German gets a pass, but not Spanish and French?The reason I don't like having all the international symbols on there is because:
To be complete [i.e. not miss some vowels/consonants], you need to fill the entire keyboard with Meta symbols. And even then each language will only use a small fraction of what you put on the keyboard. You end up cluttering the whole keyboard with symbols when no single person will use all of them. And even then you still won't get everything.
Also, each language probably has some good defaults under AltGr somewhere, which are probably already standard. (So as long as ED doesn't use AltGr for the modifier, they are ok.) Hackers can use the tool Neelix is promising to develop to set their own mappings with Meta if they are not happy with the AltGr possibilities.
[edit]
German gets a pass because ED is cool.
That does not make sense to me.
The 6 official languages of the UN are: Arabic, Chinese, English, French, Russian, Spanish.
Arabic, Chinese and Russian do not use a Latin alphabet, so we can't hope to cover those, but English, French and Spanish we can cover quite easily: we just need éèàçñ and some diacritics.
Removing the "clutter" of the more obscure letters (Æ Ø Å Đ Ł), we would get something like this:
http://www.keyboard-layout-editor.com/#/layouts/ec5f7a737b72261c295b823b52ce73e3
~1200 million people speak English
460 million people speak Spanish
336 million people speak Portuguese
274 million people speak French
over 150 million people speak Swahili
145 million people speak German
73 million people speak Tagalog (Filipino)
64 million people speak Italian
28 million people speak Dutch
~17 million people speak Afrikaans
That means 2674 million people can type their language well without needing any remapping or having to remember unlabeled things (English, Spanish, Portuguese, French, Swahili, German, Tagalog, Italian, Dutch, Afrikaans).
If you remove the letters and keep only the diacritics, those languages can still be typed (e.g. by doing tilde+n to get ñ or umlaut+u to get ü), but it becomes a lot less efficient. Only the ones in bright green remain efficient with diacritics-only (1459 million).
If you remove also the diacritics, then all of the above become near impossible to type except English and Swahili (1350 million).
So roughly speaking, the amount of people who can write their language (efficiently) roughly doubles by adding the diacritics (at least ′ ` ¨ ~ ˆ, preferably also ¸ .ˇ ) and dedicated Meta-keys for ñçéèàüöäß.
Sorted from most important to least important extra letters, I think this would be roughly the right order:
ñ ç é è à ü ö ä ß å Ł æ ø Đ
Here's the link to the big table again, for convenience:
https://docs.google.com/spreadsheets/d/14mVNNzLhmhyA4sCIAPSoDv404xtctfLSiEMTtTlEzfw/edit?usp=sharing