Re: New feature negotiation syntax

Koen Holtman wrote:
> 
> Larry Masinter:
> >
>   [Koen Holtman:]
> >> Problem: The HTTP/1.x negotiation infrastructure (Accept headers,
> >>   User-Agent header, Vary) is not good enough.
> >
> >This problem description is not good enough.
> 
> I happen to think it is good enough.  Sure, it begs the questions `how
> do we know this is a problem' and `what can we do to solve it', but
> these are derived questions, they are not the core problem.  And I'm
> perfectly willing to repeat the answers to these questions.

I *really* don't want to butt in here, but, ...

A good requirements document states the problem and a set of tests that
would verify that the problem has been solved.  I think Ted Hardie made
a start on those two items, whereas I don't think Koen's statement does
either.  And I am guessing that that was Larry's objection, too.

Dave Kristol

Received on Thursday, 5 June 1997 06:30:51 UTC