Re: Proposal to Adopt HTML5

Hi Josef,

On Apr 11, 2007, at 3:14 AM, Josef Spillner wrote:

>
> Am Mittwoch, 11. April 2007 04:10 schrieben Sie:
>> HTML 4 is so underspecified it's a miracle it's even implemented.
>> And I am one of the few here (with Murray, Chris and Dan) who were
>> already here in the HTML WG for HTML 4.
>
> Has there ever been an analysis of why this happened? I would be  
> interested to
> know if the spec was already considered vague at the time of  
> finishing the
> draft, or if it was granting too much freedom to implementors.

I think at the time it was written, we had very different norms for  
what constitutes a good standard in the web standards community.  
Since then, I think the following norms have evolved:

* More preference for specified behavior over undefined or  
implementation-specific behavior, even though the latter gives more  
freedom to implementations.

* An expectation of clearly stated conformance requirements.

* Review of specifications with the idea that the conformance  
requirements will become testable assertions for a comprehensive test  
suite.

Given these changes in attitude, the web standards community today  
has much higher expectations for what makes a good spec.

Regards,
Maciej

Received on Wednesday, 11 April 2007 22:30:35 UTC