[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Monotone-devel] Monotone CVS sync
From: |
William Uther |
Subject: |
Re: [Monotone-devel] Monotone CVS sync |
Date: |
Fri, 16 Feb 2007 11:43:59 +1100 |
On 16/02/2007, at 10:28 AM, Christof Petig wrote:
Christof Petig schrieb:
I will put some work into cvssync3 tonight, stay tuned.
I'm not finished yet, but I will continue tomorrow.
Wow, thanks. I wasn't expecting a rush to get it done.
At the moment I'm using rev cd65adc2a113afeb020dc3a0eea88cebe743e3f9
of the .refactor branch (2007-01-30T18:26:40). It generates an
mtn_cvs binary that seems to work, although I haven't tested it very
thoroughly.
You mention different versions though:
State:
- - cvssync1: finished, tested, slow
- - cvssync3: work in progress, did work before the summit, needs a
bit
more work to compile again, passes test cases but I saw some
strange issues
- - cvssync4: proposed by Nathaniel, still to get properly designed
How do these map to the source code/branches? How do I know which
I'm using? e.g. grep for cvssync1 doesn't match anything in the source.
Does the fact that cvssync3 "did work before the summit" mean that my
reverting to a revision from the end of January means I might have a
working cvssync3?
Are these different sync mechanisms going to be compatible? Will
there be an upgrade path? e.g. if I start using cvssync3 and then
you get cvssync4 going, am I going to be able to keep syncing?
The way things stand, mtn ends up with a complete history, and CVS
ends up with a limited history. That is fine, but if the update
means throwing out the mtn db and re-pulling from CVS then I'd lose
history (which may not be the end of the world, but...). If the
update means changing the CVS repository, that is more problematic.
One final question... what should cvssync do with moves and renames?
It seems to break things at the moment. It isn't clear what should
happen... is it best to do a grep for "Renamed entries:" in the log
and call for help if it matches? Then the user manually disapproves
of that revision and things work again?
Thanks for writing cvs-sync :)
Cheers,
Will :-}