Open Stack

On May 11, 2015, at 1:05 PM, Dean Troyer <dtroyer at gmail.com<mailto:dtroyer at gmail.com>> wrote:

On Mon, May 11, 2015 at 11:44 AM, Rosa, Andrea (HP Cloud Services) <andrea.rosa at hp.com<mailto:andrea.rosa at hp.com>> wrote:
> Agreed. Violating the HTTP spec is something that should be avoided.

Actually it is not violating the HTTP spec, from RFC:
"  A payload within a DELETE request message has no defined semantics;
   sending a payload body on a DELETE request might cause some existing
   implementations to reject the request."

The RFC prohibit the use of a body just for the TRACE:
" A client MUST NOT send a message body in a TRACE request."


When playing in undefined areas such as this it is best to keep in ming Jon Postel's RFC 1122 principle: "Be liberal in what you accept, and conservative in what you send".

I'll put it this way:  An RFC not prohibiting something does not make it a good idea.  This is not how we build a robust API that developers and user can easily adopt.

I’m agreed that this is not a good idea. I’d also invoke the principle of least astonishment here. Because it’s a de facto standard (so much so that some proxy in the middle may alter or reject such a request), it would be surprising to developers to allow such requests.

REST APIs are difficult enough. Let’s avoid things with "no defined semantics".

I’d comment on the review but Gerrit is having a fit so I filed a bug.

https://code.google.com/p/gerrit/issues/detail?id=3361

Everett

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150511/b50006a8/attachment.html>