Takes the branch name and source location as parameters. After you add a branch, you can cg-pull it whenever you want and it will keep your objects database in sync with it. Its latest commit is accessible as .git/refs/heads/branchname (or - more conveniently - as $(commit-id branchname)). For example, to make a diff between Linus (after you added him) and your current tree, do:

$ cg-pull linus
$ cg-diff linus

The possible location specifiers are:

  1. Local path - note that pulling will hardlink the objects, not copy them.

  2. rsync - either rsync://host/path or host:/path (the latter can change)

  3. HTTP - http://host/path Note that this does not work for repository containing packed files! (e.g. the current Linus' Linux Kernel repository)

  4. SSH - git+ssh://host/path or git+ssh://user@host/path - note that the path must be absolute.

The URL can have a fragment part, which identifies a branch inside of the repository. Therefore, if you have a repository git+ssh://host/path/repo.git and you are interested in its testing branch, you can e.g.:

$ cg-branch-add repo-testing git+ssh://host/path/repo.git#testing

and refer to it as repo-testing anytime later.


-h, --help

Print usage help.


Copyright © Petr Baudis, 2005


cg-branch-add command is part of cogito(7), a toolkit for managing git(1) trees.