fractal_timescales

Mathematician working in epidemiology, love to go down the rabbit hole on technical stuff. Computers, chess, general relativity, they're all past and future obsessions. I really like to hear experts talk about things they know well! The more I know, the more subjects I can make puns in!

Last Notes

2026-05-10 16:36:12 GMT

- reply

Something about this Borg comparison is reminding me of how @npub1fdr…lvhs is always saying that Margaret Thatcher's "There is no alternative" really means "Stop trying to think of alternatives". Given how often Janeway outfoxed the Borg, I guess resistance pays off.

2025-09-24 01:27:13 GMT

- reply

my wife and I don't own a home yet, but when we do we want to make solar a priority. Are there any resources you would recommend looking at to try to gain the knowledge to do this ourselves? I'm not sure I'd be able to, but I'd like to know for sure before I have to put my future in these crap ass companies' hands.

2025-09-08 13:41:56 GMT

- reply

So just to check my understanding here, you'd make a special release, binaries only for the patches. Someone outside the project decompiles the binary, uses it to discover the vulnerability, which they then report to you. Then you fix it for regular Graphene users.
If I'm understanding correctly, then is there a chance the regular Graphene users don't get the patches if nobody outside the project does the decompiling? I would love to contribute, and maybe this is a reason to get good at doing this kind of thing. But I don't think my technical skills are up to scratch yet.

2025-08-22 17:20:46 GMT

- reply

last time I saw a diagram with rotations like this it was an explanation of Grover's algorithm, and now all I can think is "watch me reach good for people in square root time by repealing the DMCA".

2025-07-07 15:58:53 GMT

- reply

any chance you (or others) would be willing to explain roughly how one would go about mitigating something like this? Do you restrict the ability of the app to use custom animations or display settings?

2025-07-01 20:09:28 GMT

- reply

Happy to help! The class was quite interesting. He did work previously with a student on the "can you hear the shape of a drum problem" restricted to triangular drums, using validated numerics. We basically spent the last few weeks applying what we had learned from Evans' PDE book to understand their proof. Instructive and interesting!

2025-07-01 11:30:27 GMT

- reply

I've taken classes from the professor in question, he's well-known for validated numerics. I imagine any proof they have will use those techniques, it was mentioned briefly in their article. Checking their proof will be a lot more like debugging a computer program than a traditional proof. My guess is that the strategy is to use validated numerics to prove that a physics inspired neural network has provably found a solution with a loss of regularity.
https://en.m.wikipedia.org/wiki/Validated_numerics