L2/17-158 Reconsidering ScriptExtensions added for supporting Tamil fractions Shriramana Sharma, jamadagni-at-gmail-dot-com, India 2017-May-07 In ScriptExtensions.txt of Unicode 9.0, I note: 0BAA ; Gran Taml # Lo TAMIL LETTER PA 0BB5 ; Gran Taml # Lo TAMIL LETTER VA The above two entries are under the presumption that these characters are also used to denote the fractions 1/4 and 1/20 and since all fractions are common to both scripts. I realize I have affirmed this in L2/15-085. This was based on the evidence available then. I had also mentioned that this should be documented in ScriptExtensions but I had not realized that such addition would be added even before the situation of the rest of the fractions was finalized. In retrospect it seems one should have waited for the fractions proposal to crystallize before applying these ScriptExtensions. For 1/4, in L2/16-062 p 16, the Gov't of Tamil Nadu (GoTN) had requested to encode a separate character with raised "head" compared to வ. In L2/16-114 p 2, I had argued against this. For 1/20, in L2/16-062 p 17, the GoTN had requested to encode a separate character with rounded corners compared to ப. In L2/16-114 p 1, I had supported this. The GoTN convened an experts meeting in Chennai on 2016-Oct-20 and invited me to resolve these and other pending issues. Based on the decisions taken at this meeting, the GoTN has recently submitted the finalized proposal L2/17-069. Epigraphists present at the meeting affirmed that older writings do consistently distinguish between letters and fractions/symbols and thus the conflation must have happened during the transition to manuscripts/printing. To enable capturing the distinction in the epigraphs faithfully in digital text, the encoding should provide characters for the fractions separate from the letters. Based on this input, it was decided to disunify both these characters from the "lookalikes" existing in the Tamil BMP block. This is recorded in L2/17-069 p 6 §3 points 1 and 2. In the light of this proposal, these ScriptExtensions entries would no longer be valid as the distinct characters are those that should be used. I am not sure to what extent ScriptExtensions are covered under the stability policy. So the UTC should decide what is to be done about this. -o-