[OSM-talk] Languages

[Stephan Plepelits]

On Tue, May 05, 2009 at 08:45:26PM +0100, Emilie Laffray wrote:
> I was actually thinking about anything that carries a name.
> In the example you have given me, you have partially answer the question
> that I was asking: name is expressed in the local language. If you want
> to add translation, you need to append :isolanguage to the named field.
> > What might be interesting and worth a discussion: A tag to describe the
> > default language of this object, e.g. "language=en". This could als be
> > several tags, e.g.
> > name=België - Belgique - Belgien
> > name:nl=België
> > name:fr=Belgique
> > name:da=Belgien
> > (and some more)
> > language=nl;fr;da    <- this would be new
> > (I hope I got the languages right, sorry if not)
> >
> > Comments?
> >   
> I like the idea of the language element. I would like to add an extra
> precision in this case. I think the order of language should be by
> importance.
Great to have a supporter :) I think the people who enter the name and the
language-tag to the database should decide, because they know the political
situation. As a general rule the importance of languages sounds reasonable
to me.

I even had a second idea. When you look at Google Maps you see that there's
an English translation (for places which have an English translation) in a
second row. This is something that I also applied to the
OpenStreetBrowser[1]. I thought it might be interesting to apply a tag
"second_languages" to an element (or to an administrative boundary), so
that the application can decide which languages should be put in the second
row. And if this languages is already part of the "main-name" it will be
ignored. This could even be inherited from the next-higher administrative
boundary.

Examples:
name=Sibiu
name:de=Hermannstadt
language=ro
second_languages=en;de (in the administrative boundary)
-> FIRST: Sibiu SECOND: Hermannstadt

name=België - Belgique - Belgien
name:nl=België
name:fr=Belgique
name:da=Belgien
name:en=Belgium
(and some more)
language=nl;fr;da    <- this would be new
second_languages=en;da (in the administrative boundary)
-> FIRST: België - Belgique - Belgien SECOND: Belgium
(da is already part of the "main" name, so it will not be put in second row)

Just a few thoughts ...

> Things like different alphabet might be also interesting to look at even
> if it is likely that it can be subsumed under translation.
Yes. I was thinking if it wouldn't be better to not use the english
translation as a second row, but rather the "int_name"? or the name
rewritten in latin letters? Could make a difference ... Ideas?

[1] http://www.openstreetbrowser.org

greetings,
        Stephan
-- 
Seid unbequem, seid Sand, nicht Öl im Getriebe der Welt! - Günther Eich
,---------------------------------------------------------------------.
| Stephan Plepelits,                                                  |
| Technische Universität Wien   -    Studium Informatik & Raumplanung |
| > openstreetbrowser.org > couchsurfing.com > tubasis.at > bl.mud.at |
| skunk at xover.htu.tuwien.ac.at       -       My Blog: http://plepe.at |
`---------------------------------------------------------------------'