What is the use case here? Why would I need to mob next if I’m already remote pairing with the person?
I think it’s for people that like the idea of taking turns on the keyboard when pair/group programming, but are working remotely on some project that’s not well suited to working on at a shared remote server or something like that (like desktop application?).
I’m working with backend stuff so in my team we can just as well just log in to the same development VM.
Yeah, usually I pair using an app that can take over the remote keyboard (such as tmate or Tuple)
What happens when there are multiple teams on the same code base? Will things collide? Not rethorical, they might’ve already thought of this.
You can specify a suffix for the branch name. If you don’t, and two different teams target, say, master, I believe that would create chaos on mob/master (with two histories merged together). From the docs:
start [<minutes>] Start a <minutes> timer
[--branch|-b <branch-postfix>] Set wip branch to 'mob/<base-branch>-<branch-postfix>'
In my team, even code that is mob-programmed still has to go through formal code review, so I normally create an empty branch that will ultimately be used for the PR. Then, collaboration happens on mob/that_branch_name, which prevents conflicts neatly.