I mean Louvain actually

what would i give to be a fly on the wall at the agda dev meeting in Leuven

you'd think.. ....

I'm bad at researching, the first thing that came up was some sort of stutter bisimulation but I'm not sure that's exactly what I need and not much beyond that.

I'll have the weekend to idly think about the problem, maybe something will come of it

defining this "quotient" is what's proving very difficult, so much so I might just change my cfg (again (don't ask))

indeed, but in my dvlpmt, a cek state could be related to multiple nodes, as not all of them carry semantic meaning. this is a big problem and to fix it I either have to write a new CFG or quotient the one I have in some way. I thought I could do without functionality but it seems not :/

the lecture notes i found carry out the same proof i've been meaning to carry out, using the same approach as me, to establish the main result, but they assume there's a one-to-one index correspondence between code points and CFG nodes -> no correspondence theorem needed. lucky...

thank you for giving it a read!!! i' stealing valor here, a lot of good people at jb are working on snakt and i'm sort of satellite to them. it is a great project though !!!

re incorrectness logic: i do see it as some form of dual reasoning, which imho has *some* validity? but it's def not ideal

the main questions in this are "is the approach sound" and "if it is, why do i require functionality" and "is there a way to get functionality".

i have found lecture notes that do the proof i want to do, with the approach i want, whose approach requires functionality.

which is good, but now work.

heartbreaking: the big problem you were bashing your head on was the good problem you should keep bashing your head on.

save me CMU lecture notes save me

hood Elon Musk be like: "yo wassup every member of my immediate family hates me"

tsubi club - trinket [full album] YouTube video by tsubi club

speaking of random cyberk1d music, do yourself a favor and listen to tsubi club new album

Absolutely Mental, super catchy. glad they're back

listening to bladee is crazy because every song is either the most lush soundscape you've ever heard in your life or a random mishmash of sound effects with the dumbest singing ever recorded on it, and the two songs will sound basically the same. true artistry.

this has got to be one of the wordiest Wikipedia articles I've ever seen. not too long just reads like a math arxiv

i love opening a new lean file and stuff is just

broken

like wdym you don't know what set is anymore

iowJdoaijsldkAWiosjdiwoasd that is indeed a previous draft that was published :((( will be fixed soon! thanks for pointing it out

thank you!!! i spent a great deal of time finetuning it, it was a struggle

glad to hear it worked well :3

anarcho-mistralism

anarcho-macronnism with von der leyen characteristics

i'm still very new at this writing thing. hopefully i'll get better! looking for stylistic advice as well.

as promised, the first "real" writeup on my website: on dataflow soundness. looking for all kinds of feedback and input!

expect a writeup on formalizing dataflow soundness in Lean fairly soon! end of week normally. i'll take any input i can get atp.

job so niche within my department that my new task is "draft a writeup about what you're doing so we can understand you and why you're blocked" i love it here

TRVTHNVKE

that's what i have proposed to my supervisor but she has been ghosting my slacks... hopefully it's enough.

want to prove soundness
need to prove relation validity
want to prove relation validity
need to break functionality
prove relation validity
go back to soundness
soundness requires functionality

are you kidding me

posting to distract myself from the abysmal local train performance

chainsaw man is so fucking goofy I don't think there's a single story beat that lands how it's supposed to and it's great

frame composition is CRAZY though. he does some shit with depth of field that makes characters pop it's super fun

writing documentation for Lean is so fun you get to cosplay being a mathematician

writing proofs (documenting my code as proof i understand what i'm doing)

i lied i don't want to be an engineer please don't give me a regular job