- From: Anne van Kesteren <annevk@opera.com>
- Date: Wed, 02 May 2007 10:34:48 +0200
- To: "Gareth Hay" <gazhay@gmail.com>, "David Hyatt" <hyatt@apple.com>
- Cc: "HTML WG" <public-html@w3.org>
On Wed, 02 May 2007 10:23:30 +0200, Anne van Kesteren <annevk@opera.com> wrote: > This doesn't take into account HTML content created from script. It's > also unclear when you would show an error and how you would handle > extensibility in this case. The other assumption "just show an error" makes is that you assume perfect implementations. I suppose this may be reasonable for XML parsers (though I'm not sure how many are actually conforming in the face of RFC 30323 and character encoding stuff), but HTML is already much more complex. (Let alone CSS.) Also, implementation experience with previous iterations of HTML should make it clear that this assumption is incorrect. (Besides of course the many other arguments against this idea as not wanting to show error messages to the user, etc.) To be honest, if you like that approach, XHTML2 seems more appropriate. -- Anne van Kesteren <http://annevankesteren.nl/> <http://www.opera.com/>
Received on Wednesday, 2 May 2007 08:35:09 UTC