GNU bug report logs - #24634
25.1 on Windows (native and cygwin+w32): high CPU load with "wavy" characters

Previous Next

Package: emacs;

Reported by: Georges Ko <gko <at> gko.net>

Date: Fri, 7 Oct 2016 05:01:02 UTC

Severity: normal

Found in version 25.1

Done: Eli Zaretskii <eliz <at> gnu.org>

Bug is archived. No further changes may be made.

Full log


View this message in rfc822 format

From: Georges Ko <gko <at> gko.net>
To: 24634 <at> debbugs.gnu.org
Subject: bug#24634: 25.1 on Windows (native and cygwin+w32): high CPU load with "wavy" characters
Date: Sat, 08 Oct 2016 10:59:06 +0800
Eli Zaretskii <eliz <at> gnu.org> writes:

>> > What font(s) is/are used on your system to display these characters?
>> 
>> I did some tests with "runemacs -Q", with the following results as an
>> org-mode buffer.
>> 
>> I can mitigate the problem by using language environment "English" and
>> font "MingLiu-SIZE" (2nd step), the worst case being language
>> environment "UTF-8" and font "Courier NEw-SIZE" (3rd step).
>
> So Symbola and Courier New don't cause slowdown, while MingLiu does?
>
> Do you have some custom fontset setup, or does Emacs use MingLiu for
> the WAVY LOW LINE by default, in "emacs -Q"?

Emacs uses MingLiu in "emacs -Q" on my system.

In "Initial setup" I just do "emacs -Q" and insert these characters and
see the fonts used to display ASCII characters (1st line), WAVY LINE
(2nd line) and WAVY LOW LINE (3rd character), which are:

  - ASCII         -O-Courier New-NNN-mono-13-*-*-*-c-*-iso8859-1
  - WAVY LINE     -O-Symbola-NNN-serif-13-*-*-*-p-*-iso8859-1
  - WAVY LOW LINE -O-MingLiU-NNN-mono-13-*-*-*-c-*-iso10646-1

and WAVY LINE     (Symbola) is OK (no slowdown) and
    WAVY LOW LINE (MingLiu) is NOK (slowdown).

After switching to "MingLiu-10" with (set-frame-font "MingLiu-10),
everything is smooth and using all use MingLiu-10 (2nd row of table):

  - ASCII, WAVY LINE -O-MingLiU-NNN-mono-13-*-*-*-c-*-iso8859-1
  - WAVY LOW LINE    -O-MingLiU-NNN-mono-13-*-*-*-c-*-gb2312.1980-0

> (But the way, setting language environment to UTF-8 on Windows is not
> a good idea, it will cause you trouble unless you make further
> customizations to countermand that.)

OK, let's narrow the scope and not consider UTF-8 cases (rows 3 and 4).

Georges
-- 
 Georges Ko                     gko <at> gko.net                      2016-10-08





This bug report was last modified 8 years and 286 days ago.

Previous Next


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