[standards-jig] Re: [Council] i18n JEP

Peter Saint-Andre stpeter at jabber.org
Wed May 29 22:39:18 UTC 2002

Hi Max, this sounds like a good solution, did you have a chance to work on
this again?


Peter Saint-Andre
Jabber Software Foundation

On Sun, 12 May 2002, Max Horn wrote:

> Sorry for the extremly late response to this, but a) I didn't see 
> Jer's mail in the first place (stpeter told me about it recently 
> <g>), and b) I was swamped with other stuff.
> Jer's mail was on the council list, but I think a discussion of this 
> might be more fit for standards-jig - not sure on which list to 
> continue this best, though, if you think it rather should continune 
> on the council list, please tell me.
> At 16:38 Uhr -0500 15.04.2002, Jeremie wrote:
> >IMHO it's important to try and take this one in baby-steps, start doing a
> >little i18n so as to get others with more experience involved to tackle
> >the harder problems.
> Agreed.
> >
> >In doing that, what about tackling just one aspect in this JEP, the
> >expression of the language used within that element?  The JEP could be
> >limited to having a client send an xml:lang attrib on any
> >iq/message/presence it wanted to express a language preference for.  That
> >information could be used by any iq-responder (service) to select an
> >appropriate reply, or by a client receiving a message/presence to indicate
> >the language used (or offer the use of a translation service :).
> >
> >This would simplify the JEP significantly and solve DW's immediate
> >comments (leaving the rest, like querying and negotiation, for future work
> >based on these steps).  It is simply a recommendation to all clients and
> >services to express their language, and at this point it would really help
> >get the ball rolling towards building better i18n support.
> Indeed. This way, the I18N-JEP would indeed be rather short, but 
> that's not bad. We would lay the foundation for future I18N 
> improvements.
> In the future, we probably want some negotiation protocols for 
> prefered languages, which allow two entities to agree on a 
> communication language. E.g. I speak German and English, my prefered 
> language is german. Some service might be available in English and 
> French (with "available" I mean here that the texts in it, e.g. for 
> forms, are available in both languages). Thus, the logical decision 
> for the service should be to offer the english version to me. This 
> requires some way this information can be exchanged. Either the 
> service has to query us for the "supported" languages, or (probably 
> the better approach), the client will sent on each query a list of 
> favorite languages, in order, and the service would (if it supports 
> that feature) pick the first language from that list it supports.
> On the user side, this would equal to a preference for prefered 
> languages, just like most web browsers offfer it already.
> I imagine this could be done with a simple <x> protocol, attaching 
> the list of prefered languages to already existing <iq> messages; 
> e.g. one could send an iq:register request to a service, including a 
> list of the users prefered languages; existing services would just 
> ignore this list, and new ones could take advantage of it. OTOH, if 
> an old client makes contact to a new service (that is, it sends now 
> list of prefered languages), the service would just assume some 
> default language.
> I think of something like this:
> <iq type="get" id="1001" to="aim.denmark">
>    <query xmlns="jabber:iq:register"/>
>    <x xmlns="jabber:x:lang">
>      <item xml:lang="de-DE"/>
>      <item xml:lang="de-DE"/>
>    </x>
> </iq>
> If nobody objects, I will update the JEP in the couple of days 
> according to Jer's suggestion.
> Cheers,
> Max
> -- 
> -----------------------------------------------
> Max Horn
> Software Developer
> email: <mailto:max at quendi.de>
> phone: (+49) 6151-494890
> _______________________________________________
> Standards-JIG mailing list
> Standards-JIG at jabber.org
> http://mailman.jabber.org/listinfo/standards-jig

More information about the Standards mailing list