W3C home > Mailing lists > Public > www-forms-editor@w3.org > March 2005

Re: Compatibility between HTML, XForms, and WSDL

From: Eric Prud'hommeaux <eric@w3.org>
Date: Mon, 21 Mar 2005 04:32:27 -0500
To: Ian Hickson <ian@hixie.ch>
Cc: www-html@w3.org, www-forms-editor@w3.org, public-ws-desc-comments@w3.org, gerald@w3.org
Message-ID: <20050321093227.GA13015@w3.org>
On Mon, Mar 21, 2005 at 02:10:11AM +0000, Ian Hickson wrote:
> On Thu, 24 Feb 2005, Eric Prud'hommeaux wrote:
> >
> > HTML4.01 (and probably XHTML, though I didn't find it) specifies that
> > a GET URI is constructed as such:
> >   action + '?' + url-encoded form parameters
> > 
> > An cursory implementation survey showed that Lynx 2.8.4, Opera 7.5.4,
> > Mozilla 1.7.5 all appear to chop off the '?' and everything after it.
> >   <http://q.example/what?inference=owl-lite&q=SELECT%20%3>
> > 
> > WSDL [WZ] follows XForms's [XF] example and specifies that the
> > separator between the action and the parameters be a '?' if there is
> > not one already, otherwise, a '&' .
> >   <http://q.example/what?foo&inference=owl-lite&q=SELECT%20%3>
> > 
> > I propose an errata to HTML to reflect either current practice or the
> > XForms way of doing it. I will propose that the SPARQL protocol [SP]
> > use the XForms approach as well.
> 
> One advantage of the current practice is that if the action is "" then you 
> don't end up adding more and more arguments:
> 
>    <form action="">
>     <p><input type="submit" name="test" value="test"></p>
>    </form>
> 
> The first click would return
> 
>    http://www.example.com/q?test=test
> 
> With the current practice of HTML UAs, a second click returns the same 
> thing. With the other proposal, you would get:
> 
>    http://www.example.com/q?test=test&test=test
> 
> ...then:
> 
>    http://www.example.com/q?test=test&test=test&test=test
> 
> ...and so forth.

Good point, that is an advantage. But what do we do about it? It
violates both the HTML spec and the XForms spec. What I'd like to see
is a uniform policy between HTMLs and XForms, perhaps as an errata to
one or both.  What policy they choose is largely immaterial to me;
they all have their strengths and weaknesses.

What path do you think best?
-- 
-eric

office: +81.466.49.1170 W3C, Keio Research Institute at SFC,
                        Shonan Fujisawa Campus, Keio University,
                        5322 Endo, Fujisawa, Kanagawa 252-8520
                        JAPAN
        +1.617.258.5741 NE43-344, MIT, Cambridge, MA 02144 USA
cell:   +81.90.6533.3882

(eric@w3.org)
Feel free to forward this message to any list for any purpose other than
email address distribution.

Received on Monday, 21 March 2005 09:52:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 10 June 2009 18:12:13 GMT