L2/04-432 Source: Cathy Wissink Title: Indic collation: action items 99-20 and 99-29 Date: 2004-12-31 I had two action items concerning Indic collation: 99-A20 Get more information on the correct ordering for U+0CBD from the author of L2/04-102 99-A29 Determine the collation order of Bengali Letter Khanda Ta After a bit of research and discussion with my colleagues, here's what I know: 1. Concerning U+0CBD (Kannada Sign Avagraha): the author of L2/04-102 is correct. Avagraha should be treated as a sign. It's not clear where it should sort relative to the other signs (as it's used for Sanskrit and it's difficult to contrast it in real collated text with other sign usage; we had no luck tracking this down). For the time being, it may be useful to sort the avagraha in the default table analogous to how it is treated respectively in Gujarati or Oriya. 2. Concerning Bengali Khanda Ta: our current research suggests it should sort as ta + hasant (virama). The dictionaries and word lists are consistent in their treatment of the Khanda Ta as equivalent to the virama form, but given the amount of discussion that Khanda Ta is distinct from ta + hasant, perhaps the dictionaries have not yet caught up to this usage. I would recommend a conservative approach, making it equivalent to ta + hasant until there is sufficient evidence that it should be treated differently.