freefont-bugs
[Top][All Lists]
Advanced

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

[Freefont-bugs] Re: [Indlinux-group] Re: Testing of FreeSans - devanagar


From: Primoz PETERLIN
Subject: [Freefont-bugs] Re: [Indlinux-group] Re: Testing of FreeSans - devanagari script - latest version
Date: Sat, 4 Feb 2006 00:20:38 +0100

Hello Rahul,

2006/2/3, rahul bhalerao <address@hidden>:
The open-source font editor "Fontforge"(formerly known as pfa-edit) allows merging
different fonts automatically. The only problem is that it doesnt do the header definition
for you. Thus even the achor pints remain undefined. All you have to do is define all of
the used anchor points in the target file first. Solve conflicts with anchoring, if any(something
like you wont expect conflicts in unicode positions of different scripts glyphs). This procedure
is documented on "Samyak" site :
http://www.gnowledge.org/Gnoware/localization/merge.htm
The step-6 on this page suggests copy-pasting all the characters. But this can be replaced with
'Element ->Merge Fonts' function in Fontforge(Though I haven't tried this yet, it should work).
Or copy pasting can be even  done  by selecting entire block of characters at a time and paste
into target file, this wont take any extra time or efforts. Also it will retain all the typographic
rules as well.

I believe another problem with merging fonts using Fontforge might arise: namespace polution. If I am correct, OpenType tables rely on unique glyph names. However, quite often I encounter two types of glyph names of unencoded characters (code point -1) that might pose problems:

* generic glyph names, e.g. "glyph331"
* too general glyph names, e.g. "kaa" (I believe that "kaa" can occur in any of the Indic scripts; a better solution would be, say "dvng_kaa" or something along this line)

When merging two fonts, it may occur that there are two characters named "glyph331" or "kaa". I haven't tested yet the reaction of Fontforge in this case, but in any case I believe it is the responsibility of font maker/font merger rather than the software to assure unique glyph names. I might write a script for spotting problematic cases, though.

With kind regards, Primoz

reply via email to

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