Re: IDN - RTL

At 03:30 05/02/22, Najib Tounsi wrote:

 >Indeed, the consequence is that
 >"ABC.DEF.ma" is to  display
 >"CBA.FED.ma" (actually displayed "FED.CBA.ma")

The later display ("FED.CBA.ma") is correct. For examples,
please see http://www.w3.org/International/iri-edit/BidiExamples.

 >and
 >"www.ABC1.2DEF.ma" is to display
 >"www.1CBA.FED2.ma  (actually displayed "www.FED1.2CBA" (in stead of 
"www.FED2.1CBA.ma"))

These examples are (currently) illegal, so discussing how
they should be displayed is not very productive.

 >>Isn't the whole point that they just use their normal company names, as
 >>regular text, and they display the same as normal, like regular text?

More or less. But please note that this isn't actually the
case for English/Latin names either. In some cases, it works
out exactly; in others, it doesn't. The domain name system
may give the impression that it accomodates arbitrary names,
but it doesn't exactly do that in any and all cases.


 >I don't know what companies exactly want. I think if my cie name was 'ABD 
DEF', I'd likely be led to register it as "ABC-DEF" (and not "ABD.DEF"). 
And then  type "http://www.FED-CBA.ma" to see "http://www.ABC-DEF.ma".

If FED-CBA is the logical representation, and ABC-DEF the visual
one, that's fine. It would be totally wrong the other way round.

 >Note that I am already forced to type "http://" (though it's not the same 
problem).

Yes. It's not yet clear how that problem will get solved.
Maybe browsers will offer a little drop-down menu in the
language of the user interface, maybe things will work
out more or less because everything is mostly HTTP anyway
(on billboards,... one rarely if ever sees ftp://, and
usually also http:// is just left off), or maybe there
will be a new standards effort.

However, the issue at hand in this case is quite different
in various ways. Different domain names in different scripts
are just different domains. Different URI schemes in different
scripts would identify the same URI scheme. This creates
quite different issues for interoperability.

Regards,    Martin. 

Received on Wednesday, 23 February 2005 09:48:54 UTC