This discussion reminds me about the current status, in the Bugis-Makassarese Lontara script, for the encoding of the (non-traditional) diacritics for noting vowelless consonnants : notably an unusual virama sign (probably used only in technical documents used by linguists), and its alternatives for specific cases (an anusavara sign for noting the post-nasalisation, or a sign to note the gemination).
I wonder also if a Bugis font should map a ligature for noting 'yi' (based on a pair of zero-consonnant letters 'a', both with a diacritic vowel 'i') occuring before another vowel (that would then need to use a 3rd zero-consonnant 'a'). Are there other known ligatures in Bugis-Makassarese Lontara script (except this one, or the preencoded ligatures that are already part of the encoding in the UCS for prenasalized letters, considered as distinct letters by their own)? Are there also pending extensions for these modern use of the script, not used in its traditional form ? There remains only two free slots in the current Bugis block, not enough to fit all of these extensions (even if some of them were already described in 2003 in the last proposal for encoding this script in the UCS, submitted to the UTC and JTC2 by Michael Everson). May be they were not encoded because the problem was a lack of agreement about the choice of the unusual virama or its alternatives for more common cases occuring in "modern" spoken Bugis (or because these extensions were introduced by specific authors, i.e. each one with a single source). -- Philippe.
