@simontatham
May I ask in what use case you are sending patches by mail? I assume you have read permissions to clone the repo before you start working. It seems pushing back to a pr branch is not an option. Is this for security reasons?
@SpaceshipOp if you both have accounts on the forge site where the code is hosted, then sure. But if it's a forge site with limited membership, so that the other collaborator _can't_ make an account, or if it's not even a forge at all but just a git repo put up in someone's plain web space, then you have to send patches in some more ad-hoc way.
In this case, it's the former: the code is hosted somewhere that I have no reason to believe I'd be able to create an account. And it's a limited-time collaboration – we're polishing up one code change together until both of us are satisfied with it – so there's no point going to lots of effort to try to change that.
@simontatham
Understood. I did that as well with CVS/SVN repos, but that's more than 20 years ago. Was not aware that this ist still a thing. Seems my view on oss is too narrow. Thanks for widening it.
@SpaceshipOp it is weird that git usage is so conventionalised these days. Its whole design is geared towards supporting many wildly different structures and use cases – more than CVS/SVN could manage, in particular. But now everyone is so strongly expected to use a standard forge that it's a surprise to find git can be _even_ as flexible as SVN!