wwilson

Nope, it's a generic PBT harness. The human who built it knew about the bug, but the workload was specifically designed not to specifically target this particular bug, and may have found several others as well (we are in the process of reporting these upstream).

Brutally honest answer: it only recently became easy to get random systems under test with Antithesis without a bunch of manual/expert effort. Expect us to be reporting a lot of bugs against open source projects in the near future.

Discussion here: https://news.ycombinator.com/item?id=49277799

As the post (and linked repo) pretty clearly indicate, zero info about the existence of the bug went into this work.

Oh btw, we also found some other ones… Stay tuned!

We are working on getting self-serve signups to you ASAP. For now, just email us. We don’t bite.

We wanted to publish this ASAP since people are talking about this bug today. In a follow-up, we will show how our automated causality analysis could have saved Tailscale and SQLite from 6 months of root cause analysis. Stay tuned!

RECEIPTS: https://antithesis.com/blog/2026/wal-reset-bug/

Stay tuned!

Was curious so we checked and yep, Antithesis finds this bug in about 15 minutes. Will post a repro/writeup here soon.

I’ve been running one for a while and love it. Have also built a Nix -> OpenWRT config language transpiler so that I can keep my router state in Nix files and have nice deterministic rollbacks etc. It’s been great!

Author here. That’s right — the challenge was representing a tree in an existing SQL database that we’d chosen for its pricing model. I think encoding a B-tree in SQL would be a lot less fun even than what we did.

Thanks,our company is in the DC area so I just reached out with an offer to chat. Wesnoth is an incredible project, I can't believe he doesn't have a programming job.

[Disclaimer: Started Antithesis]

If you run your Hegel tests in Antithesis, you get this for free (along with various sorts of “non-local” assertions, perfect reproducibility even for concurrent or distributed code, etc.).

But yeah, not hard to hack together basic coverage guidance outside Antithesis. That works well for large classes of programs, just not a majority of them.

If I had been wearing my fiendish CEO hat at the time, I might have even said something like: "somebody pointing this out will be a great way to jumpstart discussion in the comments."

One of the evilest tricks in marketing to developers is to ensure your post contains one small inaccuracy so somebody gets nerdsniped... not that I have ever done that.

Glad you enjoyed the talk! Making Bombadil able to take advantage of the intelligence in the Antithesis platform is definitely a goal, but we wanted to get a great open source tool into peoples’ hands ASAP first.

Not our first customer, but one we certainly appreciate!

Antithesis | Distributed Systems Breaker Extraordinaire | Full-time | SF or London or DC

---

Have you always wanted to channel your inner Kyle Kingsbury? We are looking to hire somebody with a deep understanding of distributed systems who can bring open source or customer software under test in our platform and find fun issues in it.

Preferred locations: SF, London, DC. Remote is a possibility for this role if you're really good (though be aware that we're a very in-person company culturally speaking, so it might be annoying for you).

To apply: email 'will' at our company domain with a description of something interesting that you've worked on.

I am one of the people who was on the panel. As always, it's a very lossy process when a journalist is summarizing another journalist summarizing a 90 minute discussion. Happy to expand in the comments here on any of the issues that got brought up.