Non-Latin characters in URL

Greek, etc.

[jonte]

My Web page is being translated to Greek by a kind soul. However, I’m having doubts on how to name URLs on the site. I mean, if the English page refers to:

http://www.example.com/earth.html

and the Greek translation of earth is χώμα (these are supposed to be Greek letters), then the Greek page would refer to:

http://www.example.com/χώμα.html

in order to optimize search engine indexing (in my opinion).

Now, I’m really not very keen on using characters different than a-z0-9_ in URLs (Unicode, etc.), even though in some languages maybe it is what you really do. Or perhaps Greek pages use Latin equivalence of the words appearing in the URLs?

Now, does anyone know if there are any standards on how to design URLs containing non-Latin characters in general?

Thanks
Jon

[edited by: encyclo at 9:58 am (utc) on May 6, 2006]
[edit reason] switched to example.com [/edit]

[kaled]

Why not use http://www.mydomain.com/greek/earth.html?

Or even consider

http://greek.mydomain.com/earth.html?

Using non-english characters is urls doesn't sound like a good idea to me.

Kaled.

[encyclo]

Here is a recent thread about non-ASCII characters in file names which might help you:

Accessibility issues with foreign characters in folder names [webmasterworld.com]

In short, only ASCII characters are truly "safe", other characters within the ISO-8859-1 (western European) range can be used but escaped (such as

%2F

), but anything else is not safe for URL use.