GNU bug report logs - #50042
26.3; Post-command hooks slow with M-x commands

Previous Next

Full log

View this message in rfc822 format

From: Lars Ingebrigtsen <larsi <at> gnus.org>
To: Simon Katz <sk <at> nomistech.com>
Cc: 50042 <at> debbugs.gnu.org
Subject: bug#50042: acknowledged by developer (control message for bug #50042)
Date: Fri, 20 Aug 2021 15:09:53 +0200

[Message part 1 (text/plain, inline)]

Simon Katz <sk <at> nomistech.com> writes:

>> The file for reproducing the problem was apparently mangled during
>> transmission, too.  Can you re-send the file as an attachment?
>
> Sure. Please find it attached.

Thanks.

When I hoisted the final waiting from the command, I totally didn't
notice that there was even more waiting there (in the case of a command
that had displayed a message).

I've now tweaked this further, and it now seems to work correctly in
all the message/no keybinding/keybinding cases, I think...

Can you test this, too, to see if I got all the cases right?  That is,
it shouldn't display anything if you do anything else before the
suggestion triggers, and it shouldn't inhibit entering other commands,
and...  er...  no other peculiarities.

One thing that is peculiar, but no more than before, is this bit:

                       ;; FIXME: Can be slow.  Cache it maybe?
                       (setq binding (execute-extended-command--shorter
                                      (symbol-name function) typed))))

It takes five seconds on my laptop, which means that these suggestion
arrive five seconds after I did the `M-x'.  But that hasn't changed
since before these patches.  

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

[demo-post-command-hook-slow-with-m-x-commands-v2.el (application/emacs-lisp, attachment)]

Previous Next

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