- From: <gclemm@rational.com>
- Date: Wed, 30 Jan 2002 22:41:48 -0500
- To: ietf-dav-versioning@w3.org
Reverting to a previous state of the workspace will often involve reverting to earlier states of several resources. In the particular case described below, you need to remember the state of two resources: that of /src (so that you revert back to when it had a member called "a"), and that of /src/a (so that you revert to a particular state of /src/a. If you want to capture the state of several resources in a workspace, you can use a "baseline", which would have remembered both the state of /src, and the state of /src/a. If you only wanted to revert back to part of the state captured by a baseline, you can use the DAV:baseline-collection of that baseline to identify what versions were selected by that baseline, and then use UPDATE requests to bring back the parts of the baseline you are interested in. Cheers, Geoff -----Original Message----- From: Kirmse, Daniel [mailto:daniel.kirmse@sap.com] Sent: Wednesday, January 30, 2002 10:44 AM To: Ietf-Dav-Versioning (E-mail) Subject: Revert a Deletion of a VCR using version-controlled Collections Hi, what about this one here: /src/a /src is a version-controlled collection (VCC). a is a VCR. By now there is a version V1 of VCC /src that contains the binding (a) and a "pointer" to the version-history of a. Now I delete VCR a. Doing this I get a new version V2 of VCC /src that does not contain the binding for a anymore. So far so good. Sometime later I want to revert the deletion of VCR a. So I do an update of VCC /src so that V1 becomes the checked-in version of VCC /src again. VCR a will be created again. It points to the right version-history. But what is the checked-in version of VCR a by now? This property was gone to the dump earlier at deletion time. The only guess would be, the checked-in version of VCR a is the root version of the version-history. Well I my eyes this would be no real revert of the deletion, cause the state before the deletion and the state after the revert are not equal. Regards, Daniel
Received on Wednesday, 30 January 2002 22:42:51 UTC