On 19 Jun 2013, at 07:54, Denis Jacquerye <[email protected]> wrote: > Marshallese uses the letters L/l, M/m, N/n, and O/o with cedilla. > > The Ad Hoc http://www.unicode.org/L2/L2013/13128-latvian-marshal-adhoc.pdf > concluded that encoding > LATIN CAPITAL LETTER MARSHALLESE L WITH CEDILLA > LATIN SMALL LETTER MARSHALLESE L WITH CEDILLA > LATIN CAPITAL LETTER MARSHALLESE N WITH CEDILLA > LATIN SMALL LETTER MARSHALLESE N WITH CEDILLA > would cause the least architectural disruption and would be the best > way to proceed. > > How can that be the best way?
It treads most lightly on the existing data. > How would one rationalize using one diacritic U+0327 with M/m and O/o but not > with L/l and N/n in Marshallese? The same way one would rationalize using precomposed ãẽĩñõũỹ (aeinouy with tilde) but a necessarily de-composed g̃ (g with tilde) in Guaraní. > A single combining diacritic to use with Marshallese L/l, M/m, N/n and O/o > would be easier to deal with. It would introduce yet another confusability problem. > It would require less new characters to be encoded and would make it easier > to support in fonts (adding 1 instead of 4). No! Because if you added a single new character you'd have to make sure you had good glyph placement with LlMmNnOo which is eight glyphs. > It would also be easier to implement on keyboard layouts (same behaviour four > all Marshallese letters with cedilla instead of 2 different behaviours) . First off, what does a Marshallese keyboard look like anyway? Second, well, maybe, but I am still convinced that this is the best solution. Keyboards aren't that hard to implement. It's interesting what to do in an internationalized situation. One might implement a keyboard that followed the glyph (cedilla or comma below) rather than the underlying encoding for instance. Michael Everson * http://www.evertype.com/
