New maintainer/canonical fork

[John Baber]

I just forked, fixed compilation errors, updated the README, etc. and realized that a dozen other people have also done so in the last 10 years.

minisat is a powerful, simple, open source SAT solver and I'm delighted the Niklases gave it to the world.  But the first step for anyone learning SAT is to figure out the situation and download one of 200 forks.

Would the Niklases want to crown a successor to accept pull requests to the canonical github repo.?

Of course an alternative is for one of the many forks to become the go-to resource.

[Norbert Manthey]

Hi John,

I followed the SAT developments from the recent years, and merged future techniques into a minisat fork that is still capable of the old interface (2.2.0, not the latest MiniSat github state). Most SAT competition solvers do not come with a git history, so just using those is hard.

My solver lives here, and from a performance point of view it can compete with CaDiCaL and other top solvers from 2020.

https://github.com/conp-solutions/mergesat

The question is whether the future fork should stay as close to where MiniSat os today, or whether future developments should be taken into account. With MergeSat, I try the latter approach in a lazy way, i.e. by merging development of competition submissions, testing them and fixing implementation issues that arise when conbining different techniques.

Other variants like using the Glucose solver sith the same (2.2.0) interface might also be a good successor.

Best,

Norbert

--

---
You received this message because you are subscribed to the Google Groups "MiniSat" group.
To unsubscribe from this group and stop receiving emails from it, send an email to minisat+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/minisat/65d11943-f891-4fc0-afc1-d6b4ac60ecf2n%40googlegroups.com.