23.0.60; wrong glyph widths of some Unicode codepoints

23.0.60; wrong glyph widths of some Unicode codepoints

[Torsten Bronger]

The sub/superscripts (e.g. ⁴ or ₃) are narrower than the ordinary
characters if using a monospace font (monospace-13 in my case), with
the exceptions of ¹²³.  The same is true at least for the prime
character ′.  The right arrow → and the cubic root ∛ used to have a
"correct" width, but with the recent Emacs, it is too wide (the
cubic root even too tall, too).

Is it sensible at all to report such observations, maybe on a more
systematic/complete basis?


In GNU Emacs 23.0.60.3 (x86_64-unknown-linux-gnu)
 of 2008-05-22 on wilson
Windowing system distributor `The X.Org Foundation', version 11.0.10400090
configured using `configure '--prefix=/usr/local/'
--mandir=/usr/local/share/man/' '--infodir=/usr/local/share/info/'
--with-x-toolkit=no' '--with-xpm' '--with-jpeg' '--with-tiff' '--with-gif'
--with-png' '--with-x''

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: de_DE.UTF-8
  value of $XMODIFIERS: nil
  locale-coding-system: utf-8-unix
  default-enable-multibyte-characters: t

Major mode: Summary

Minor modes in effect:
  global-auto-revert-mode: t
  savehist-mode: t
  display-time-mode: t
  desktop-save-mode: t
  mouse-wheel-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  line-number-mode: t

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[Kenichi Handa]

In article <8763t6r1tp.fsf@physik.rwth-aachen.de>, Torsten Bronger <bronger@physik.rwth-aachen.de> writes:

> The sub/superscripts (e.g. ⁴ or ₃) are narrower than the ordinary
> characters if using a monospace font (monospace-13 in my case), with
> the exceptions of ¹²³.  The same is true at least for the prime
> character ′.  The right arrow → and the cubic root ∛ used to have a
> "correct" width, but with the recent Emacs, it is too wide (the
> cubic root even too tall, too).

Please type C-u C-x = on normal ascii characters and those
problematic characters and check which fonts are used.

By the way, are you sure that your monospace-13 font has
glyphs for all of them.  In my environtment, monospace font
is actually mapped to "bitstream vera sans mono" and that
font doesn't have glyphs for those characters.  In such a
situation, it's difficult to expect to see them with the
same metrics as others.

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

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[Torsten Bronger]

Hallöchen!

Kenichi Handa writes:

> In article <8763t6r1tp.fsf@physik.rwth-aachen.de>, Torsten Bronger
> <bronger@physik.rwth-aachen.de> writes:
>
>> The sub/superscripts (e.g. ⁴ or ₃) are narrower than the ordinary
>> characters if using a monospace font (monospace-13 in my case),
>> with the exceptions of ¹²³.  The same is true at least for the
>> prime character ′.  The right arrow → and the cubic root ∛ used
>> to have a "correct" width, but with the recent Emacs, it is too
>> wide (the cubic root even too tall, too).
>
> Please type C-u C-x = on normal ascii characters and those
> problematic characters and check which fonts are used.

Unfortunately, another issue with the current CVS version was a
blocker for my particular case, so I switched back to an older
version a couple of minutes ago.  Hmpf.  My current Emacs version
(which displays the sub- and superscripts with correct widths) says

     dejavu sans mono:pixelsize=17:foundry=unknown:weight=regular:slant=normal:width=normal (#x6A6)


> By the way, are you sure that your monospace-13 font has glyphs
> for all of them.  In my environtment, monospace font is actually
> mapped to "bitstream vera sans mono" and that font doesn't have
> glyphs for those characters.

It's the same here.  However, the sub/superscripts used to work
until recently, see above.

> In such a situation, it's difficult to expect to see them with the
> same metrics as others.

If the adjustment of glyphs from different fonts is not feasible, I
must and can live with it.  But it's certainly a useful thing to
have in my opinion.

Tschö,
Torsten.

-- 
Torsten Bronger, aquisgrana, europa vetus
                                      Jabber ID: bronger@jabber.org
               (See http://ime.webhop.org for further contact info.)

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[Kenichi Handa]

In article <87zlqhv71h.fsf@physik.rwth-aachen.de>, Torsten Bronger <bronger@physik.rwth-aachen.de> writes:

> Unfortunately, another issue with the current CVS version was a
> blocker for my particular case, so I switched back to an older
> version a couple of minutes ago.  Hmpf.  My current Emacs version
> (which displays the sub- and superscripts with correct widths) says

>      dejavu sans mono:pixelsize=17:foundry=unknown:weight=regular:slant=normal:width=normal (#x6A6)

> > By the way, are you sure that your monospace-13 font has glyphs
> > for all of them.  In my environtment, monospace font is actually
> > mapped to "bitstream vera sans mono" and that font doesn't have
> > glyphs for those characters.

> It's the same here.  However, the sub/superscripts used to work
> until recently, see above.

Ah!  "dejavu sans mono" and "bitstream vera sans mono"
happen to have the same metrics.

> > In such a situation, it's difficult to expect to see them with the
> > same metrics as others.

> If the adjustment of glyphs from different fonts is not feasible, I
> must and can live with it.  But it's certainly a useful thing to
> have in my opinion.

I agree.  It's possible to implement a code to prefer a font
of the same metrics (e.g. ascent, descent, avgwidth) as the
default one in font-selection.  But, for that, Emacs
consumes more memory and time for font finding.  And even
with that, people may prefer a different font.

I need more time to consider this problem.

BTW, if you know that you want "dejavu sans mono" for them,
how about using it as the fallback font as this:

(set-fontset-font
 "fontset-default" nil
 '("dejavu sans mono" . "iso10646-1") nil 'prepend)

It worked in my environment.

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

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[David Hansen]

Not that it solves any Emacs devel related problems but maybe the this
specific usage problem...

On Fri, 23 May 2008 16:55:53 +0900 Kenichi Handa wrote:

> Ah!  "dejavu sans mono" and "bitstream vera sans mono"
> happen to have the same metrics.

For the "common" characters they are the same.  The DejaVu fonts are
just the Bitstream fonts plus additional glyphs.

> BTW, if you know that you want "dejavu sans mono" for them,
> how about using it as the fallback font as this:

Or use only the DejaVu fonts.

BTW Kenichi:  Do you know urxvt aka rxvt-unicode?  It's a terminal
emulator for X that claims to have good unicode support.  IMHO (not that
I'm in any way qualified ;) it's doing a very good job in picking the
right fonts.  It supports xft as well as old school X fonts.

David

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[Kenichi Handa]

In article <87zlqhiej5.fsf@localhorst.mine.nu>, David Hansen <david.hansen@gmx.net> writes:

> Not that it solves any Emacs devel related problems but maybe the this
> specific usage problem...

> On Fri, 23 May 2008 16:55:53 +0900 Kenichi Handa wrote:

> > Ah!  "dejavu sans mono" and "bitstream vera sans mono"
> > happen to have the same metrics.

> For the "common" characters they are the same.  The DejaVu fonts are
> just the Bitstream fonts plus additional glyphs.

Ah!  But, then why does fontconfig put higher priority to
bistream fonts?  In my environment (debian etch),
/etc/fonts/conf.avail/60-latin.conf has these lines:

		<family>monospace</family>
		<prefer>
			<family>Bitstream Vera Sans Mono</family>
			<family>DejaVu Sans Mono</family>
[...]

> > BTW, if you know that you want "dejavu sans mono" for them,
> > how about using it as the fallback font as this:

> Or use only the DejaVu fonts.

> BTW Kenichi:  Do you know urxvt aka rxvt-unicode?  It's a terminal
> emulator for X that claims to have good unicode support.  IMHO (not that
> I'm in any way qualified ;) it's doing a very good job in picking the
> right fonts.  It supports xft as well as old school X fonts.

I've just tried rxvt-unicode to display HELLO and
UTF-8-demo.txt.  In my environment, it seems that the latest
Emacs does better job.  But, at least, rxvt-unicode aligns
characters better.  It's perhaps because rxvt-unicode pad
some space for that.

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

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[Jason Rumney]

Kenichi Handa wrote:
> Ah!  But, then why does fontconfig put higher priority to
> bistream fonts?

My best guess would be because they have good enough coverage for most 
US/European users, and less memory is used due to the narrower coverage. 
Either that or the GNU/Linux distributors are just being conservative 
with their default configurations and prefering the older original font 
rather than its still under development offshoot.

Re: 23.0.60; wrong glyph widths of some Unicode codepoints

[David De La Harpe Golden]

Kenichi Handa wrote:

> I agree.  It's possible to implement a code to prefer a font
> of the same metrics (e.g. ascent, descent, avgwidth) as the
> default one in font-selection.  But, for that, Emacs
> consumes more memory and time for font finding.  And even
> with that, people may prefer a different font.
> 

FWIW, it's precious monospaced alignment going wrong that makes me do
bad things like the below...

Here, for reference, is an evil way to *force* emacs to use one (1) font
at all times, rather than trying anything frightfully clever :-).

This way, the xft fontbackend only gets to choose from one font
in the first place.  Alignment tends to become perfect, but sometimes
(of course) glyphs may be missing, depending on your choice of font.


*** First, set emacs to use xft FontBackend only.

echo "Emacs.FontBackend: xft" | xrdb -merge
(and add it to your ~/.Xdefaults )

*** Then, make a directory and copy one font into it (might
*** still be a few .ttf files to cover different faces)

mkdir ~/.lonelyfont
cp /usr/share/fonts/truetype/ttf-dejavu/DejaVuSansMono* ~/.lonelyfont/

*** make an alternate fontconfig cache dir.

mkdir ~/.lonelyfontconfig

*** Now, create a fontconfig file that we're going to use
*** to override the usual search paths.

place in ~/.lonelyfont.conf  (the only changes from a typical
~/.fonts.conf are the <dir> and <cachedir> entries - if you
don't include the rest, you'll probably lose your antialiasing and
subpixel rendering settings)


<?xml version="1.0"?><!DOCTYPE fontconfig SYSTEM "fonts.dtd">
<fontconfig>
 <dir>~/.lonelyfont</dir>
 <cachedir>~/.lonelyfontconfig</cachedir>
 <match target="font" >
  <edit mode="assign" name="rgba" >
   <const>rgb</const>
  </edit>
 </match>
 <match target="font" >
  <edit mode="assign" name="hinting" >
   <bool>true</bool>
  </edit>
 </match>
 <match target="font" >
  <edit mode="assign" name="hintstyle" >
   <const>hintfull</const>
  </edit>
 </match>
 <match target="font" >
  <edit mode="assign" name="antialias" >
   <bool>true</bool>
  </edit>
 </match>
</fontconfig>


*** Now start emacs, using the FONTCONFIG_FILE environment variable

FONTCONFIG_FILE=~/.lonelyfont.conf emacs


*** Laugh maniacally.