Re: COPY collection to an existing resource

Geoffrey M Clemm <[email protected]> Thu, 13 Jul 2006 08:44:24 -0400
Newsgroups gmane.ietf.deltav
Message-ID <OF0CD84BC5.F8B200EE-ON852571AA.0045D4A1-852571AA.0045FB96@us.ibm.com>
This is a multipart message in MIME format.
--=_alternative 0045FAE0852571AA_=
Content-Type: text/plain; charset="US-ASCII"

Yes, and note that a client can also just DELETE the target before 
requesting the COPY, in which case a fresh set of resources would be 
created at the destination (each with a new version history, if the server 
automatically puts new resources under version control).

Cheers,
Geof

Werner wrote on 07/13/2006 05:53:37 AM:

> 
> In other words, the position is taken that when a user
> copies a collection to an existing destination, both are
> likely to be related to the user. Otherwise strange
> version histories would result for internal members that
> happen to coincide. This position seems reasonable.
> 
> There is a practical consequence for the user, however.
> Since the operation is recursive, the user has to be aware
> of the internal members that coincide at any depth and
> make sure that either they are checked out or that their
> "auto-version" property is set properly. It is also
> necessary that all subcollections at the destination are
> checked out or have the appropriate auto-version property.


--=_alternative 0045FAE0852571AA_=
Content-Type: text/html; charset="US-ASCII"


<br><tt><font size=2>Yes, and note that a client can also just DELETE the
target before requesting the COPY, in which case a fresh set of resources
would be created at the destination (each with a new version history, if
the server automatically puts new resources under version control).</font></tt>
<br>
<br><tt><font size=2>Cheers,</font></tt>
<br><tt><font size=2>Geof</font></tt>
<br>
<br><tt><font size=2>Werner wrote on 07/13/2006 05:53:37 AM:<br>
<br>
&gt; <br>
&gt; In other words, the position is taken that when a user<br>
&gt; copies a collection to an existing destination, both are<br>
&gt; likely to be related to the user. Otherwise strange<br>
&gt; version histories would result for internal members that<br>
&gt; happen to coincide. This position seems reasonable.<br>
&gt; <br>
&gt; There is a practical consequence for the user, however.<br>
&gt; Since the operation is recursive, the user has to be aware<br>
&gt; of the internal members that coincide at any depth and<br>
&gt; make sure that either they are checked out or that their<br>
&gt; &quot;auto-version&quot; property is set properly. It is also<br>
&gt; necessary that all subcollections at the destination are<br>
&gt; checked out or have the appropriate auto-version property.<br>
<br>
</font></tt>
--=_alternative 0045FAE0852571AA_=--