W3C home > Mailing lists > Public > public-bpwg@w3.org > March 2008

Re: Statistics on mobileOK Basic

From: Dominique Hazael-Massieux <dom@w3.org>
Date: Thu, 06 Mar 2008 09:19:59 +0100
To: Sean Owen <srowen@google.com>
Cc: public-bpwg@w3.org
Message-Id: <1204791599.15448.86.camel@localhost>

Le mercredi 05 mars 2008 à 12:33 -0500, Sean Owen a écrit :
> What I am
> referring to is the fact that XHTML imposes more in its schema/DTD
> beyond well-formedness, like, "<span> must appear in <p>" or things
> like that. Could some of that have been removed, and would it have
> hurt its goals, and would that have left more valid documents out
> there, and would that have been a net win?

I think this is a complex question; in my mind, there are several types
of errors that DTD validation spots:
 * errors that really need to be highlighted: e.g. if you put a link <a
href> inside another link, you'll probably want to know that this is
broekn
 * errors that could be important: if you spell your href attribute
jref, your link won't work; if you add a "lang" attribute in a document
that only allows xml:lang, this is much less important
 * errors that are hard to care about: why is <ul> forbidden in <p>? why
can't <input> be a direct child of <form>?

Clearly one of the limitations of DTD validation (and that maybe
unfortunately we directly import in mobileOK) is that there is only one
level of granularity: any error makes you fail validation, there is no
notion of warning.

Dom
Received on Thursday, 6 March 2008 08:21:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 6 March 2008 08:21:20 GMT