[OpenStreetMap Serbia] Oznake

Miloš Komarčević kmilos at gmail.com
Fri Feb 24 17:13:55 GMT 2012


On Fri, Feb 24, 2012 at 4:50 PM, Thomas <toba0211 at yahoo.de> wrote:
> The language code for Serbian is sr according to the ISO Code used by OSM.

Yes, but only ISO 639 though, which is not enough.

> If someone wants to deviate from the recommendation, he could use
> name=whatever
> name:sr=cyrllic [optional if name=cyrillic]
> name:sr_lat=latin.

Baaaaaaaaad.

Once again, please see the links I posted to Mark Davis' BCP 47 discussion:

http://icu-project.org/repos/icu/icuhtml/trunk/design/language_code_issues.html

OSM should really implement BCP 47, not only one ISO tag, and than do
kludges with non-standard arbitrary tags like until now.

So you'd have support for sr-Cyrl, sr-Latn, zh-Hant, zh-Hans, etc.

OSM should then have transliteration rules from one script to the
other for each language.

Mappers shouldn't input this data multiple times, it should be generated.

It's great news the grant is there, as it's a good opportunity to
something about OSM i18n.

Regards,
M



More information about the Talk-rs mailing list