[Pkg-sugar-devel] Bug#790693: lintian: Aymara language code is ayc (not ay - not yet, at least...)

[Jonas Smedegaard]

Package: lintian
Version: 2.5.32
Severity: normal
Tags: l10n

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Seems to me that language code "aym" should not be corrected to "ay" but
instead (for the time being" to "ayc".  Here's my reasoning:

Some of th sugar-*-activity packages trigger lintian warnings like this:

  W: sugar-read-activity: incorrect-locale-code aym -> ay

Long description of the lintian warning mentions that language codes are
those is ISO 639-1 and ISO 639-2, but does not say which wins if listed
in both but by different name.

Aymara is listed in ISO 639-1 as "ay" but in ISO 639-2 as "aym", but are
are both listed as scope "Macrolanguage" - i.e. what some refer to as
"Aymaran", the _family_ of Aymara languages.

Wikipedia article on Amayran references the individual languages as
"Southern Aymara" with code "ayc" and "Central Aymara" with code "ayr".

Only Amaya-related locale listed in /usr/share/i18n/SUPPORTED is ayc_PE
- - i.e. Southern Aymara in Peru.


As I understand it, renaming locale files to "ay" renders them unusable
on a standard Debian system (needing /usr/local/share/i18n/SUPPORTED
hack).  Only Amayan language code usable in standard Debian currently is
"ayc".

I am right now in Peru, discussing this issue with Sebastian Silva - a
local activist involved in deploying Sugar there.  As I (vaguely)
understand it, the ayc_PE slot was created few years ago to cover one of
the two Aymara dialects, whereas current translation efforts (involving
Sugarlabs, Peruvian ministry and OLPC) intends to cover both dialects as
a single language.  In the short term the only option I see is to
"abuse" the "ayc" slot for "ay/aym". Sebastian will try get confirmed if
above is correctly understood, and if so we should probably have the
slot "ay" (or "aym" - depending on which ISO standard has priority)
added and move dual-dialect localizations to that.

...Or instead have slot "ayr" created, if in fact current translation
covers only one of the dialects _and_ there is interest in localizations
also for the other.


Phew.  Hope that all made some sense...

 - Jonas

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
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=HxE6
-----END PGP SIGNATURE-----