- From: John Hudson <tiro@tiro.com>
- Date: Sun, 20 Jul 2014 11:14:55 -0700
- To: Hariraam <hariraama@gmail.com>, indic <public-i18n-indic@w3.org>
- CC: prashant verma <vermaprashant1@gmail.com>, Swaran Lata <slata@mit.gov.in>, Somnath Chandra <schandra@mit.gov.in>, Richard Ishida <ishida@w3.org>
This is a terrible idea. U+094E, which you propose entering before consonant letters or conjuncts as the stem portion of ikar, is itself a left-ordering vowel sign, and would be reordered by Indic shaping engines. So not only would you be affecting the encoding of Devanagari text to obtain a particular display in the absence of a shaping engine, you would be doing so in a way that guarantees the text displays incorrectly when a shaping engine is used. Further, while your method might appear to work in some situations for the particular case of the Devanagari ikar, other Indian writing systems involve left ordering and split vowel signs that cannot be handled with this kind of hack. Whether you like it or not, Unicode has encoded Indian scripts in a way that requires intelligent reordering of glyphs by shaping engines based on string analysis and shaping engine support. The response to this should be to ensure that such support is as widespread and consistent as possible, not to invent hacks that involve mangling text encoding and display. JH
Received on Sunday, 20 July 2014 18:15:25 UTC