I took a first attempt at the Macrolanguages table. I only marked those as macrolanguages if there was a language category for the macrolanguage and none of the individual codes were in use (though this might miss some text uses in translation tables). Please fill in those that are de-facto decided or link to previous discussions. --Bequw → ¢ • τ 07:51, 14 April 2009 (UTC)
As there's two types of unfinished mergers, I'll make statements that people can comment on separately.
Language Families: I believe there's only three of these left (excluding {{LG.}}). I propose we make some code exceptions for the final few language families not covered by ISO 639-5. I see two options, (1) use the reserved ISO 639 alpha-3 space (qaa-qtz), (2) make extensions from the existing ISO 639-5 codes (e.g. cel-bry for Brythonic, cel-gae for Gaelic, and sem-jar for Judeo-Aramaic)
--Bequw → ¢ • τ 03:34, 28 September 2009 (UTC)
roa-gal), so I think we don't have any reason to differentiate then from Brythonic, Gaelic and Judeo-Aramaic by using the qaa-qtz group. And, by comparison, these three extensions cel-bry, cel-gae and sem-jar all look good. Furthermore, in my opinion, they're easy to remember. --Daniel. 04:17, 28 September 2009 (UTC)gmw-anf for "Anglo-Frisian" and sai-car for "Cariban"? --Bequw → ¢ • τ 15:26, 5 October 2009 (UTC)Dialects: After re-reading the discussions at Template_talk:etyl#Regional_language_tags, User_talk:Robert_Ullmann/2008b#Standardizing_dialects, and the RFDO of fr-ca I get two things. First, what we want out of these templates is very simple, just a wiki link and category. Second, coding every regional/temporal difference could get out of hand. In an effort to find a different solution, could we add a simple text substitution parameter into I forgot we were doing stuff like {{etyl}} so that {{etyl|la|dial=Late}} would produces ]]? --Bequw → ¢ • τ 03:34, 28 September 2009 (UTC){{etyl:Late Latin}}. --Bequw → ¢ • τ 15:26, 5 October 2009 (UTC)
SIL just completed their review of 2009 change requests for ISO 639-3. They officially gave Latgalian {{tlg}}, which we had coded as lat-ltg. They retired (via splitting/merging/etc.) many codes, but most of them we hadn't used yet. For the following retired codes I've contacted all the editors that have used them to see if they can update the content.
{{lut}} was retired as it was a group name. Per SIL, "Lushootseed is a group name, not an individual language name. Existing languages in this subgroup of Salish languages are Southern Puget Sound Salish , Skagit , and Snohomish ". Usage: 1 translations{{mof}} was "split into Mohegan-Pequot and Narragansett ". Usage: {{rmr}} was "split into Caló and Erromintxela ". is a new identifier. Usage: {{sum}} was "split into Mayangna and Ulwa ". Usage: The only thing that may take some discussing/changing is what to do about Latvian (see WT:BP#Latvian). --Bequw → ¢ • τ 23:01, 21 January 2010 (UTC)
For languages that have both an ISO 639-1 and 639-3 code (eg French has {{fr}} and {{fra}}), would any have a problem if we redirected the -3 code to the -1 code? That would keep the language names the same and allow easy identification of the duplicate codes. They could still be subst'ed and used in places where the language name is used. Now that we scan for use erroneous, "literal" usages of the -3 code, I don't think this will encourage hard-to-deal-with errors. --Bequw → ¢ • τ 22:44, 23 January 2010 (UTC)