I'm currently in Cilamitay, in the east of Taiwan. There is a little meeting of Taiwanese Free Software people and people from the Amis, Taroko and Puyuma tribes, with the idea of starting localisation efforts for some aboriginal languages.
These are some of the issues we are going to discuss:
Language code
A new ISO standard (639-3) will hopefully be formalised in January that will include the language codes for the Taiwanese aboriginal tribes. We'll have to work some temporary solution, but there's good hope that it won't have to be temporary for long.
List of characters
Because of Christian missionary influence, both Amis and Taroko use a roman alphabet, with accents. We need to work out the complete list of character and accent combination, see if everything is in Unicode, see how they sort.
We then need to find a comfortable way to input them using the keyboards normally available here (English US layout): compose key? Dead keys? How about on Windows?
Womble2 on IRC tells me that on Windows one can works with MSKLC.
Technical terms and country list
We need to work out how to map terms that do not exist in the language.
Technical terms are usually borrowed from Japanese.
Names for all the countries in the world probably do not exist.
Translation interface
We need to find an easy to use interface to input the translations.
There is Rosetta.
There is Pootle. (Thanks to Christian Perrier for pointing me at it)
There is Webpot.
Update: there is now a wiki page on the Debian wiki.