Font regression on latest CVS

Font regression on latest CVS

[İsmail Dönmez]

[-- Attachment #1: Type: text/plain, Size: 445 bytes --]

Hi all,

Font display has regressed on latest CVS once again. Currently when I
insert a non-ascii characters into a line the lines height changes.
Attached is two screenshots. In ascii.png you can see that height of
"ERC" is OK but when I insert i-dotless the height goes up
unnecesseraly. This is a regression introduced in last 2-3 days.

Regards,
ismail

-- 
I do object-oriented programming - if the customer objects, I do more
programming.

[-- Attachment #2: ascii.png --]
[-- Type: image/png, Size: 2118 bytes --]

[-- Attachment #3: non-ascii.png --]
[-- Type: image/png, Size: 1441 bytes --]

Re: Font regression on latest CVS

[İsmail Dönmez]

On Thu, Jul 10, 2008 at 7:12 PM, İsmail Dönmez <ismail@namtrac.org> wrote:
> Hi all,
>
> Font display has regressed on latest CVS once again. Currently when I
> insert a non-ascii characters into a line the lines height changes.
> Attached is two screenshots. In ascii.png you can see that height of
> "ERC" is OK but when I insert i-dotless the height goes up
> unnecesseraly. This is a regression introduced in last 2-3 days.

FWIW this only happens with Droid font and when using Turkish characters.

Regards,
ismail

-- 
I do object-oriented programming - if the customer objects, I do more
programming.

Re: Font regression on latest CVS

[Kenichi Handa]

In article <19e566510807100912l59f8b8cv132541a71dec886a@mail.gmail.com>, "=?UTF-8?Q?=C4=B0smail_D=C3=B6nmez?=" <ismail@namtrac.org> writes:

> Font display has regressed on latest CVS once again. Currently when I
> insert a non-ascii characters into a line the lines height changes.
> Attached is two screenshots. In ascii.png you can see that height of
> "ERC" is OK but when I insert i-dotless the height goes up
> unnecesseraly. This is a regression introduced in last 2-3 days.
[...]
> FWIW this only happens with Droid font and when using Turkish characters.

The line heigh will change if the height of the font itself
is taller than the height of the current line, even if the
shown glyph itself is not that tall.

What is the full font name shown by C-u C-x = when you put
cursor on Turkish character.

And, which font was selected for Turkish previously?

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

Re: Font regression on latest CVS

[İsmail Dönmez]

Hi,

On Fri, Jul 11, 2008 at 7:21 AM, Kenichi Handa <handa@m17n.org> wrote:
> In article <19e566510807100912l59f8b8cv132541a71dec886a@mail.gmail.com>, "=?UTF-8?Q?=C4=B0smail_D=C3=B6nmez?=" <ismail@namtrac.org> writes:
>
>> Font display has regressed on latest CVS once again. Currently when I
>> insert a non-ascii characters into a line the lines height changes.
>> Attached is two screenshots. In ascii.png you can see that height of
>> "ERC" is OK but when I insert i-dotless the height goes up
>> unnecesseraly. This is a regression introduced in last 2-3 days.
> [...]
>> FWIW this only happens with Droid font and when using Turkish characters.
>
> The line heigh will change if the height of the font itself
> is taller than the height of the current line, even if the
> shown glyph itself is not that tall.

The problem is Turkish ı is actually smaller than i, thats why I think
this is a bug.

> What is the full font name shown by C-u C-x = when you put
> cursor on Turkish character.

It says:

    character: ı (305, #o461, #x131)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x0131
           syntax: w 	which means: word
         category: h:Korean j:Japanese l:Latin
      buffer code: #xC4 #xB1
        file code: #xC4 #xB1 (encoded by coding system utf-8-unix)
          display: by this font (glyph code)
    xft:-unknown-Kayrawan-normal-normal-normal-*-14-*-*-*-*-0-iso10646-1 (#xF5)

Character code properties are not shown: customize what to show

There are text properties here:
  auto-composed        t
  fontified            t


> And, which font was selected for Turkish previously?

Since it was same as the other characters I assume Droid Sans Mono.

Regards,
ismail

-- 
I do object-oriented programming - if the customer objects, I do more
programming.

Re: Font regression on latest CVS

[Kenichi Handa]

In article <19e566510807102158j1dea9a94v32b97e6489acb530@mail.gmail.com>, "=?UTF-8?Q?=C4=B0smail_D=C3=B6nmez?=" <ismail@namtrac.org> writes:

> > And, which font was selected for Turkish previously?

> Since it was same as the other characters I assume Droid Sans Mono.

Ah, I see.  I deleted the code of preferring the specified
font for Latin characters while doing various experiments,
and forgot to recover that part.  :-(

I've just installed a fix.

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

Re: Font regression on latest CVS

[İsmail Dönmez]

Hi,

On Fri, Jul 11, 2008 at 10:36 AM, Kenichi Handa <handa@m17n.org> wrote:
> In article <19e566510807102158j1dea9a94v32b97e6489acb530@mail.gmail.com>, "=?UTF-8?Q?=C4=B0smail_D=C3=B6nmez?=" <ismail@namtrac.org> writes:
>
>> > And, which font was selected for Turkish previously?
>
>> Since it was same as the other characters I assume Droid Sans Mono.
>
> Ah, I see.  I deleted the code of preferring the specified
> font for Latin characters while doing various experiments,
> and forgot to recover that part.  :-(
>
> I've just installed a fix.

Fix confirmed, thanks!

/ismail

-- 
I do object-oriented programming - if the customer objects, I do more
programming.