pseudo-localization variants

Peter Constable petercon at microsoft.com
Tue Dec 18 05:18:45 CET 2012


A lot of companies use pseudo-localization methods for testing localizability in which a source string in the development (e.g., in English) is transformed into a different spelling that uses other characters, changes the length and perhaps other things to provide something that can still be read in the same language by the development team / localization engineers but that emulates some of the qualities that might arise when the strings are truly localized. For example, “Bing App custom protocol” might be pseudo-localized to “[HoKVu]《ßíηģ Дþφ čũşťøm рґöŧõςοŀ伪》” or (for bidi testing) to “حفظ(Bing App custom protocol!)שמור”.

Within MS, there’s been a bit of up and down history with how pseudo-loc content has been handled. When this was first done, all content got tagged using Windows numeric locale IDs (LCIDs), and there wasn’t one for pseudo, so the LCID for some language that Windows wasn’t localized into was used, which worked until the day came when Windows actually _was_ localized into that language. After some iterations, today a private-use language subtag “qps” is used. Sadly, none of the complete tags are valid BCP 47 tags, since non-registered variant subtags are used. (Even more sadly, the main one that’s most frequently used wouldn’t even be viable as a variant subtag since it’s only 4 letters.)

That’s made me wonder if it wouldn’t make sense to register one or more variant subtags for pseudo-localized content.


Thoughts?


Peter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.alvestrand.no/pipermail/ietf-languages/attachments/20121218/f579180b/attachment.html>


More information about the Ietf-languages mailing list