Re: [Monotone-devel] cvsimport branch reconstruction

[Nathaniel Smith]

On Fri, Apr 07, 2006 at 05:11:25PM +0200, Markus Schiltknecht wrote:
> During RCS file parsing, for every branch three times are extracted, if
> available:
> 
> a) the time of the last commit before branching
> b) the time of the first commit in the branch
> c) the time of the commit following the branching in the parent

I'm not really qualified to even judge whether this approach is good
or not.  But maybe I could become more qualified, and everyone else
too, if you worked out what cvs2svn does to solve this problem, and
posted a description of that?  (Perhaps with a comparison of what it
does differently, too.)

I'm not suggesting this to just be annoying -- rather, because, that's
by far the most battle-hardened cvs parser freely available[1], and
this is the sort of nasty problem that just can't be solved except by
lots and lots of grunt labor, trying solutions, seeing how they fail,
improving the solutions, etc.  It's just not the sort of question that
we really _can_ evaluate on the basis of a mailing list post :-(.  So
we could either do that labor ourselves, or steal it from someone
who's already done it :-).

[1] We could also all nag Canonical Inc. to release their proprietary
cvs importer, but I've never gotten very far with that.  I guess it's
part of their strategy to attract developers to their proprietary
Luanchpad portal, by being the only purveyors of accurate imports from
CVS?  I don't really know.

-- Nathaniel

-- 
"Lull'd in the countless chambers of the brain,
Our thoughts are link'd by many a hidden chain:
Awake but one, and lo! what myriads rise!
Each stamps its image as the other flies"
  -- Ann Ward Radcliffe, The Mysteries of Udolpho