[OSM-talk] name tags on place=country and how they're rendered on lowzoom

andrzej zaborowski balrogg at gmail.com
Mon Jan 19 16:01:17 GMT 2009


2009/1/19 Hakan Tandogan <hakan at gurkensalat.com>:
>
> On Thu, January 15, 2009 14:31, Stefan Baebler wrote:
>>>> where is a local language being set for a country or a region?
>>> It isn't at all. The software doesn't know which language "name" is in,
>>> only the mapper does.
>> Software can only gues the language of the default name from an
>> identical name value with a specified language.
>>
>> eg:
>> name=Venezia name:en=Venice
>> name:sl=Benetke
>> name:de=Venedig
>> ...
>> can additonally be tagged with name:it=Venezia
>>
>>
>> This could be used to make nice maps of default languages, if only it
>> was used more than just on major cities (smaller places rarely have foreign
>> names)
>>
>> With just one tag precisely describing the name (with language) all
>> other translations could be pulled from wikipedia (enriching maps with
>> additional languages) when needed eg such precise tag could be
>> wikipedia:en=Venice

I was wondering about links to wikipedia in national languages and
asked about it on IRC recently and settled on tagging with
wikipedia=en:Venice rather than wikipedia:en=Venice (actually I use
the form wikipedia=<language>:Page_Title only for non-english articles
and wikipedia=Page_Title if English is available).  I try to tag not
only major place names but also anything else that has a page.

Using "wikipedia=" instead of "wikipedia:<language>=" is good because
it strongly suggests that there is only one such tag per object
instead of many and avoids having conflicting translations with those
in wikipedia.

>
> This is one reason why I add the geonames ID of country objects to the
> node. If necessary, a renderer could pull additional names from geonames,
> if / when they become available, without having to go to Wikipedia and
> parse the information found there.

Slightly off-topic, the recently added low-zoom country map display
(mapnik) apparently pulls data from outside OSM database.  I spotted a
typo (or what I think is a typo) and wanted to correct it and found
that the name with a typo was nowhere in the database.  I think the
renderer should only pull data from sources that we can fix.  Where
does the data come from?

Cheers




More information about the talk mailing list