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

Re: Versioning and html[5]

From: Maciej Stachowiak <mjs@apple.com>
Date: Sun, 15 Apr 2007 15:53:43 -0700
Message-Id: <C04FB065-2364-4585-B24C-A56C4BDD42BD@apple.com>
Cc: "public-html@w3.org" <public-html@w3.org>
To: Boris Zbarsky <bzbarsky@MIT.EDU>


On Apr 15, 2007, at 1:17 PM, Boris Zbarsky wrote:

> Maciej Stachowiak wrote:
>> Safari and Firefox handle sites with misnested forms in tables  
>> reasonably well, as far as I know, but make a proper DOM tree in  
>> these cases. I'd love to hear of sites where this is not true, in  
>> case we have some bug fixing to do.
>
> To be entirely honest, Gecko has had four bugs filed on issues  
> related to this in the last 8 years or so.  In all cases, the page  
> used innerHTML on the <form> or the table elements in question and  
> was surprised that the form had no kids or that setting innerHTML  
> on the <tr> in the following markup:
>
> <table>
>   <form>
>     <tr><td></td</tr>
>   </form>
> </table>
>
> did not add the form controls in the new HTML to the form.

It sounds to me like these issues could be addressed while still  
maintaining a tree-formed DOM, although innerHTML outputting  
something other than an actual reflection of the DOM tree would be a  
little ugly. Still, adjusting innerHTML in this way would actually be  
more round-trippable than making innerHTML reflect the true DOM.

I'm not sure if it is worth the effort if it is really very few  
sites; thanks for the heads-up though.

Regards,
Maciej
Received on Sunday, 15 April 2007 22:54:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:15:53 GMT