Dear Chris Fynn, This font should be tried without Uniscribe support and without any other fully conformant Tamil fonts to understand the scientific principles behind the current recommendations. Of course in real use these supports are essential. I'll soon be publishing the new version font for real use where the opportunity to see the raw data will be reduced. Sinnathurai Srivas << The proposal is not acceptable. The Current state of allocations is based on scientific principles. The new proposal is of usage based principle. It is not only "ai" (I guess its not AI as described below), but also au, e, ee, o, oo have the similar charactersitics. Unless all of these are changed to usage based one, it should not be accepted as a solution/change. A mixed solution is a recipi for disaster. Contextual is only acceptable if all of the recommendations are contextual based. In my opinion the current scientific solution should be kept in tact so that the process handling language becomes sophisticated as it is scientific based. As I do not have detail information on this proposal, if my assumption about the proposal is wrong please correct me. I have published a Tamil Unicode Font for test purposes. This proposal and the other characters I mentioned above may be better understood by visiting (please do not visit if you do not wish to view sicML) http://www.geocities.com/avarangal/tamilunicode.html Sinnathurai Srivas < They involve a change to the contextual processing model involving the AI vowel. John F. >> ============ Dear Chris Fynn, here it is. Though intended, somehow it missed the Unicode list before. <<< Perhaps it would be more worthwhile to discuss the proposed changes to Tamil on the Unicode list [EMAIL PROTECTED] rather than on the OpenType list - at the very least some mention should be made on the Unicode list as well. So far nothing seems to have been said there about these proposed changes. BTW Though I don't read Tamil, I seem to get your web page rendered correctly in all but one or two places without your font installed (I do have Microsoft's Tamil IME installed under Win 2K) Chris Fynn Dzongkha Computing Project Thimphu Bhutan
url address corrected http://www.geocities.com/avarangal/tamilunicode.html << The proposal is not acceptable. The Current state of allocations is based on scientific principles. The new proposal is of usage based principle. It is not only "ai" (I guess its not AI as described below), but also au, e, ee, o, oo have the similar charactersitics. Unless all of these are changed to usage based one, it should not be accepted as a solution/change. A mixed solution is a recipi for disaster. Contextual is only acceptable if all of the recomondations are contextual based. In my openion the current scientific solution shold be kept in tact so that the process handling language becomes sophisticated as it is scientific based. As I do not have detail information on this proposal, if my assumption about the proposal is wrong please correct me. I have published a Tamil Unicode Font for test purposes. This proposal and the other characters I mentioned above may be better understood by visiting (please do not visit if you do not wish to view sickML) http://www.geocities.com/avarangal/tamilunicode.html Sinnathurai Srivas < They involve a change to the contextual processing model involving the AI vowel. John F. >> >>
