Re: Extension Specification for XHTML5 entity definitions.

On 18/01/2013 13:17, Henri Sivonen wrote:
> Trivially easy but bad in terms of the Degrade Gracefully Design
> Principle (which might not matter in this case in the Rapid Release
> world).

There are some theoretical bad cases one could imagine but no cases as
bad as the spec being changed to break the existing documents using
these entities including the document that _defined_ them. (The MathML2
spec) Which in its XHTML form is a well formed document and was parsed
as such by any browser that could handle mathml at all until they
changed to match the unfortunately drafted xhtml parse rules in HTML5
drafts. So current browsers reject the entire document as not well formed.

That was a simple straight spec bug that broke existing content and it
should have been fixed when reported. Since it wasn't fixed then it
should be fixed now, better late than never.

However putting that history to one side, do I read your message
correctly that you wouldn't be totally against gecko implementing this
if it were to be added to the html spec? (Even if you think it's not the
most interesting or pressing topic that one could be discussing:-)


Received on Friday, 18 January 2013 15:21:39 UTC