Re: Formal definition of HTML5 (was Re: Version information)

On Apr 14, 2007, at 09:49, Mike Schinkel wrote:

> Just curious: How is it possible to write validation software that  
> is free of bugs whereas isn't not possible to create a reference  
> implementation without said bugs?

Non-trivial software (TeX being the possible exception), in general,  
has bugs. Therefore, it is safe to assume that a conformance checker  
has at least one bug.

When the software is not a normative reference implementation, the  
bugs can be fixed without reissuing the spec when the spec prose and  
the software disagree.

-- 
Henri Sivonen
hsivonen@iki.fi
http://hsivonen.iki.fi/

Received on Saturday, 14 April 2007 11:08:19 UTC