W3C home > Mailing lists > Public > public-html@w3.org > June 2007

Re: Raising issues in a way that the editors will pay attention to them

From: Ian Hickson <ian@hixie.ch>
Date: Mon, 4 Jun 2007 20:55:05 +0000 (UTC)
To: Henrik Dvergsdal <henrik.dvergsdal@hibo.no>
Cc: HTML WG <public-html@w3.org>
Message-ID: <Pine.LNX.4.64.0706042046320.13385@dhalsim.dreamhost.com>

On Mon, 4 Jun 2007, Henrik Dvergsdal wrote:
> On 1 Jun 2007, at 02:40, Ian Hickson wrote:
> > 
> > * Include links to relevant research on the wiki page. That could be:
> > 
> >    * Links to pages that are working around the lack of the feature being
> >      proposed.
> > 
> >    * Surveys (even of a few dozen sites) showing authoring practices, so
> >      that we can determine authoring patterns around the topic. (I might
> >      take such surveys to greater lengths if possible and useful by
> >      running similar types of scans at Google.)
> > 
> >    * Test cases showing what existing browsers do.
> > 
> >   Making proposals with no research is another good way to lose
> >   credibility fast.
> 
> This requirement conforms very well to design principles such as 
> "Support Existing Content", "Don't Reinvent The Wheel", "Pave The 
> Cowpaths" etc. However, it effectively blocks out *novel* proposals that 
> may be related to principles such as "Solve Real Problems", "Media 
> Independence", "Universal Access" etc. Do we really need to be this 
> conservative?

Take <datagrid>, which is a very novel proposal (in HTML terms) in the 
spec now. For that proposal, the research could take the form of:

 * People want data grids so much that there are entires businesses built 
   around the concept of providing custom controls for them:

      http://www.activewidgets.com/grid/
      http://www.treeview.net/

 * Big sites are working around the lack of such a feature. Such sites 
   include the new Yahoo! Mail and GMail, with millions of users.

 * Demand hasn't been high enough for browser vendors to do it themselves, 
   though.

 * Few authors are hand-authoring such grids.

 * Fallback: UAs ignore <datagrid> today, and render the contents. Thus we 
   could provide backwards-compatible fallback by having a <datagrid> 
   element contain the data in the form of a <table> or other structure.

Hopefully this clarifies that "Research" doesn't necessarily bias us 
against novel proposals.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Monday, 4 June 2007 20:55:22 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:45 UTC