RE: Some problems with the WebDAV protocol

At 01:08 4/20/99 -0700, Greg Stein wrote:
>On Tue, 20 Apr 1999, Henrik Frystyk Nielsen wrote:
>> >b) If this scenario is a problem for DAV, it seems to me it's also a
problem
>> >for downlevel HTTP/1.1 clients too, since what you're alluding to is a
>> >general problem having to do with the implications a restriction on one
end
>> >of a containment relationship has on the other end.
>> 
>> No, as HTTP/1.1 doesn't have the requirement that "all ancestors MUST
>> already exist" then it can create the resource
>> "http://example.com/foo/fuzz/bar.html" just fine without caring about
>> whether /foo and /foo/fuzz exist or not.
>
>But HTTP/1.1 also does not require that servers create those intermediate
>collections. Your point is bogus... you're simply relying on some
>precedent rather than the specification.

I don't follow what you are saying here. 

>As a server author myself, I have stated that my response to PUT will
>return an error if you PUT to a collection that doesn't exist. Forget DAV
>-- that is my statement for PUT itself. I don't believe you have a basis
>to tell me that my response is incorrect.

You can do whatever you want but that doesn't mean that everybody else
thinks that this is the right thing. The purpose of a specification is to
allow people to interoperate without inflicting "sanity rules" of whatever
you feel is the right thing to do at the moment.

>Ergo, why is this issue w.r.t. PUT being argued? Server authors are free
>to return an error in this scenario.

which is fine

>(DAV simply states they must.)

which is broken as I have pointed out. The notion of whether a resource
exists or not depends on who is looking. The example that I gave leaves the
client no means of knowing whether the server complied with the
specification or not.

Henrik

--
Henrik Frystyk Nielsen,
World Wide Web Consortium
http://www.w3.org/People/Frystyk

Received on Wednesday, 21 April 1999 11:41:57 UTC