[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Monotone-devel] Re: cvs_import rewrite
From: |
graydon hoare |
Subject: |
[Monotone-devel] Re: cvs_import rewrite |
Date: |
Fri, 16 Dec 2005 19:40:40 -0800 |
User-agent: |
Mozilla Thunderbird 1.0.7 (X11/20050923) |
Markus Schiltknecht wrote:
what's up with the cvs_import rewrite branches? Anybody still working on
such a thing?
I completed the branch and merged it back to mainline, quite a while ago.
cvs2svn uses a multiple passes of processing the cvs repository.
Yes. The .rewrites.cvs_import branch copied their algorithm reasonably
closely. It's a good algorithm, that's why I copied it.
There is one thing you might notice missing. Others have characterized
this as "not processing branches". That's not true: it processes
branches, it just doesn't arrange them in a *hierarchy*. All branches
are considered flat, starting from the root revision, with no historical
relationship between them.
There is a reason for this. It is essentially a pessimistic decision
based on the fact that there is not necessarily any single "parent
branchpoint" for any branch in CVS. Each file can enter a branch "foo"
from some different branch "bar", at any point in time. There need be no
common branching structure between files. So synthesizing one is
guesswork at best.
Currently we make no guess. It would be an improvement to make a guess.
If you want to help write the code to make such guesses, I'd be happy to
show you around the cvs_import code and answer any questions you have.
Otherwise I will eventually get around to it, as I know it's an
important feature.
-graydon
- Re: [Monotone-devel] cvs_import rewrite, (continued)
- [Monotone-devel] Re: cvs_import rewrite,
graydon hoare <=