Question

Currently I always worked with GIT and was quite satisfied.

Due to some project restrictions we took over an already started project which recently hit the 1.0 production version.

Now I would like to branch the stable version 1.0 and we keep on developing in the trunk.

Now my structure looks like:

branche
--app-1.0

trunk
--app

Now two scenarios: I keep on developing in the trunk. I fix a bug which also hits version 1.0. in one Java class. It’s only 1 line of code that changes. How do I commit this to app-1.0 without overriding the “old” class except with this one single line?

2nd scenario: Someone asks me to check if there’s a bug. I switch to app-1.0, find the bug, fix it, I find by chance a second bug, fix it too, commit it to app-1.0. How do I commit this to the trunk?

Is in both cases the merge command correct?

What are best practices?

On my GIT I alway had it the other way round: I created a branch app-1.0 app-2.0 where I develop and when I found it to be stable I merged it to the trunk which was always a representation to my stable version.

I hope you guys can help me 🙂

Answer 1

For both scenarios : you merge!

From SVN Book:

Merge a branch back into the trunk (assuming that you have a working copy of the trunk, and that the branch was created in revision 250):

$ svn merge -r 250:HEAD http://svn.red-bean.com/repos/branches/my-branch
U  myproj/tiny.txt
U  myproj/thhgttg.txt
U  myproj/win.txt
U  myproj/flo.txt

If you branched at revision 23, and you want to merge changes on trunk into your branch, you could do this from inside the working copy of your branch:

$ svn merge -r 23:30 file:///tmp/repos/trunk/vendors
U  myproj/thhgttg.txt

http://svnbook.red-bean.com/en/1.0/re16.html