* Words cannot contain apostrophes or dashes due to ambiguity.
* Contractions need to be split, translated, rejoined.
* Separation and mixing rules of ASCII and native alphabets.
* Separation and mixing rules of ASCII and native digits.
* Related languages to inherit from language family base class.
* Borrowing from related languages.
* Transcription / Transliteration.
* Preserving case and other attributes.
* Encoding conversion besides MacOS.
* Choice of scripts (Belarusian and Serbian).
* Use SIL codes for languages not supported by ISO 639-2.
* Proper UTF-8 support.

* Missing languages:
	* Amharic
	* Arabic
	* Armenian
	* Chinese
	* Farsi
	* Georgian
	* Hebrew
	* Indic languages
	* Japanese
	* Latin
	* Urdu
	* Vietnamese

* Basque
	* What is the correct name and spelling of the language in Basque?
	  Euskara	187,000	  Euskera	83,000

* Belarusian
	* Both Cyrillic and Latin (Lacinka or Latsinka) scripts really are used.
	* Two Cyrillic versions exist (Taraskevitza and Narkomovka).
	* See http://www.belarus-misc.org/bel-ling.htm
	* What is the correct name and spelling of the language in English?
	  Belarusian	194000	  Belorusan	52
	  Byelorussian	48400 	  Belarussan	23
	  Belarussian	47800 	  Byelarusian	22
	  Belorussian	23600 	  Byelorusan	4
	  Belarusan	5150  	  Byelarusan	3
	  Belorusian	952   	  Belorussan	2
	  Byelarussian	188   	  Byelorussan	0
	  Byelorusian	89    	  Byelarussan	0
	* What is the correct transliteration of the Belarusian name for Belarusian?
	  Beloruski	1480	  Belaruski	719
	  Beloruska	1260	  Belaruska	184

* French
	* Handle contractions with apostrophe: l' j'
	  Specifically: j' m' t' s' qu' s'il s'ils c'est
	* Handle hyphens: est-ce
	* Have I handled the oe ligature correctly?
* German
	* Compound nouns need to be split, translated, and rejoined.
	* Normalize ss=>, ae=>, oe=>, ue=> for dictionary lookup.
	* Normalize ss=>, ae=>, oe=>, ue=> for dictionary lookup.
	* Pluralization and generalization rules could point to last
	  noun in compounds.
* Hebrew
	* Cannot find grammar on internet.  Encoding standards and
	  conversions difficult to understand.  Behaviours of browsers
	  and editors difficult to understand.
* Japanese
	* Words are not separated.  We need custom code to separate them.
* Korean
	* Hangul endings can be applied to ASCII words!
* Luxembourgish
	* What is the correct name and spelling of the language in English and Luxembourgish?
	  Luxembourgeois	59,400
	  Luxembourgish		46,700
	  Ltzebuergesch	8,410
	  Luxemburgian		1,660
	  Ltzeburgesch		213
	  Letzburgisch		29
	  Ltzebuergeusch	25
	  Ltzebrgesch		2
* Malay
	* What is the correct name and spelling of the language in English and Malay?
	  Malay			426,000
	  Bahasa Melayu		31,600
	  Bahasa Malaysia	27,200
	  Bahasa Malayu		332
	  Bahasa Melaju		22
* Norwegian
	* How should we decide between Bokml and Nynorsk?
	  ISO 639 has three tags: nb = Bokml, nn = Nynorsk, and no = Norwegian which
	  is what we use currently but it's ambiguous!
	* Should we have two separate language modules or one containing both variants?
* Romanian
	* Should try to use the characters with "comma below" rather than "cedilla"
	  whenever possible.
* Romansch
	* What is the correct name and spelling of the language in English and Romansch?
	  Romansh		31,800
	  Romanche		9,950
	  Romansch		3,420
	  Rhaeto-romance	3,340
	  Rheto-romance		275
* Serbian
	* Both Cyrillic and Latin scripts really are used
* Serbo-Croatian
	* What to treat as a language and what to treat as a regional variety?!
	  Unix locales and ICU seem to support Croatian, Serbian, and Serbo-Croatian.
	  In these cases, Serbo-Croatian seems to refer to a Latin variant and Serbian
	  always refers to a Cyrillic variant.
* Tahitian
	* What is the Tahitian name for the Tahitian language?
* Thai
	* Words are not separated.  We need custom code to separate them.
	* Native digit handling could be improved.
* Vietnamese
	* Not enough information on the internet.

