Hello, I came across this interesting question (and the author talked about pijul too). I wonder it is possible to use syncthing to sync pijul repository. Any pitfalls?
I’d be interested to know what’s a good way to back storage for the time being. I wondered almost if I could host a pijul repository inside a git one, just for the ease of the free hosting and rememebring where I put stuff. But I have just set up SSH for now and will see how that goes, and will be intrigued if you find any answer!
So I’ve lost the link now but basically the author successfully syncs git repository across his devices using syncthing and doesn’t do a git push. Only git commit.
I am going to try to do the same with pijul
At the moment you can probably do that, Pijul stores stuff in files, and you can sync those, the backend (Sanakirja) is meant to be cross-platform even with the same binary files.
Unlike Git though, the datastructures are a bit different in Pijul, the source of truth is the list of patches, and the big “db” file is a “cache” (in first approximation) of that.
I have ideas to change the current situation, but no time to implement them.
I think for now I wil just host all my pijul repos inside a private gitlab project; yes, it’s bad but it probably provides fewer risky points than always syncing via SSH, and so I shouldn’t be able to overwrite things by accident that I didn’t mean to. It’s probably terrible on space eficiency but that’s gitlab’s problem, not mine
So workflow will be to pull any changes from other machines from the git repo to my local machine, in one centralised location. Then somewhere in the bowels of a project, I work on my changes, when I’m done, push all my channels from the local project to a subdirectory in the centralised nest, and then push this back up to git at the end of the day. Then when I am on a different machine, basically the same thing.
I’ll let you know how I go!