Re: PaceErrVerb

[Henry Story <henry.story@xxxxxxxxxxx>]

Should the discussion about ERR not move now towards 
ietf-http-wg@xxxxxx?
Or am I meant to defend this position over there all alone?

It does not seem to make sense to me for one lonely API like atom to 
demand something as important as a new HTTP method all by itself. If 
so, should one not first convince the http-wg that it is important?

How did the WebDav folk do it? Did they just write those methods 
themselves?

Henry

On 24 Jun 2004, at 08:20, Asbjørn Ulsberg wrote:

> On Wed, 23 Jun 2004 08:14:41 -0600, Antone Roundy 
> <antone@xxxxxxxxxxxxxx> wrote:
>
> > Something like REPORT would work better...and might be useful for
> > not-yet-imagined things besides errors.
>
> I've thought about using ERR as a more general assertion method, and 
> not only to report errors. I'm not sure how that's useful, but in such 
> a scenario, we could add an HTTP status code with the assertion to the 
> resource, which then was the entity that told if this assertion was an 
> error or just an informative message.
>
> In such a case, we could use ASSERT, and even POST would work then. 
> The problem with POST is that it is abused, and that its inner 
> semantics is never fully understood, even if you're able to parse the 
> body. Hiding semantics inside a POST is imho a bad idea. ASSERT + 
> status code is a better idea then, imho.
>
> --
> Asbjørn Ulsberg         -=|=-        asbjornu@xxxxxxxxxxx
> «He's a loathsome offensive brute, yet I can't look away»