- From: Leif Halvard Silli <lhs@malform.no>
- Date: Tue, 19 Aug 2008 13:18:27 +0200
- To: Richard Ishida <ishida@w3.org>
- CC: 'Ian Hickson' <ian@hixie.ch>, 'HTMLWG' <public-html@w3.org>, www-international@w3.org, hsivonen@iki.fi
Richard Ishida 2008-08-18 16.48: > Leif, I'm not sure that makes sense... The prompt for the content > negotiation should come from the user agent, since it reflects the > preferences of the user, not the language of a given page. The user agent > sends the user's preferences (or its defaults) via the Accept-Language HTTP > field, and the server uses that information to do content negotiation (if > enabled on the server). If the server finds a match, it returns meta > information about the document it is serving in the Content-Language field > of the HTTP header. I did not intend to propose a change in anything. I just wondered whether it could be tested whether servers actually read the META elements in order to present things as you said. The trouble is that if you have 'page-english.html' and 'page-german.html', and both of them contain <meta ... content="en, de">, then such a negotiaton would not work. [...] >> Richard Ishida 2008-08-15 21.42: >>>> From: Henri Sivonen [mailto:hsivonen@iki.fi] >> >>>> What purpose does metadata serve if it isn't actionable? >>> >>> Metadata is actionable if some application is written to >>> use it. It is not actionable if the information is not >>> available. >> >> Regarding the question of "actionable": Ideally, authors should be >> able to add content-language information via the META tag, and >> then experience that the web server - and the Web browser - use >> this information to perform language negotiation. >> >> Richard, you made many tests of how UAs react to language tagging: >> Perhaps it is possible to make test case for what web servers and >> browsers do with the content-langauage information with regard to >> content negotiation -- leif halvard silli
Received on Tuesday, 19 August 2008 11:19:17 UTC