Re: ERROR verb

[Robert Sayre <mint@xxxxxxxxxxxxxxx>]

On 6/19/04 9:51 AM, "Henry Story" <henry.story@xxxxxxxxxxx> wrote:

> 
> This should not be an Atom only thing. The problem is larger than Atom.
> But if it is why not tackle it, instead of inventing another ad hoc
> Atom only solution?
> 

Solving problems for everyone is harder than solving problems for Atom.

I think someone should definitely write this up. It will be interesting to
contrast with PaceServiceError.

There are some differences here. One is that you're sending the error back
to the URI it was retrieved from. Another is that it doesn't work in the
absence of an HTTP response.

Also, I'm concerned that the definition of "error" could
creep--PaceServiceError is pretty specific. Julian's suggestion of
referencing  the error by URI is interesting.

When I first picked up this idea[1], I thought of making a new verb, too.
Watching people *freak out* about PUT and DELETE made me pretty skeptical it
would ever happen.


Robert Sayre


http://www.franklinmint.fm/blog/archives/000091.html