- From: <ccjason@us.ibm.com>
- Date: Tue, 7 Dec 1999 19:02:48 -0500
- To: "Eric Sedlar" <esedlar@us.oracle.com>
- cc: "Geoffrey M. Clemm" <geoffrey.clemm@rational.com>, w3c-dist-auth@w3.org
>> Couldn't you make PUT and MKCOL all create strong bindings, and then have weak BINDings from BIND? Then you wouldn't have to introduce fake bindings to maintain persistence? >> You could do it many ways. If I interpret Geoff's proposal correctly, a bindings to collections would use a symbolic link and all bindings to non-collections would us bindings. Your suggested approach could also be taken. In WebDAV bindings created via GET/MKCOL are no more privledged than those created with BIND. If you decided to implement them differently, then you'll have two types of situations to handle when a DELETE is requested. You'd have handle deletion of bindings to directories implemented as hard links and as symbolic links. The deletion of a symbolic link just works. Deletion of a hard link isn't quite so easy. You can't just delete it because all those symbolic links get broken. ... so instead you'll have to delete one of the symbolic links and then move the "deleted" directory to the location where the symbolic link was. And then also redirect any other symbolic links. Of course in order to do all this, you have to keep track of all your symbolic links and manage them during method processing. Geoff's approach makes these methods map to trivial file system operations and allows GC'ing to be defered
Received on Tuesday, 7 December 1999 19:04:20 UTC