freetype-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ft-devel] a flag making FT_Open_Face() to use fallback family name


From: mpsuzuki
Subject: Re: [ft-devel] a flag making FT_Open_Face() to use fallback family name
Date: Fri, 27 Aug 2010 02:00:45 +0900

Thank you for comment.

On Thu, 26 Aug 2010 11:35:36 -0400
David Bevan <address@hidden> wrote:

>Two brief comments:

>We've been treating unnamed fonts from PDFs as "tricky" in
>our code for some time, with the following addition to
>tt_check_trickyness():

>    /*** Emtex change ***/
>    /* ER38292 if no family name in the font program */
>    /* treat the font with no name as one of the 'tricky' fonts */
>
>    if ( !name )
>      return TRUE;

Interesting.

>Given the expiry of the TT patents, is there still a need
>for handling tricky fonts specially? We assumed not, and
>that when we got around to upgrading to the new FreeType
>(2.4.1), we could handle all fonts the same. Have we misunderstood?

You don't misunderstand. Yes, it's one of the reasonable
attitudes: older complains about FreeType's hinted "unwanted"
result are applicable to the unpatented hinter, but they are
not applicable to FreeType 2.4.x's "correct" rasterization
result by genuine TrueType hinter. If correct rasterization
is unwanted quality, the font production should be complained,
the font rasterizer with genuine hinter should not be complained.

Although the question "which is better result, hinted or
unhinted?" should be evaluated since FreeType 2.4.x
(because of fundamental change of default hinter) again,
I'm afraid it cannot make everybody agreed to use hinted
result. Some people may say: "I don't know the complex
story in TrueType rasterizer, but I prefer unhinted result.
I don't want to enable hinter by default".

So, my patches are basically for the case "if the application
developer don't agree to enable the hinter by default".

# If I remember correctly, some GNU/Linux distributions had
# enabled patented TrueType hinter for a long time, thus,
# some people had complain the "correct" rasterization result.

OK, I will try to check some fonts that made Okular people
decided to disable the hinter:
  http://mail.kde.org/pipermail/okular-devel/2009-October/004913.html

Regards,
mpsuzuki



reply via email to

[Prev in Thread] Current Thread [Next in Thread]