Notices by blumlaut (blumlaut@hounds.online)

just watched project hail mary, amaze amaze amaze

finally a movie where both the aliens and us are roughly "on the same page", i was waiting a long time for that

bro if i had access to furtracks source code i would go absolutely fucking ham with bugfixes

RuhrFurMeet 05/2026.

So glad to be back again after a few months of absence, it's been great seeing and talking to you guys again!

All Photos on Furtrack:
https://www.furtrack.com/index/event:ruhrfurmeet+photographer:blumlaut

#fursuit #photography #ruhrfurmeet

i built another keyboard

fab

tfw nginx proxy manager explodes

📍Snetterton Circuit

this happened during the track day, mr2 did great for the most part otherwise, loved the feedback that chassis provides, its almost telepathic in how clearly it talks to you

Snetterton!

add complexity: ['error', { max: 10 }] to your nodejs repo's eslint config i dare you

i cant snap oversteer and die this close to the trackday, i must resist

i wanna drive midship but its too cold

took a photo of Jupiter and it's moons

extreme track package
look inside
1750kg😴

@nepi@gts.nepi.gay this specific one was about the camaro zl1 extreme track package... and yes, that's actually what its called.

adding a toggle to my app called "hide generative ai content" that just closes the app

mounting my old RAID5 one last time before it gets paved over to make way as a backup storage to my primary server

you when i disneyfy you

RE: https://social.entan.gl/users/furrygirls/statuses/01K5M133MGW62ZE1FN3YCM1SV5

I don't mind implementing AI Features if they make sense, i'm the most knowledgable person at my work related to LLMs, but the fact that my direct advice then gets ignored is what irks me, shit is implement just because it might work

Theres a reason our calculators arent fucking AI Powered, it's because a hammer is not a very good screwdriver, and AI is terrible at spitting out numbers that make sense, fuck.

had to explain my boss and my direct superior how a token predictor works and that the feature they want just doesnt work with LLMs
their response "well implement it anyway, see how it goes"

I'm so tired.