Esperanto is useless in IT. It's words are taken from existing languages and doesn't have for example 12 words for a different type of snow (eskims have 12 of varieties while my language covers only two types of snow). Esperanto's gramatics is also not rich enough to cover just about anything possible.
If you know anything about databases then "collating sequence" is not a strange term for you - it's used to assure sorting records to different alphabets since relying on ASCII can be used only for english. Symbols used in other alphabets are not sequentially added to it, so words beginning with for example Ä would be at the end of the list instead of getting a spot right after those words that begin with A.
Collating sequence doesn't have to be even a bit similar to ASCII, you may put a symbol A on the spot where brackets stand, for example. The point is that machine doesn't care for someone's patriotic feelings nor cares if something exists in reality. Nor should we actually be.
Something similar can be developed for realtime language translator, so you practically "upgrade" or "downgrade" english gramatics and dictionary to be used for translation to some specific language. But the problem here is that the goal is not to make translate english to any language, but translate any language to any language. So instead thousands of "collating sequences", a completelly new, universal one, nonexisting one, should be designed.
Yea, I'm aware that would stop the idiocy "my language is superior" by different kinds of nationalists and patriots, but get real. No existing language can be compatible with all existing languages out there. If it was, hell, the star-trek-like universal translator software would exist ages ago.