Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
Tony Hoyle wrote:
[in reply to me saying:]
> >Now, if the mergepoint processing could handle the merge from
> >branch-to-mainline as well as all the mainline-to-branch merges we do,
> >I'd be a very happy man :-) ('cos we'd then have a clear indication on
> >the mainline about which branch delivered every revision).
> It does - we merge branch to mainline all the time...
> Maybe I'm not understanding what you're meaning?
Correct - what I'm talking about is the messier situation of handling
mainline-to-branch merges followed by a branch-to-mainline merge
(possibly repeated multiple times), and making it so you only ever
have to use the single (appropriate) "-j" form for update.
As far as I can work out, there is sufficient information being held
by the commited mergepoints to work out which base version should be
used by the merge (obviously the other two versions are "branch head"
and "mainline head").
I'm not talking about arbitrary inter-branch merges - that is somewhat
harder to do :-)
I'll see if I can knock my ideas into something semi-understable -
my current drafts resemble gibberish...
phil
--
change "spam"/"news" to "phil" to email me