[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Branch mess, and starting a project with import
From: |
Jim.Hyslop |
Subject: |
RE: Branch mess, and starting a project with import |
Date: |
Thu, 3 Jun 2004 16:55:02 -0400 |
Peter S. May wrote:
> Tom Copeland wrote:
>
> > On Thu, 2004-06-03 at 14:43, Jim.Hyslop wrote:
> >
> > Hm, that's interesting. I've always started projects in
> CVS by doing an
> > import of some existing directory tree... just seems to
> remove some of
> > the tediousness of lots of cvs adds of directories...
>
> Indeed, the CVS Book 3rd edition (
> http://cvsbook.red-bean.com/ ) around
> page 31 outlines cvs import as the way to start a new module in the
> repository.
Well, that is _one_ way to do it, but I really think it's a mis-use of the
import command. You still have to create the entry in CVSROOT/modules (a
very important step for a repository of any significant size).
Haven't you ever wondered why you are *required* to provide a branch tag
that will never be used?
> The vendortag and releasetag arguments are a bit of
> bookkeeping for CVS. Don't worry about them now;
The tags are not "bookkeeping for CVS", they are required in order for the
command to operate as it was intended to operate.
> That does seem to be less of a hassle (and probably less error-prone)
> than editing the repository guts...
Who said anything about editing the repository guts?!? The command sequences
that Mark and I posted earlier do not directly change the repository. Every
change to the repository is done through the CVS interface.
--
Jim Hyslop
Senior Software Designer
Leitch Technology International Inc. (http://www.leitch.com)
Columnist, C/C++ Users Journal (http://www.cuj.com/experts)
- RE: Branch mess, and starting a project with import,
Jim.Hyslop <=