Clarify language code documentation for NLLB

[johnml1135]

It roughly says what we want, but we should do a few things:

• In all places, point people to engine creation for more documentation on language codes
• In engines, for language codes, say "See engine type for more information"
• For NMT, say something to the effect:

  If you use a code from the NLLB-200 list, Serval will utilize everything the NLLB-200 model already knows about that language when doing the translation. If you use an IETF language tag, Serval will convert it to a NLLB-200 code. If the language you are working with is not in the NLLB-200 model, any language code that is not in the NLLB-200 list can be chosen as name will have no effect. When possible, the use of IETF language tags are recommended.

[johnml1135]

This issue is coming about because of frequent questions about how to use language codes.

[johnml1135]

Should we also say something about simplified Chinese is assumed?

[Enkidu93]

I'm going to take care of this now. I would lean toward it not being necessary to mention defaulting to simplified Chinese. As a user, I would have assumed it would default to simplified characters. Unless we expect to have numerous language projects working out of Taiwan (and I know there are a handful of un-Bibled, indigenous languages there still), I don't see the question of how to use traditional characters even coming up.

[johnml1135]

Good enough for now.