Re: Comments on HTML WG face to face meetings in France Oct 08

Elliotte Harold wrote:
>> It's working OK for CSS so far, no?  It's working OK for HTML4 (modulo 
>> interop problems due to everything being undefined).
> 
> No, it isn't. CSS parsing is a colossal mess in which many stylesheets 
> depend on specific bugs in specific versions of specific browsers.

Hold on.  By "work OK" do you have some other criterion than "is used on 
the web, largely interoperably"?

> The inability of vendors to correctly implement the much simpler CSS 
> parsing rules makes me wonder whether we have any hope that they will 
> correctly implement the far more complex HTML 5 parsing algorithm.

Note that a lot of the "bugs" have to do with unimplemented features 
(and correct implementations of error recovery from there).  For the 
rest, most take advantage of places where the CSS parsing rules are so 
simple as to not fully specify behavior (something that's been worked on 
over the years).  I agree that HTML 5 should avoid this pitfall, and I 
think the state machine approach is a good technique for doing so.

-Boris

Received on Monday, 17 November 2008 15:54:25 UTC