W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2005

Re: Is validity the real issue?

From: Maurizio Boscarol <maurizio@usabile.it>
Date: Sun, 06 Nov 2005 21:53:08 +0100
Message-ID: <436E6D34.1030007@usabile.it>
To: michele@diodati.org, w3c-wai-gl@w3.org

Michele, I agree with your premises, but I can't understand your solution.

For my point of view, application/xhtml+xml mime type decrease the 
chance of a page to be rendered: it will be rendered only if prefectly 
well-formed.

At the moment, a lot of pages won't be rendered. So I can't understand 
how this would help accessibility. At the opposite, I'm evaluating the 
idea that wcag 2.0 should reccomend to use text/html mime type (and 
backward compatibility guidelines of Apendix C) for any page whose you 
are unsure you can grant the present and future validity. You should 
only use application/xhtml+xml mime type when you're absolutely sure 
that nothing can go wrong. That's rarely the case.

As you know, I personally think that something is better than nothing. 
:) So, the XML specification that say that UA shouldn't attempt to 
render invalid pages are wrong and against accessibility, because 
decrease the chance of someone accessing something. But this is a 
different topic: I just wanted to know if I understand your proposal.

Maurizio

Michele Diodati wrote:

>I would propose to the group to examine another point of view about
>validity issue.
>
>It seems to me that the real problem with requesting validity to
>developers is the ability of browsers to compensate for lacking of
>validity: they succeed in rendering even a web page with thousand of
>errors within. Though there are many good reasons to publish pages
>without (X)HTML errors, no one of these is decisive. A "&nbsp" at the
>end of a page is enough to get invalid code; at the same time, today
>the front page of The New York Times <http://www.nytimes.com/> has 17
>errors and I can nevertheless read it.
>
>I think that requesting validity as a basic requisite for
>accessibility need something stronger than a vague possibility that
>something goes wrong with an AT. As developers, we need such a
>guideline that, if we do not apply it, the effect is immediately
>visible and unequivocal. So, if we think validity is indispensable, a
>solution could be to request, for every web page published, a
>content-type of application/xhtml+xml. In that case, validity is
>actually necessary: no validity no rendering of the page. (To address
>old browsers incapability with a content-type of application/xhtml+xml
>is always possible to create a server side switch for serving a
>content-type of text/html to all the old user agents.)
>
>Best regards,
>Michele Diodati
>--
>
>----------------------------------
>M i c h e l e    D i o d a t i
>Via Pian due Torri 86 - 00146 Roma
>Tel. 06 5503533 - Fax 06 233212132
>http://www.diodati.org
>----------------------------------
>
>
>
>
>  
>
Received on Sunday, 6 November 2005 20:49:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:40 GMT