- From: Clemm, Geoff <gclemm@rational.com>
- Date: Sun, 28 Jul 2002 08:59:39 -0400
- To: ietf-dav-versioning@w3.org
The only behavior that seems to me to be consistent with the standard is (a), rejecting the request. The only way that CHECKOUT can succeed on a version that has DAV:checkout-fork as DAV:discouraged, and already has a successor, would be if the CHECKOUT had a DAV:fork-ok argument, and there is no way for a versioning-unaware client to add a DAV:fork-ok argument to the auto-checkout. Cheers, Geoff -----Original Message----- From: Julian Reschke [mailto:julian.reschke@greenbytes.de] Sent: Sunday, July 28, 2002 4:37 AM To: ietf-dav-versioning@w3.org Subject: Interaction of DAV:auto-version and DAV:checkout-fork Hi, consider the following scenario: - "a" is a version controlled resource with some kind of auto-versioning enabled, for instance DAV:checkout-unlocked-checkin, - it's DAV:checked-in property refers to a version with a DAV:checkout-fork property of DAV:discouraged and a non-empty DAV:successor-set property. Non-versioning aware client attempts to modify the resource. Possible outcomes: a) request is rejected (409 with DAV:checkout-of-version-with-descendant-is-discouraged), b) request is accepted and the version history will fork. I think both behaviours are allowed, but which one makes more sense? Right now, I'm preferring to reject the request (standard operations on a checked-in resource shouldn't introduce forks into the version history). Opinions?
Received on Sunday, 28 July 2002 09:00:13 UTC