- From: Sierk Bornemann <sierkb@gmx.de>
- Date: Thu, 26 Apr 2007 12:25:30 +0200
- To: Shane McCarron <shane@aptest.com>
- Cc: "www-validator@w3.org Community" <www-validator@w3.org>
Am 26.04.2007 um 04:20 schrieb Shane McCarron: > > olivier Thereaux wrote: >> Are you begging to differ on the wording alone (on which I stand >> corrected) or are you also suggesting that the validator should >> not give a warning for not respecting the SHOULD recommendation in >> XHTML 1.1's conformance section? What is the XHTML's working group >> stance on the matter? > > Personally, I feel the validator should not raise an error if the > content type differs. [..] > It could raise a warning, I suppose. That's exactly, what validator 0.8. beta does. Validator 0.8 beta raises a warning, it doesn't raise an error. And that's very OK, I can live with it, because it follows the spec rules. BUT: My webserver *is* able to serve "application/xhtml+xml", the validator *would* get this mimetype *if* the validator would send an accept header like any other client out there. Because validator 0.8 beta doesn't send an accept header to the webserver, the webserver doesn't know, that the validator also accepts "application/xhtml+xml". So validator 0.8 beta receives "text/html". Because my rewrite rule doesn't catch due to a lacking accept header of validator 0.8 beta. Maybe the problem is solved on my side, if I reverse the mod_rewrite rule concerning the mimetype and only rewrite to "text/html", if the accept header of the client does *not* have "application/xhtml+xml" in its string. I have to play around with this a little bit. The easist and in my eyes the user-friendliest solution would be, if the validator 0.8 beta simply would send an accept header and would behave in this manner like a common client. Sierk -- Sierk Bornemann ICQ: 221105136 email: sierkb@gmx.de WWW: http://sierkbornemann.de/
Received on Thursday, 26 April 2007 10:25:45 UTC