There have been a lot of Pijul releases recently—roughly eight beta versions from late May.
It is great to see this level of activity. Could the maintainers share what is currently being worked on and whether there is a rough roadmap for the next stable release?
No firm dates needed—just a brief overview of the priorities and direction would be very helpful.
We do have a roadmap, and a small group of people knows it, but it can’t be shared publicly at the moment, for various reasons.
The main striking point of these recent releases is: we now have a fuzzer, which was inspired by recent bugs surfaced by AI agents using Pijul in parallel in unintuitive way. I never had the time to write that, but since this is isn’t part of hard-core algorithms it’s easy to generated with LLMs.
I now believe that the recurring bugs around unrecord are gone, all the other commands have been stable for years. There are still minor performance issues here and there, but that’s it. And many projects, I’ll talk about them on the blog as soon as I can.