bug#1990: 23.0.60; Emacs trying too hard to find fonts that don't exist

[Kenichi Handa <handa@m17n.org>]

In article <f7ccd24b0901250357p6ac606beh67a4f2d4562e6230@mail.gmail.com>, Juanma Barranquero <lekktu@gmail.com> writes:

> On Sun, Jan 25, 2009 at 06:22, Jason Rumney <jasonr@gnu.org> wrote:
> > That looks much better though. Before, each "font for:" log entry was
> > followed by a dozen or so "list:" entries, which is what takes excessive
> > time on Windows.

> They still are. That was the output of M-occur on the *Help* buffer.

> For example, that's the result of

> emacs -Q
> C-h H
> M-: (setq font-log nil)
> M-x font-show-log

> font for: (1513 . hebrew-iso8859-8)
> list: -outline-DejaVu Sans Mono-*-mono-*-iso10646-1:script=hebrew
> list: -outline-DejaVu Sans Mono-*-iso10646-1:script=hebrew
> list: -*-DejaVu Sans Mono-*-mono-*-iso10646-1:script=hebrew
> list: -*-DejaVu Sans Mono-*-iso10646-1:script=hebrew
> list: -outline-*-mono-*-iso10646-1:script=hebrew
> list: -outline-*-iso10646-1:script=hebrew
> list: -*-mono-*-iso10646-1:script=hebrew
> list: -*-iso10646-1:script=hebrew
[...]

As there's no logs from the font-backend's `list' method,
the above listings actually did cache looking up only, and
thus they should not be that slow.  If you do:

M-: (setq font-log nil)

before C-h H, you'll see logs from the font-backend (labeled
"uniscribe-list:" or "w32font-list:" on Windows).

---
Kenichi Handa
handa@m17n.org