Re: CoAP and new methods?

Mark Nottingham wrote:
> Well, they've got a separate registry for methods, so I'm not too concerned.

Right.  In the end, of course, we would be most happy with a solution
that maps cleanly into HTTP in a cross-protocol proxy.  (Consider this a
mild expression of interest in getting SEARCH to RFC.)

> I'd be a little happier if they didn't seem to justify everything they did in terms of HTTP.

We do have the use cases (e.g., resource directory for iPATCH, COMI/COOL
YANG-based management for all three new methods).  The text that points
to HTTP is mainly there to justify what we selected among some of the
particular bike-shed details one could come up with.  Now if your
experience indicates HTTP got those details wrong, and we should do it
in a different way, we would like to know that!

(Any other tidbits about the experience with integrating PATCH into HTTP
would also be most welcome -- RFC 5789 has been out for six years now;
time to advance it to STD?)

Grüße, Carsten

Received on Monday, 9 May 2016 13:56:36 UTC