The Super Challenge - rules

[Serpent]

I think we need two-letter codes though? I know there are longer ones...
@rdearman would this just be for list purposes (ie sorted separately on the list of the languages you can sign up with, and then some additional information somewhere)? essentially, would I be using whatever Romance etc tags we come up with, or the tags for individual languages, which would all get processed according to your sistem? idk, unless we need a list of sub-tags, imo it's best to keep things flexible, and for example not be picky about the strictly genetic classification, instead taking transparency into account, so that Latin, Esperanto or Papiamento were counted as Romance (if the person prefers - btw these specific examples don't matter much to me). If possible, any list of languages should be "including but not limited to".

And well I assumed there's some kind of list/database that lets the bot know which tag #german and #deutsch are related to.

(Oops let's move to the technical subforum?)

[emk]

I think we need two-letter codes though? I know there are longer ones...

If you want to handle all the world's languages (not even counting language families), you need both two- and three-letter codes. There are currently 184 languages with two-letter ISO 639-1 codes, which isn't very many. The ISO 639-3 standard defines three-letter codes for 7,865 languages, and it's pretty comprehensive. ISO 639-5 includes the language families, if you need to code them for some reason. This whole issue has been nicely sorted out by various ISO committees and the SIL, and we should just reuse their work instead of trying to redo it from scratch.

One of my "fun" languages (ancient Egyptian) has the official code "egy", and there's no two-letter code for it. So I care about three-letter codes!

[rdearman]

I think we need two-letter codes though? I know there are longer ones...

If you want to handle all the world's languages (not even counting language families), you need both two- and three-letter codes. There are currently 184 languages with two-letter ISO 639-1 codes, which isn't very many. The ISO 639-3 standard defines three-letter codes for 7,865 languages, and it's pretty comprehensive. ISO 639-5 includes the language families, if you need to code them for some reason. This whole issue has been nicely sorted out by various ISO committees and the SIL, and we should just reuse their work instead of trying to redo it from scratch.

One of my "fun" languages (ancient Egyptian) has the official code "egy", and there's no two-letter code for it. So I care about three-letter codes!

OK, OK.... I have modified the DB tables to allow 3 character codes, and I even inserted "egy" after testing it as much as I could on my test system. I have forked the source code and I'll see if I can hack in the "If I tweet german, I really want you to count it in Germanic as well" option.

But I still don't want to do massive inserts. So what families are people actually going to study? Don't forget that right now, as far as the bot is concerned a language family is different from the language. So if you're studying "romance languages" then tweeting French page numbers doesn't count, and visa-versa. Until I manage to put a hack in (or someone smart steps in) then if you are going to register with a family, there isn't any point in registering for a family member.

[Serpent]

Well, I'd like to use the bot for Romance, Germanic, Slavic and Finno-Ugric, as I've said before. I don't know if anyone needs any other languages, but it would be fair to add Semitic, Turkic and possibly Sino-Tibetan (which would allow counting Mandarin and Cantonese together if one chooses). Also I guess this way it should be allowed to count Ancient and Modern Greek together?
Anything else can be added on request as needed, of course.

@emk, I kinda assumed there's a technical limitation that allows only two-character codes. If not, then it's obviously the best solution

[rdearman]

There was a technical issue. But i fixed it.

[daegga]

Well, I'd like to use the bot for Romance, Germanic, Slavic and Finno-Ugric, as I've said before. I don't know if anyone needs any other languages

+Scandinavian please

btw, the SC page isn't linked from the main site yet

[kanewai]

Romance for me too.

[rdearman]

Romance for me too.

Yeah, yeah, we all want romance... but what languages do you want?

[rdearman]

Well, I'd like to use the bot for Romance, Germanic, Slavic and Finno-Ugric, as I've said before. I don't know if anyone needs any other languages

+Scandinavian please

No such thing, I'm using the ISO standards, so you need to find what you want in this list: http://www.loc.gov/standards/iso639-5/id.php

btw, the SC page isn't linked from the main site yet

[/quote]
We are working on a challenge page which will link to all the various challenges. (I think someone volunteered anyway, I've tried to dodge that bullet)

[rdearman]

OK, I've added the following family groups.

• egx = Family-Egyptian
• fiu = Family-Finno-Ugrian
• gem = Family-Germanic
• grk = Family-Greek
• roa = Family-Romance
• sem = Family-Semitic
• sit = Family-Sino-Tibetan
• sla = Family-Slavic
• trk = Family-Turkic

Please note the "Family-" before these. I believe (yet to be tested live) you can tweet if you use the hashtag #Family-Finno-Ugrian for example. However it will work with the code: "fiu" (and use less characters in your tweets)

If anyone needs another let me know, and lets all hold hands and pray this doesn't do something horrible and unexpected to the bot.