Being very new to git, I'm trying to get an overview of how to use multiple work directories for different branches off one github project. In particular I want/need to work on two branches - one the 'master' the other one a maintenance release, say maintenance/project1. The plan is to run the work under Linux/Mint 17.2. As well, I would like to make use of the latest git worktree features introduced in 2.5.
Assuming I have a github account, have forked my own fork off the main project and have cloned the fork to my local machine in ~/user/myproject.
As things stand now, I don't expect to contribute code via git push, all the results of my work & testing will be passed on to others for possible fixes via e-mail - doing it via git might be nice down the road, but won't be necessary to start.
What I see as the steps I would need a corresponding set of git commands for are:
Creating the two work trees
git clone https://github.com/someone/someproject.git
(this will give you directory someproject on the master branch)
cd someproject; git worktree add ../someproject-devel devel
(this will give you directory someproject-devel on the devel branch, assuming such a branch exists)
Switching between these trees
Just use the cd command.
Keeping both trees up-to-date with the upstream master
Just cd in the appropriate directory and run git pull. Doing this in the main working tree first may be best, but is not required (the man page does not address this specifically, and I haven't looked into the worktree implementation in detail. If an issue existed it would only be one of optimal use of space, not functionality).
any other cautions/ special considerations for working this way.
From the man page:
Multiple checkout in general is still experimental, and the support for
submodules is incomplete. It is NOT recommended to make multiple
checkouts of a superproject.