On Wed, Jun 26, 2013 at 9:17 AM, Hans Schmidt <z0idb...@gmx.de> wrote:
> No, transcriptions/transliterations usually cannot be done > automatically. There are so many irregularities or difficulties where it > is not possible. You also have to account for creating spaces when there > are none in the original example, the actual pronunciation is irregular > (especially in Japan). Why should it not be stored inside the > appropriate name tag? > I wouldn't tag street names that aren't signposted in name:*. Even the name=* tag, I'd go with the primary name, and include any other languages included in name:*=*. Regional examples that I can think of (though may not have actually implemented yet due to not having the time to deal with multilingual signage in areas relatively distant from me for the most part, particularly in areas of the Muscogee Nation and Creek Nation where either the TIGER import botched it because the Census worker has had little to no exposure and didn't copy the sign verbatim despite it being latin characters; or in Cherokee's case, no comprehension for the writing system and just made stuff up as they went): (from Okmulgee, Oklahoma) name=Noksvlke name:en=Bear name:mus=Noksvlke (though interestingly, my limited understanding is "nokose" would be the literal translation "bear," whereas "noksvlke" would be "bear people" or "bear tribe," but the English on the sign just says Bear; if I'm right on the translation (and correct me if I'm wrong), I'd be interested to know why the finer point got glossed over in the English name). (from Sapulpa, Oklahoma) name=Mission Street name:mus=Yugheeha Yusten name:en=Mission Street Renderers for specific languages can key on name:xx for a specific language otherwise default to name=*. _______________________________________________ josm-dev mailing list josm-dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/josm-dev
