[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: incremental conversion from other SCM to svn by vcp

From: <kfogel_at_collab.net>
Date: 2003-06-19 22:24:04 CEST

Chia-liang Kao <clkao@clkao.org> writes:
> for a), VCP core deduces per-file placeholder rev at the branching
> point. my VCP::Dest::svn::create_branch makes use of them and deduces
> the `cp point' from those placeholder revisions.
> for b), that's the case i mentioned the current branch point guessing
> is aggressive, and i need to do more strict checks.
> let me make sure if the following is producing the case you meant:
> cvs add file1 file2
> <some modification and commits>
> cvs tag -b somebranch
> cvs add file3
> <manually edit file3,v to have the time ealier than the branching time>

That's one of the cases, but not the only one.

Consider a simpler case:

Someone has a "mixed" working copy, that is, a working copy where the
file's revisions are from very different points in time -- they are
not all in HEAD, and never have been. Now imagine the user does *not*
run 'cvs update'. Instead, she just makes a tag or branch in the
working copy.

Obviously, there cannot be just one Subversion revision from which to
copy. Maybe there's one revision that can get *most* of the
appropriate revisions, but further copies and/or deletes will have in
the resulting tree before the tag or branch is truly correct.


To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jun 19 23:12:08 2003

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.