[cvsnt] Re: Difference between tag and branch without commits?
bo.berglund at telia.com
Thu Aug 25 20:04:10 BST 2005
On Thu, 25 Aug 2005 14:11:53 -0300, Gerhard Fiedler
<lists at connectionbrazil.com> wrote:
>is there a difference (performance, storage) between a simple tag and a
>branch tag to which I don't commit anything?
>When sending test versions to a tester, I don't know up front whether I
>will have to make changes to a certain test version.
>- I could tag the versions, and only branch them if I actually have to make
>changes to it.
>- Or I could just branch them all, and leave the ones alone where I don't
>have to make changes. This is more convenient, obviously. But I'd like to
>know whether there's a downside to this -- that is, whether a branch tag is
>more "expensive" in any way than a simple tag.
THe down side is that if you make a branch at the release point and
then later work on that branbch it is very difficult to get back to
the exact release state. There is no way to use a cvs command to get
the *base* of the branch once it has gone active.
So I vote for setting a release tag on the TRUNK and later (if needed)
branch off that point to work on some modifications to that particular
(Bo Berglund, developer in Sweden)
More information about the cvsnt