In our environment, we develop exclusively on the trunk; periodically we try to coalesce all the development and make a release of the compiled code. In spite of our efforts, we must normally create patches during the interval until the next release.
I am unclear how to keep and maintain a record of the patched release in the repository. Currently, we just tag the release, though we probably should make a branch tag also. Normally, the source revisions for the patches will be (or have already been) developed on the trunk, and are then integrated into the release by hand (eg, copying the relevant files to the right directory and running make). I would prefer to run some sort of update/commit commands on the release branch, to avoid the copy, and so that I could later checkout from the repository the patched release source. However, I get the impression from the documentation and info-cvs archives that this is either not possible or involves a potentially horrendous number of tags. Is it possible from some directory in the branch to update some files or the whole directory from the trunk with a cvs command? And then check them in as a revision to the branch? Alternatively, is there some other standard approach to this problem? I'd like to see explicit examples of the commands involved; some of the archive files talk about things like this, but it's unclear how the branch updates are occurring; /bin/cp seems to be a possibility. Thanks, -- danq _______________________________________________ Info-cvs mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/info-cvs
