[lug] speaking of HTML validation and validator.w3.org

[rm at fabula.de]

On Mon, Jul 01, 2002 at 03:37:57PM -0600, D. Stimits wrote:

> > Hm... if it's complaining about entities, maybe you need to HTML-quote
> > the ampersands?
> > 
> >    mailto:foo at bar?subject=hi%20there&body=foo%20bar
> 
> I was ahead of you on that one, and tried it out yesterday. No go, it
> then does not believe that "body" is an entity, and the emailer then
> makes the entire body part of the subject. The html encoding has the
> same effect in this regard as does URL encoding...it neutralizes the
> control notation and turns it into an uninterpreted text character
> without special meaning. Perhaps I'll send an email to the w3.org site
> and ask about it, it sure seems like either the validator is incorrect,
> or else that RFC does not apply the the loose.dtd/transitional html 4.0.


Hmm. From my humble understandings of SGML i'd say that this has nothing
to do with DTDs at all. SGML entity resolving comes way before DTD validation
(and anyway: the location part of the 'mailto' is part of the attribute
_value_ of an href tag - DTD have no way to control the content of attributes
on such a fine level). This looks like a 'reader' problem to me. As far as i
understand the content of the href tag is supposed to be a URL and hence
needs to be URL encoded. But since it's also part of an HTML document it needs
to be HTML-encoded as well. Welcome to the wounderful world of SGML escape
mania ....

  Ralf Mattes
> D. Stimits, stimits at idcomm.com
> _______________________________________________
> Web Page:  http://lug.boulder.co.us
> Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug
> Join us on IRC: lug.boulder.co.us port=6667 channel=#colug