bug#3319: 23.0.93; Default font

bug#3319: 23.0.93; Default font

[Kenichi Handa]

In article <001801c9d71a$8bade240$a309a6c0$@J.Wright@qmul.ac.uk>, "Francis Wright" <F.J.Wright@qmul.ac.uk> writes:

> Please describe exactly what actions triggered the bug
> and the precise symptoms of the bug:

> When I start Emacs 23.0.93 normally, it starts up with the default
> font set to Sanvito Pro Subhead (regular, 10 point), which is not what
> I expected and is almost unreadable in this context. This is the first
> time I have run Emacs 23 on this computer; previously I was running
> Emacs 22.3. When I start Emacs 23.0.93 with the -Q option, it starts
> with Courier New as default font, which is fine (as a default). I
> think the problem is that I had this in my .emacs file:

> (custom-set-faces
>  '(default ((t (:stipple nil :background "white" :foreground "black"
> :inverse-video nil :box nil :strike-through nil :overline nil :underline nil
> :slant normal :weight normal :height 98 :width normal :family
> "outline-lucida sans typewriter"))))
>  '(woman-italic-face ((t (:foreground "red" :slant italic)))))

I think what you intend by :family "outline-lucida sans
typewriter" is actually :foundry "outline" :family "lucida
sans typewriter", right.  If so, I've just installed a fix.
Please try again.

---
Kenichi Handa
handa@m17n.org

bug#3319: 23.0.93; Default font

[Jason Rumney]

Kenichi Handa wrote:
> I think what you intend by :family "outline-lucida sans
> typewriter" is actually :foundry "outline" :family "lucida
> sans typewriter", right.  If so, I've just installed a fix.
> Please try again.
>   

Please ensure your fix does not break cases like :family "MingLiU-ExtB", 
where the hyphen really is part of the family name.

bug#3319: 23.0.93; Default font

[Kenichi Handa]

In article <4A1227C1.2010503@gnu.org>, Jason Rumney <jasonr@gnu.org> writes:

> Kenichi Handa wrote:
> > I think what you intend by :family "outline-lucida sans
> > typewriter" is actually :foundry "outline" :family "lucida
> > sans typewriter", right.  If so, I've just installed a fix.
> > Please try again.

> Please ensure your fix does not break cases like :family "MingLiU-ExtB", 
> where the hyphen really is part of the family name.

We can't handle that kind of family name without breaking a
backward compatibility.  We discussed this matter a while
ago, but didn't reach a conclusion.  I'll put the last mail
on that thread at the tail.

Anyway, my change is just this.

-	  (setq foundry (match-string 2 family)))
-	(setq family (match-string 1 family)))
+	  (setq foundry (match-string 1 family)))
+	(setq family (match-string 2 family)))

and it is a fix of a bug introduced by this change.

2008-10-14  Chong Yidong  <cyd@stupidchicken.com>

	* faces.el (set-face-attribute): Set family and foundry before
	other attributes.

---
Kenichi Handa
handa@m17n.org

------------------------------------------------------------
From: Kenichi Handa <handa@m17n.org>
To: Miles Bader <miles@gnu.org>
In-reply-to: <buo7icb1fnp.fsf@dhapc248.dev.necel.com> (message from Miles
	Bader on Fri, 27 Jun 2008 16:49:14 +0900)
References: <87hcbi9vbg.fsf@catnip.gol.com>
	<je1w2loom0.fsf@sykes.suse.de>	<buohcbh6es9.fsf@dhapc248.dev.necel.com>
	<4862198F.4060305@gnu.org>	<E1KBTIl-0006en-Nc@etlken.m17n.org>	<jwviqvxk4bq.fsf-monnier+emacs@gnu.org>	<E1KC8Gs-0006kZ-PQ@etlken.m17n.org>
	<buo7icb1fnp.fsf@dhapc248.dev.necel.com>
Content-Type: text/plain; charset=US-ASCII
Message-Id: <E1KC94K-0007dY-At@etlken.m17n.org>
Date: Fri, 27 Jun 2008 17:12:32 +0900
Cc: schwab@suse.de, jasonr@gnu.org, monnier@iro.umontreal.ca,
        emacs-devel@gnu.org
Subject: Re: sans-serif name change

In article <buo7icb1fnp.fsf@dhapc248.dev.necel.com>, Miles Bader <miles.bader@necel.com> writes:

> Kenichi Handa <handa@m17n.org> writes:
> > And, fontconfig requires `-' in a family name to be escaped
> > by `\':
..
> > So, how about using the same rule for Emacs?
> >
> > (set-face-attribute 'variable-pitch nil :family "sans\\-serif")

> But currently other places, such as `set-frame-font' seem to work
> properly if passed "Sans-Serif" without escaping the "-".  Perhaps some
> part of emacs is already adding the backslash?

No.  In ftfont.c, these names:

  "mono", "monospace", "sans-serif", "sans serif", "sans", "serif"

are specially recognized as generic family names.  In
addition, fontconfig accepts the literal "sans-serif" when
used in the function FcPatternAddString (pat, FC_FAMILY, ...).
Fontconfig requires escaping only on parsing a font name,
and adds escaping on generating a font name.

> I think _if_ there are fonts with "-" in their name, requiring the dash
> to be escaped will annoy and confuse users, and lead to bug reports.
> I don't know how many such fonts there are, of course.

> [I'd say just get rid of the old-style foundry-family names; do we
> really care?]

As I wrote before, I tend to agree with breaking the
backward compatibility here.

---
Kenichi Handa
handa@ni.aist.go.jp

bug#3319: 23.0.93; Default font

[Francis Wright]

> From: Kenichi Handa [mailto:handa@m17n.org]
> Sent: Tuesday 19 May 2009 6:42 am
> To: Jason Rumney
> Cc: 3319@emacsbugs.donarmstrong.com; F.J.Wright@qmul.ac.uk
> Subject: Re: bug#3319: 23.0.93; Default font
> 
> In article <4A1227C1.2010503@gnu.org>, Jason Rumney <jasonr@gnu.org>
> writes:
> 
> > Kenichi Handa wrote:
> > > I think what you intend by :family "outline-lucida sans
> > > typewriter" is actually :foundry "outline" :family "lucida
> > > sans typewriter", right.  If so, I've just installed a fix.
> > > Please try again.
> 
> > Please ensure your fix does not break cases like :family "MingLiU-
> ExtB",
> > where the hyphen really is part of the family name.
> 
> We can't handle that kind of family name without breaking a
> backward compatibility.  We discussed this matter a while
> ago, but didn't reach a conclusion.  I'll put the last mail
> on that thread at the tail.
> 
> Anyway, my change is just this.
> 
> -	  (setq foundry (match-string 2 family)))
> -	(setq family (match-string 1 family)))
> +	  (setq foundry (match-string 1 family)))
> +	(setq family (match-string 2 family)))
> 
> and it is a fix of a bug introduced by this change.
> 
> 2008-10-14  Chong Yidong  <cyd@stupidchicken.com>
> 
> 	* faces.el (set-face-attribute): Set family and foundry before
> 	other attributes.

I just edited the above change into faces.el and recompiled it, but it
doesn't seem to have made any difference. I'll recheck tomorrow.

Francis

bug#3319: 23.0.93; Default font

[Kenichi Handa]

In article <006601c9d8a5$563b2550$02b16ff0$@j.wright@qmul.ac.uk>, "Francis Wright" <f.j.wright@qmul.ac.uk> writes:

> > Anyway, my change is just this.
> > 
> > -	  (setq foundry (match-string 2 family)))
> > -	(setq family (match-string 1 family)))
> > +	  (setq foundry (match-string 1 family)))
> > +	(setq family (match-string 2 family)))
> > 
> > and it is a fix of a bug introduced by this change.
> > 
> > 2008-10-14  Chong Yidong  <cyd@stupidchicken.com>
> > 
> > 	* faces.el (set-face-attribute): Set family and foundry before
> > 	other attributes.

> I just edited the above change into faces.el and recompiled it, but it
> doesn't seem to have made any difference. I'll recheck tomorrow.

Did you re-build emacs itself?  Just re-bytecompiling
faces.el is not enough.

---
Kenichi Handa
handa@m17n.org

bug#3319: 23.0.93; Default font

[Francis Wright]

> From: Kenichi Handa [mailto:handa@m17n.org]
> Sent: Wednesday 20 May 2009 1:40 am
> To: Francis Wright
> Cc: jasonr@gnu.org; 3319@emacsbugs.donarmstrong.com
> Subject: Re: bug#3319: 23.0.93; Default font
> 
> In article <006601c9d8a5$563b2550$02b16ff0$@j.wright@qmul.ac.uk>,
> "Francis Wright" <f.j.wright@qmul.ac.uk> writes:
> 
> > > Anyway, my change is just this.
> > >
> > > -	  (setq foundry (match-string 2 family)))
> > > -	(setq family (match-string 1 family)))
> > > +	  (setq foundry (match-string 1 family)))
> > > +	(setq family (match-string 2 family)))
> > >
> > > and it is a fix of a bug introduced by this change.
> > >
> > > 2008-10-14  Chong Yidong  <cyd@stupidchicken.com>
> > >
> > > 	* faces.el (set-face-attribute): Set family and foundry before
> > > 	other attributes.
> 
> > I just edited the above change into faces.el and recompiled it, but
> it
> > doesn't seem to have made any difference. I'll recheck tomorrow.
> 
> Did you re-build emacs itself?  Just re-bytecompiling
> faces.el is not enough.

No, I didn't rebuild. I realised after I sent my previous email that more
than just recompiling might be necessary. I might not be able to test this
properly until Emacs 23.0.94 becomes available since I'm running Jason's
precompiled binaries.

bug#3319: 23.0.93; Default font

[Stefan Monnier]

>> I think what you intend by :family "outline-lucida sans
>> typewriter" is actually :foundry "outline" :family "lucida
>> sans typewriter", right.  If so, I've just installed a fix.
>> Please try again.
> Please ensure your fix does not break cases like :family "MingLiU-ExtB",
> where the hyphen really is part of the family name.

Breaking backward compatibility might be OK, but I still wonder: how
come we ended up with "Sanvito Pro Subhead" from a spec of
a non-existing "outline-lucida sans typewriter"?


        Stefan

bug#3319: 23.0.93; Default font

[Francis Wright]

> From: Stefan Monnier [mailto:monnier@iro.umontreal.ca]
> Sent: Tuesday 19 May 2009 3:33 pm
> To: Jason Rumney
> Cc: 3319@emacsbugs.donarmstrong.com; Kenichi Handa; Francis Wright
> Subject: Re: bug#3319: 23.0.93; Default font
> 
> >> I think what you intend by :family "outline-lucida sans
> >> typewriter" is actually :foundry "outline" :family "lucida
> >> sans typewriter", right.  If so, I've just installed a fix.
> >> Please try again.
> > Please ensure your fix does not break cases like :family "MingLiU-
> ExtB",
> > where the hyphen really is part of the family name.
> 
> Breaking backward compatibility might be OK, but I still wonder: how
> come we ended up with "Sanvito Pro Subhead" from a spec of
> a non-existing "outline-lucida sans typewriter"?

Here's a bit more background. The following is without Kenichi Handa's fix.
I'll try to install the fix tomorrow and see what difference it makes.

I think I selected Lucida Sans Typewriter from a font menu and an earlier
version of Emacs then added the outline prefix. I currently have Calibri set
as my default font, but the Emacs 22.3 customize buffer displays the font as
Font Family: outline-calibri. If I resave this customization using Emacs
22.3, it is saved in my .emacs file as :family "outline-calibri". So I think
that's where the "outline-" prefix came from. Now, when I start Emacs
23.0.93, it has its default font set to Adobe Caslon Pro. If I set the
default font back to Lucida Sans Typewriter in Emacs 22.3, Emacs 23.0.93
still uses Adobe Caslon Pro as the default font.

To be precise, I set the default font in Emacs 22.3 by using the Options
menu item Set Font/Fontset... and then selecting Lucida Sans Typewriter from
the resulting dialogue box. Save Options doesn't seem to save this default
font setting in Emacs 22.3, so I use the customize buffer to save it. Then I
open Emacs 23.0.93, which uses Adobe Caslon Pro as the default font. This is
on a computer running Windows Vista. When I originally reported this bug I
was using a different computer running Windows XP and probably a slightly
different set of installed fonts, although they both have Sanvito, which I
think is another Adobe font.

Francis

bug#3319: 23.0.93; Default font

[Kenichi Handa]

In article <jwviqjxyxim.fsf-monnier+emacsbugreports@gnu.org>, Stefan Monnier <monnier@iro.umontreal.ca> writes:

> Breaking backward compatibility might be OK, but I still wonder: how
> come we ended up with "Sanvito Pro Subhead" from a spec of
> a non-existing "outline-lucida sans typewriter"?

When there's no font of the specified family, Emacs tries to
find a font without specifing family.   In that case, which
family is selected is just arbitrary.

---
Kenichi Handa
handa@m17n.org

bug#3319: 23.0.93; Default font

[Stefan Monnier]

>> Breaking backward compatibility might be OK, but I still wonder: how
>> come we ended up with "Sanvito Pro Subhead" from a spec of
>> a non-existing "outline-lucida sans typewriter"?

> When there's no font of the specified family, Emacs tries to
> find a font without specifing family.   In that case, which
> family is selected is just arbitrary.

Shouldn't we fallback on the default font families
(Sans/Courier/...) instead?


        Stefan

bug#3319: 23.0.93; Default font

[Kenichi Handa]

In article <jwvvdnwwmcc.fsf-monnier+emacsbugreports@gnu.org>, Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> Breaking backward compatibility might be OK, but I still wonder: how
>>> come we ended up with "Sanvito Pro Subhead" from a spec of
>>> a non-existing "outline-lucida sans typewriter"?

> > When there's no font of the specified family, Emacs tries to
> > find a font without specifing family.   In that case, which
> > family is selected is just arbitrary.

> Shouldn't we fallback on the default font families
> (Sans/Courier/...) instead?

It sounds good, but the proper fallback font family is
different on platforms.  For instance, xfns.c has this code:

      char *names[]
	= {
#ifdef HAVE_XFT
	    /* This will find the normal Xft font.  */
 	    "monospace-12",
#endif
	    "-adobe-courier-medium-r-*-*-*-120-*-*-*-*-iso8859-1",
	    "-misc-fixed-medium-r-normal-*-*-140-*-*-c-*-iso8859-1",
	    "-*-*-medium-r-normal-*-*-140-*-*-c-*-iso8859-1",
	    /* This was formerly the first thing tried, but it finds
	       too many fonts and takes too long.  */
	    "-*-*-medium-r-*-*-*-*-*-*-c-*-iso8859-1",
	    /* If those didn't work, look for something which will
	       at least work.  */
	    "-*-fixed-*-*-*-*-*-140-*-*-c-*-iso8859-1",
	    "fixed",
	    NULL };

and w32fns.c has this:

      static char *names[]
        = { "Courier New-10",
            "-*-Courier-normal-r-*-*-13-*-*-*-c-*-iso8859-1",
            "-*-Fixedsys-normal-r-*-*-12-*-*-*-c-*-iso8859-1",
            "Fixedsys",
            NULL };

---
Kenichi Handa
handa@m17n.org

bug#3319: 23.0.93; Default font

[Francis Wright]

> From: Stefan Monnier [mailto:monnier@iro.umontreal.ca]
> Sent: Wednesday 20 May 2009 3:17 am
> To: Kenichi Handa
> Cc: jasonr@gnu.org; 3319@emacsbugs.donarmstrong.com;
> F.J.Wright@qmul.ac.uk
> Subject: Re: bug#3319: 23.0.93; Default font
> 
> >> Breaking backward compatibility might be OK, but I still wonder: how
> >> come we ended up with "Sanvito Pro Subhead" from a spec of
> >> a non-existing "outline-lucida sans typewriter"?
> 
> > When there's no font of the specified family, Emacs tries to
> > find a font without specifing family.   In that case, which
> > family is selected is just arbitrary.
> 
> Shouldn't we fallback on the default font families
> (Sans/Courier/...) instead?

I think that would be much better. A known useable font would be better than
something arbitrary, which could presumably be some kind of symbol font.

Incidentally, Emacs 22 doesn't offer me Sanvito or the Adobe fonts, whereas
Emacs 23 does.

bug#3319: 23.0.93; Default font

[Francis Wright]

> From: Kenichi Handa [mailto:handa@m17n.org]
> Sent: Tuesday, 19 May 2009 2:30 am
> To: Francis Wright; 3319@emacsbugs.donarmstrong.com
> Subject: Re: bug#3319: 23.0.93; Default font
> 
> In article <001801c9d71a$8bade240$a309a6c0$@J.Wright@qmul.ac.uk>,
> "Francis Wright" <F.J.Wright@qmul.ac.uk> writes:
> 
> > Please describe exactly what actions triggered the bug
> > and the precise symptoms of the bug:
> 
> > When I start Emacs 23.0.93 normally, it starts up with the default
> > font set to Sanvito Pro Subhead (regular, 10 point), which is not
> what
> > I expected and is almost unreadable in this context. This is the
> first
> > time I have run Emacs 23 on this computer; previously I was running
> > Emacs 22.3. When I start Emacs 23.0.93 with the -Q option, it starts
> > with Courier New as default font, which is fine (as a default). I
> > think the problem is that I had this in my .emacs file:
> 
> > (custom-set-faces
> >  '(default ((t (:stipple nil :background "white" :foreground "black"
> > :inverse-video nil :box nil :strike-through nil :overline nil
> :underline nil
> > :slant normal :weight normal :height 98 :width normal :family
> > "outline-lucida sans typewriter"))))
> >  '(woman-italic-face ((t (:foreground "red" :slant italic)))))
> 
> I think what you intend by :family "outline-lucida sans
> typewriter" is actually :foundry "outline" :family "lucida
> sans typewriter", right.  If so, I've just installed a fix.
> Please try again.

This problem seems to be fixed in Emacs 23.0.94, which starts with the
default font Lucida Sans Typewriter even when it is specified as :family
"outline-lucida sans typewriter" in my .emacs file.

Thanks, Francis

bug#3319: 23.0.93; Default font

[Francis Wright]

Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:

When I start Emacs 23.0.93 normally, it starts up with the default
font set to Sanvito Pro Subhead (regular, 10 point), which is not what
I expected and is almost unreadable in this context. This is the first
time I have run Emacs 23 on this computer; previously I was running
Emacs 22.3. When I start Emacs 23.0.93 with the -Q option, it starts
with Courier New as default font, which is fine (as a default). I
think the problem is that I had this in my .emacs file:

(custom-set-faces
 '(default ((t (:stipple nil :background "white" :foreground "black"
:inverse-video nil :box nil :strike-through nil :overline nil :underline nil
:slant normal :weight normal :height 98 :width normal :family
"outline-lucida sans typewriter"))))
 '(woman-italic-face ((t (:foreground "red" :slant italic)))))

I think Emacs 23.0.93 is misinterpreting this default font setting,
which worked fine with Emacs 22. Erasing this customization seems to
have solved the problem but there seems to be a potential font
incompatibility when upgrading from Emacs 22. I think I saw something
similar when I upgraded to Emacs 23.0.92 on my work computer, but it
was not such a problem and I just reset the font using the Set Default
Font menu item.

If Emacs crashed, and you have the Emacs process in the gdb debugger,
please include the output from the following gdb commands:
    `bt full' and `xbacktrace'.
If you would like to further debug the crash, please read the file
d:/NTEmacs/emacs-23.0.93/etc/DEBUG for instructions.


In GNU Emacs 23.0.93.1 (i386-mingw-nt5.1.2600)
 of 2009-05-02 on SOFT-MJASON
Windowing system distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-gcc (3.4)'

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: ENG
  value of $XMODIFIERS: nil
  locale-coding-system: cp1252
  default-enable-multibyte-characters: t

Major mode: Emacs-Lisp

Minor modes in effect:
  tooltip-mode: t
  tool-bar-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  blink-cursor-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
C-x C-f ~ / . e m <tab> <return> C-s f o n t C-s C-s 
C-s C-s C-s C-s C-s C-g C-g <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<menu-bar> <help-menu> <about-emacs> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <menu-bar> <help-menu> <emacs-news> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <menu-bar> 
<options> <menu-set-font> <help-echo> C-x k <return> 
<help-echo> <C-down-mouse-1> <next> <next> <next> <next> 
<next> <next> <next> <next> <next> <up> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <menu-bar> <help-menu> <se
nd-emacs-bug-report>

Recent messages:
For information about GNU Emacs and the GNU system, type C-h C-a.
Quit [2 times]
For information about GNU Emacs and the GNU system, type C-h C-a.
View mode: type C-h for help, h for commands, q to quit.