Hi, I'm swapping a bit between IDE, gitk and the command line much like I did with svn.
I'm also still trying to iron out some wrinkles in the process. Mostly I think that is getting git configured correctly. I'm just going to use master and 8.5.x below but it should work with 7.0.x as well. Initial setup cd ~/repos git clone g...@github.com:apache/tomcat.git asf-tomcat-9.0.x git clone g...@github.com:apache/tomcat.git asf-tomcat-8.5.x cd asf-tomcat-8.5.x git checkout 8.5.x # I don't want a tracking branch for master in this clone # because if I do git push complains that it isn't up to date git co master git branch --unset-upstream git checkout 8.5.x git branch -D master #Commit ... cd ~/repos/asf-tomcat-9.0.x # edit files git commit -a -m "Some message" git push # ... and backport cd ~/repos/asf-tomcat-8.5.x git fetch git cherry-pick <hash> git push I am sure there are many different ways to handle this. If you wanted all the tracking branches in each clone, I am sure you could configure git push to just push the relevant branch for that clone. I can probably do more in the IDE than I am currently doing. I'm still finding my way a bit and figuring out what process works best for me. HTH, Mark On 28/02/2019 10:04, Rainer Jung wrote: > Hi there, > > I am struggling a bit with the new merging workflow from master to the > branches. I did a short read on the workspaces and tried cherry-picking > with and without workspaces. It seems workspaces are not for doing > actual commits that should at the end et pushed to origin. And when > using cherry-pick into a checked out 8.5.x branch I end up with a > detached head state. > > Assuming commits that apply cleanly to the branches, what is the > proposed workflow? > > Something like: > > git checkout master > > ... apply changes ... > > git add <files to commit> > git commit > git push > > git checkout 8.5.x > ???? > git push > > git checkout 7.0.x > ???? > git push > > Any initial recommendations? I am talking here about simply short > commits that might not warrant creating a local branch (eg. > b5c4c30b8037b587b8804f1e6e764bbfd4ffb727). But that's probably > independent of the backport workflow. > > Regards, > > Rainer > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org > For additional commands, e-mail: dev-h...@tomcat.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
