RE: Internationalization Requirements

It is not enough to say "Here are the different languages this document
is available in." You must also tie that into all the other axis's of
variation including CPU needed to execute, software required to read,
format, encoding, etc. We do the world a disservice if we try to solve
language variation independently of general variation. 

In addition, given the complexity of the issues and the varying
requirements from supporting variation on a single axis to supporting
variation on multiple independent axis's, I do not believe we are ready
to come out with a coherent standard. 

Variation is something the net does not do particular well. I think the
HTTP standard's attempts to deal with the problem have demonstrated
that. I can only speak from my personal experience trying to implement
HTTP on IE 4.0, a browser which is internationalized into some large
number of languages, but we were forced to cut support for the Vary
header and quality numbers from the get go. I think this problem needs
to age more, better consensus as to what the key features are need to be
developed, and then we can discuss standardization.

Standards never lead, they only follow.

		Yaron

> -----Original Message-----
> From:	Dylan Barrell [SMTP:dbarrell@bb.opentext.com]
> Sent:	Monday, August 04, 1997 1:04 AM
> To:	'Roy T. Fielding'; Martin J. Duerst; Yaron Goland
> Cc:	Dylan Barrell; w3c-dist-auth@w3.org
> Subject:	RE: Internationalization Requirements 
> 
> I agree that it would be too difficult to handle variants in general
> however language variants are a special case which would be easy to
> handle. There is always a one-to-one relationship between any given
> resource and its corresponding translation to another given language
> and the server is currently required to understand this relationship
> in order to handle the accept-language header correctly.
> 
> Cheers
> Dylan
> 
> ----------
> From: 	Yaron Goland[SMTP:yarong@microsoft.com]
> Sent: 	Mittwoch, 30. Juli 1997 21:17
> To: 	'Roy T. Fielding'; Martin J. Duerst
> Cc: 	Dylan Barrell; w3c-dist-auth@w3.org
> Subject: 	RE: Internationalization Requirements 
> 
> Um.. not that I'm aware. We decided very early on to not deal with
> variants. It brought up too many ugly issues. In fact, the problem
> most
> likely would require the introduction of a Turing complete scripting
> language. Even TCN is not powerful enough to capture all the possible
> issues.
> 
> Either way it was felt that we got a big enough win just solving the
> non-variant scenarios that it was worth achieving that goal rather
> than
> being hopelessly lost in an unending morass of negotiation
> specification
> issues.
> 
> 			Yaron
> 
> > -----Original Message-----
> > From:	Roy T. Fielding [SMTP:fielding@kiwi.ics.uci.edu]
> > Sent:	Saturday, July 26, 1997 9:21 PM
> > To:	Martin J. Duerst
> > Cc:	Dylan Barrell; w3c-dist-auth@w3.org
> > Subject:	Re: Internationalization Requirements 
> > 
> > >> Correct me if I'm wrong and show me where I can read about it,
> > >> but the server *doesn't* give you the variants list and certainly
> > >> doesn't map them to the language and this is exactly what I'm
> > >> complaining about.
> > >
> > >You may be right. I had a look at the HTTP 1.1 spec, and I only
> > >found 10.4.7  406 Not Acceptable. The behaviour described there
> > >is not very deterministic. Maybe Roy can help out?
> > 
> > That is why we have a WebDAV working group.  Both the 300 and 406
> > response bodies were left unspecified because the intention was that
> > they be specified by a group that actually had time to study the
> > problem in detail and come up with a [hopefully] better solution
> > than some off-the-cuff invention of mine.  It was one of the WebDAV
> > to-do items, last time I checked.
> > 
> > ....Roy
> 
>

Received on Monday, 4 August 1997 14:49:04 UTC