W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 2001

Re: lock tagging

From: Jason Crawford <ccjason@us.ibm.com>
Date: Tue, 13 Feb 2001 18:39:29 -0500
To: "Kevin Wiggen" <wiggs@xythos.com>
Cc: <w3c-dist-auth@w3.org>
Message-ID: <OF24B64B72.37777D82-ON852569F2.007F4369@pok.ibm.com>


I'm going to leave item (1) from Kevin's post for another thread.  I do
have an opinion on (2)...

<<
2)  /foo is a directory with no contents and is locked depth infinity
    PUT to /foo/bar, does this require a tagged list for the lock token
specifying /foo or will un untagged list work as well.
>>
My preference is for the server to ignore the URL provided with the lock
token in the tagged IF header.   Although I can dream up a case where it is
valuable to correctly specify at what URL a lock is located, I think for
the most part the server just needs to verify that the client has the lock
token for all the locks that come into play for the requested operation.

I might change my mind on the above statement in a few weeks after digging
though the spec again, but I think that if we disagree with what I said
above, we need to clearly define the scenarios where URL checking is
important and then clearly define the semantics of the URL checking to
match that.  But at the moment, I don't see URL verification as important.

Other opinions?

J.
Received on Tuesday, 13 February 2001 18:41:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:55 GMT