science fiction is misnamed. science fiction is primarily "future fantasy". most actual fiction about science is cosmic horror
17.02.2026 02:38 โ ๐ 131 ๐ 16 ๐ฌ 10 ๐ 0@heartpunk.bsky.social
trying to see if we can recover programming language semantics from implementations @ patreon.com/heartpunk 39 y/o, MDNI ๐ she/they/bitch(es) alt for pfp: the soul of the rose by john waterhouse alt for banner: sappho and erinna by simeon solomon
science fiction is misnamed. science fiction is primarily "future fantasy". most actual fiction about science is cosmic horror
17.02.2026 02:38 โ ๐ 131 ๐ 16 ๐ฌ 10 ๐ 0Counselor: These feelings of "everything sucks," how long have you had them? Butt-head: Uhh, since everything started to suck, I guess.
Here it is, your moment of zen.
17.02.2026 01:43 โ ๐ 8851 ๐ 2294 ๐ฌ 29 ๐ 45Have you ever once seen the New York Times quote Trump like this?
16.02.2026 18:24 โ ๐ 12783 ๐ 2510 ๐ฌ 517 ๐ 670ok seems like no
17.02.2026 07:32 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0ok i think i may have doubled my metformin dose this morning on accident and if so that could explain the discomfort i had all day bcz that is a huge dose change all at once (a whole extra gram!!!)
not sure tho tbc i just,,, dk
interesting!
i just notice it *soooo* much w/penny
kinda makes me feel bad for her??
ok different machine will be the test host but in any case i'm about to try it out p soon, or start to; i'll ahve to take a break tomorrow bcz no more claude quota until wednesday 9 am
17.02.2026 07:11 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0using $50 free claude extra usage and WOW that goes by fast holy *crap* the subscriptions are a good deal
17.02.2026 07:06 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0ok it should work but i have to get another machine working to test it, i think one of my laptops should work tho so i'll see if i can just use that?????
17.02.2026 06:44 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0(also pretty chill about the instability too overall too fwiw)
17.02.2026 06:34 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0claude just doing an nmap w/o asking lmfao (i did ask it to find a host for me tbf)
hope it doesn't print something like that one time i did on accident...
literally 4s after u posted that i posted my reversal update lmfao
my emotional world is so unstable
there's a new update tess
17.02.2026 06:24 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0WAIT fjlkasdfjaklsd;sdkla; lol liveblogging my research flailing here
it seems the data DOES get *generated* by the research artifact for STALAGMITE but is not *in the repo* so if i get my bigger machine up and running i can do it overnight... wheeeeeee
i was so right to be skeptical i should not have let my hope get up lol
utterly impossible
tbc *proof* of concept not *general purpose implementation of even toy quality*
literally just one example showing it likely works
tl;dr might be able to reuse data from the STALAGMITE framework to compute better grammars directly?
i am somewhat skeptical tbh
CLAUDE CAME UP W/A BRILLIANT MAYBE SHORTCUT EEEEEEEE I MAY HAVE A PLAUSIBLE PROOF OF CONCEPT SO MUCH EARLIER I HOPE IT IS NOT BULLSHIT I GOTTA SEE AFJKLAD;FJKALDSSKADL I'M TOO EXCITED FJKADFLKJADSLKASDKLAS
17.02.2026 06:16 โ ๐ 12 ๐ 0 ๐ฌ 1 ๐ 1tired of sharing a claude code subscription but glad i get to have one at all
about to hit rate limit and like i literally can't manage usage solo, just kinda annoying
hitting good rhythm for the project i'm on rn
oh lol about to be out of claude quota, so, i guess more break, enforced,,,
that ,,, well i guess i have codex quota??? i'll use that i guess.
*giggles*
17.02.2026 05:46 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0also it would tend to validate the learnability framework i tried to prove but like holy fuck it's so abstract lmfao
17.02.2026 05:37 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0if i get far enough and have some results by EOW i may submit this as a WIP report to langsec in case the other doesn't get in, because if we could get substantially improved formal grammar recovery, that's defs just as langsec relevant as the original result
17.02.2026 05:37 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0found the most well contained instance yet of this new learnability framework i'm working on, and it seems to be fixing the gaps in the STALAGMITE framework for grammar recovery from parser implementation
starting to spec out an impl already bcz i want a proof of concept for this shit ASAP
A fake diagram that's a complete mess of nonsense. There's nothing really to digest from it, except the center which says "The nexus of uncertainty"
every first pass at a diagram on google slides
16.02.2026 23:53 โ ๐ 121 ๐ 8 ๐ฌ 15 ๐ 2and iโm also pretty confident my paper wasnโt nearly the best submission i couldโve done and maybe wildly inadequate
ah well
i did what i knew how to in the time i had
i guess it is literally the best i could have for this year?
iโm p fucking burnt out from that paper sprint, yep
just like spent some time literally just lying in bed and only got back out bcz it was too boring/needed to change positions
i just dk how to manage myself after this
iโd like to keep working but i clearly need the break
Finally, we can talk to the super-persuader from the acclaimed blog post โdonโt talk to the super-persuaderโ
17.02.2026 01:06 โ ๐ 47 ๐ 8 ๐ฌ 2 ๐ 0I'm very sorry to beg again but things have not improved, in fact many things have gotten worse! please share and donate what you can.
gofund.me/22e0087c7
i just also dk how to fix it and iโm doing what i know how to to fix what i know how to
itโs just not really enough
maybe i manage to get a job somehow that could fix things
or i build a profitable enough company that would be preferable tbqh
ehhh debatable idk
iโm just overwhelmed af ok
have gotten a little break today sort of but like also half of it at least was just stretching to restore my body to basically functional and beyond that just fuck iโm lonely
i havenโt had any irl social time over like 30m so far this year (and that was just once)
i know thatโs sort of my fault