On Mon, 30 Nov 2015 07:52:19 -0800 (PST)
Anthony Hunter <hunta...@gmail.com> wrote:

> I'm new to git and looking into its usage and am responsible for the
> teams that will be using the repos.
> As development goes on within a repo and time passes, the repo will
> start to get large just do to the number of commits. Is there a way
> that for example, as a release completes, I could branch off into a
> new repo to develop the next version? This will help keep the repos
> small and release specific. 
> I'm aware of doing a shallow clone or even a branch clone, but I will
> still end up with users cloning the who thing by accident.

I think you're trying to overengeneer on the two fronts:

* If you want a new repo, just `git checkout` a version tagged
  as released, then `rm -rf .git` there (literally), then
  `git init .` + `git add .` + `git commit -q` and you're done.

  Since you explicitly want no history, then there's no sense in
  "branching off" -- because branching means just growing the history
  in yet another direction.

* Is the repo size *really* a concern?

  Here's why I'm asking: there are projects like Mozilla Firefox and
  Linux which contain all their histories from day one, and they
  are *huge.*  Still, Git handles them just okay.

  Sure, there's always a problem of a slow initial clone but it can
  be solved in different ways, with using "--reference" when
  `git clone`-ing being the most apparent.

  If you store some large assets in your repos, you might consider
  using tools like `git-fat`, `git-lfs` and `git-annex`.

You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to