Equality Saturation: An “Incomplete” Project
https://blog.sigplan.org/2026/10/01/equality-saturation-an-incomplete-project/
ACM Special Interest Group on Programming Languages (SIGPLAN) PL Perspectives post by Max Willsey

I gave a talk today about some compiler things to some compiler people, and just posted a recording (https://vimeo.com/1234257957) and slides (https://cfallin.org/pubs/vmil2026_nightmonkey_slides.pdf).

In brief: I made JavaScript-running-in-a-Wasm-module go way way faster by building a static type inference and AOT compiler, now dubbed "NightMonkey". It's been tons of fun!

(the "why?!" is covered in the first part of the talk -- certainly a unique set of requirements but IMHO it's well-motivated for our niche)

A group of 25 Fields Medalists, including myself, have made a joint declaration on Math and AI: https://mathandai.org/ . We welcome additional signatories (similar to the Leiden declaration), as can be seen on the page. See also this article in the Economist announcing the declaration: https://www.economist.com/science-and-technology/2026/09/11/top-mathematicians-are-outraged-by-openais-methods

So fun that every PR interaction is now also a Turing Test

Cooling my PC by compressing files on my server and decompressing them on my PC.

Happy 5.25 day!

“As Gregor Samsa awoke one morning from uneasy dreams he found himself transformed in his bed into a gigantic compiler front-end”

We've released a demo of the project @CanLehmann and I have been working on: @verijit is a meta-tracing Verilog emulator that specializes the simulated core to the machine code that is being executed on top of it. Was lots of fun to work on! I'm quite happy with the results too, we reach two orders of magnitude improvement over Verilator for a simple RISC-V core
https://www.youtube.com/watch?v=PXgUsEjvAOY

I'm fundamentally a tool builder, and LLM coding agents work one million times better if you give them good tools, and I wrote a thing about this

https://john.regehr.org/writing/zero_dof_programming.html

A few years ago I designed a way to detect bit-flips in Firefox crash reports and last year we deployed an actual memory tester that runs on user machines after the browser crashes. Today I was looking at the data that comes out of these tests and now I'm 100% positive that the heuristic is sound and a lot of the crashes we see are from users with bad memory or similarly flaky hardware. Here's a few numbers to give you an idea of how large the problem is. 🧵 1/5

@guenther @gabrielesvelto ah yes, bitflips georg who lives in a plutonium mine and gets a thousand bit-flip related crashes every day is an outlier adn should not have counted

When AI Writes the World’s Software, Who Verifies It?

https://leodemoura.github.io/blog/2026/02/28/when-ai-writes-the-worlds-software.html

"Writing a specification forces clear thinking about what a system must do, what invariants it must maintain, what can go wrong. This is where the real engineering work has always lived. Implementation just used to be louder."

while we're at it, let's make sure everyone has read the audiophile memcpy post

https://www.audioasylum.com/messages/pcaudio/119979/

in 2017 a popular twitter game was to type a partial phrase then see what your phone auto-completes it with.

this proved so popular that it is now the only business model in the US.

You don't use open source software because it's better (it usually isn't).

You don't use open source software because it's freer (it only sometimes is).

You don't use open source software because it's got better politics (it isn't always).

You use open source software because *it is the only option*. In the long run, if it isn't open source, it doesn't exist.

image source: keithstack.com

combined Davos/POPL keynote ideas

"You'll own nothing and you'll be happy: a garbage collection rejoinder"
"The great shift/reset: a continuation or delimiter of progress?"
"Late-staged programming: are the economics of macros sustainable?"

Scheme, the ur programming language, is 50 years old, as Jason Hemann just reminded me.

I hereby coin the term "Ptolemaic Code" to refer to software that appears functional but is based on a fundamentally incorrect model of the problem domain. As more code is generated by AI, the prevalence of such code is likely to increase.
1/7

#TheGeneralTheoryOfSlop