W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > October to December 2006

Fwd: REBIND a parent to a child

From: Chetan Reddy <chetanreddy@gmail.com>
Date: Wed, 25 Oct 2006 10:40:08 -0400
Message-ID: <125bae360610250740r39bb1ea3m5c4997fecaf5a521@mail.gmail.com>
To: w3c-dist-auth@w3.org
I wrote my question in a 'Reply' instead of 'Reply to all'. Forwarding
Julian's reply to the list.

---------- Forwarded message ----------
From: Julian Reschke <julian.reschke@gmx.de>
Date: Oct 25, 2006 6:08 AM
Subject: Re: REBIND a parent to a child
To: Chetan Reddy <chetanreddy@gmail.com>
Cc: Geoffrey M Clemm <geoffrey.clemm@us.ibm.com>

Chetan Reddy schrieb:
> The last three mails have confused me a bit.
> A request to rebind /a to /a/b:
> Request-URI    /a (identifies a collection resource R1)
> segment          b
> href                 /a
>
> case 1: /c is also bound to R1
> case 2: /a is the only binding to R1
>
> I think case 1 should succeed(with /c and /c/b both pointing to R1 and
> /a not existing anymore) and case 2 should fail.

I would agree that if the request succeeds in case 1, that's what the
outcome should be.

I'm not sure whether case 2 needs to fail, but that's certainly what I
would do.

> In general, if the collection identified by the Request-URI is not
> reachable from root after removing the  old binding(href), the server
> should fail the request.
> Is that correct?

I'm not sure it needs to, but it's certainly a plausible approach.

Best regards, Julian

(any reason why you took this off-list?)
Received on Wednesday, 25 October 2006 14:40:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:44:15 GMT