Re: A question about branching, merging
jamsden@us.ibm.com
Wed, 8 Dec 1999 13:41:08 -0500
From: jamsden@us.ibm.com
To: ietf-dav-versioning@w3.org
Message-ID: <85256841.0066E620.00@d54mta03.raleigh.ibm.com>
Date: Wed, 8 Dec 1999 13:41:08 -0500
Subject: Re: A question about branching, merging
I think there is something very different about a line of descent created
by linear checkout/checkin and merging. In the linear case, one can be
reasonably sure the working resource was derived from the predecessor, but
in merging, there is a weaker (or at least different) relationship
resulting from a very different operation. I think these relationships
should be kept separate.
"Geoffrey M. Clemm" <geoffrey.clemm@rational.com> on 12/08/99 09:09:38 AM
To: ietf-dav-versioning@w3.org
cc:
Subject: Re: A question about branching, merging
I just added to the spec this more detailed description of how
merging is performed using the protocol. In the process of
doing so, the assymetry of picking one merge contributor to
be the "predecessor" and the rest as "merge-predecessors" became
evident.
I know there was one advocate during the breakout in Washington
of keeping these properties distinct, but I remain unconvinced.
So I'd like to hash this out on the mailing list so that we
can either convince ourselves that this distinction is important,
or convince ourselves that the spec. should be simplified by replacing
the four properties:
DAV:predecessor, DAV:successors, DAV:merge-predecessors,
DAV:merge-sucessors
with the two properties:
DAV:predecessors, DAV:successors.
Note: I will interpret silence to mean consent that we simplify (:-).
Cheers,
Geoff