantly, why is it a good idea to stop running timers during
this accept-process-output call and ignore output from other
subprocesses?

Previous Next

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