On Sun, 2011-03-06 at 10:51 +0100, Andras Timar wrote: > Hi, > > Do you know why 'be-BY' code was used in LibreOffice instead of > simpler form 'be' (Belarusian)?
We should follow the bcp47 guidelines as much as possible IMO, which boils down to keep them as short as possible, and only as long as necessary. So yeah, "be" would be good, sa-IN is another one that dubiously over-specified IMO. I don't thing there was a good reason for be-BY vs be. > Do you know why 'ns' was used instead of the standard (ISO 639-2) > 'nso' for Northern Sotho? I believe that the tooling at some stage somewhere wasn't able to deal with > 2 letter codes. But it definitely is by now > If I replace them, i.e. be-BY -> be and ns -> nso, can it break anything? In i18npool/source/iso/isolang.cxx we have sufficient magic to fix up e.g. documents which had "ns-ZA" tags in them, so that's ok anyway. You'll have to fix the helpcontent2/source/auxilliary/ns/*.cfg as well, the l10ntools/java/l10nconv/java/com/sun/star/tooling/languageResolver/lang.map as well I suppose and solenv/inc/postset.mk I'm not sure where, if there is/was a place, the truncation from nso to ns took place in the past to select the "ns" translations when the system locale was nso_ZA. Should be fairly easy to at least test that. C. _______________________________________________ LibreOffice mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/libreoffice
