Re: [Indic-computing-devel] [Indic-Computing-announce] First National OpenType Font Workshop

[Krishnamurthy N. <kn...@ya...>]

Hi all,

My comments on OpenType

* This being an extension of TrueType, M$ and Adobe   
   hold patents on the format. So, though the format
is    published, it's not open. It may be possible, in
that   case, to develop open source tools to
create/edit
  opentype font files (such as extending pfaedit), but
   then M$ and Adobe always have the last say and may 
    have hidden extensions that only their
  tools would be able to handle.
* The rendering algorithms right now are embedded in  
   commecial tools such as M$'s Inscribe. So, they may
    even patent the rendering algorithms thus
  blocking any open source rendering tools.
* Now about the great advantages of the GSUB, GPOS    
   tables - something fundamental is being missed out 
  by all developers here : display rendering is
closely   tied with (keyboard) input and since Indian
  languages are phonetic, the mapping of input to     
   appropriate choice of glyphs, their relative       

  Moreover, the contextual positioning of various     
   glyphs, especially the mathras and dependent vowel 
    signs (more so with 'split vowel signs') is very
much   dependent on the input context of relevant
letters.

  Subtle contexts such as when a 'm' should be mapped 
   to an anuswara can't be specified with these GSUB  
    table rules at all.

  Though the documentation of GSUB tables says "The   
   text-processing client uses the GSUB data to manage
    glyph substitution actions", it's quite a bit of  
     burden on 'each' of the applications to figure
out      what glyph substitution to use when!

  Of the six types of glyph substitution, the one that
   is touted as the most powerful, "contextual        
    substitution", works on the context of surronding 
     glyphs and not the input context! And that's not
what   one looks for.

  All in all, every app has to bother about looking up
   the GSUB and GPOS tables to figure out various     
    substitions. So, no common library that will
  do the job (like pango/gtk).  

By contrast, the generic transliteration rules
framework for Indian languages that I developed and
presented in the Sep 2002 Indic-computing workshop
intelligently ties together the input sequence &
context with display rendering, with very
sophisticated facilities for context specification
(both input and display), glyph reordering and so on.
The C library, which is independent of any of the
input languages/scripts, is available in sourceforge,
along with sample rule files for four of the Indian
languages
(Hindi, Telugu, Tamil and Kannada).

About this upcoming OTF workshop/seminar : it's not
free and not related to open source projects. So, why
is it being advertised on indic-computing list which
is for open source development ?  

cheers,
Nagarajan



__________________________________________________
Do you Yahoo!?
Yahoo! Shopping - Send Flowers for Valentine's Day
http://shopping.yahoo.com