W3C home > Mailing lists > Public > www-international@w3.org > January to March 1997

Re: Language labelling (fwd)

From: M.T. Carrasco Benitez <carrasco@innet.lu>
Date: Mon, 24 Feb 1997 11:23:14 +0100 (MET)
To: WInter <www-international@w3.org>
Message-ID: <Pine.LNX.3.91.970224112207.19117F-100000@winter>
[I repost because my ISP was down and I do not know if it was posted]

---------- Forwarded message ----------
Date: Fri, 21 Feb 1997 18:09:22 +0100 (MET)
From: M.T. Carrasco Benitez <carrasco@innet.lu>
To: Francois Yergeau <yergeau@alis.com>
Cc: WInter <www-international@w3.org>, Misha Wolf <misha.wolf@reuters.com>,
    Martin Dürst <mduerst@ifi.unizh.ch>,
    Charles Wicksteed <charles.wicksteed@reuters.com>
Subject: Re: Language labelling

[François]> Why?  You have not given a single argument in favor of it!

I was under the impression that I had.  Lets try again.

We *agree* on the following:
  - Content-Language should be in the http header
  - Language label internal to the doc
  - Language label in only one place inside the doc
  - <meta http-equiv= ...> "may" be there and it is not not there
  - Optimization is private to the doc management system

We disagree on where to put the language label.  *One* of the following:
  - <html lang=xx>  
  - <meta http-equiv= ...>
  - (Any other suggestion ?)

As long as the language label is in *one* place only (monolingual 
docs) inside the doc and the server pick it up (optimization is 
private, hence I do not mind when the server does the actual 
picking) and the server transmit the Content-Language, I do not mind very 
much if it is one way or another as long a we all agree.

The reason for suggesting the <meta http-equiv= ...> because the stuff to be 
put in the RFC 822 type header are in these tags.

If the <meta http-equiv= ...> are there there should be used by the 
server; this is parallel to the question of transmting the char in the 
header: it is not done, but should be done. 

As I mentioned before, the question of language label is also important 
for the robots (altavista, etc).  A robot should be capable of 
indexing per language or look for docs in only one language.  It could 
be decided to do only a HEAD to check the language before doing a 
GET.  (Anybody from Altavista, Yahoo! or similar around ?).

Regards
Tomas
Received on Monday, 24 February 1997 05:10:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:16:46 GMT