killertron was their bot
#programming
fast pagoda
azure lynx
i've considered every judges decision against me a partial victory for being able to tank the other guy for the full 3 minutes.
sick owl
Oh my god he was in it from near the start
I never noticed what the hell
visual pawn
he sure looks like he'd be his dad
sick owl
Nah Chris wasn't on the team for this one apparently, there's no Chris in the name listings
Just Richard Broad
But that robot was a fan favourite so I seen he showed up for some of its many appearances
visual pawn
I wonder what the rules must be like for robot wars in terms of what you can build
sick owl
You know I miss Robot Wars, I tried watching BattleBots recently but the vibes just aren't there by comparison
visual pawn
I imagine there's a ton of room to bend the rules
sick owl
Like it's fine but it's just not as good TV
fast pagoda
chris at 14:25
sick owl
Its so crazy to me that that's the kid that would become the Chris Broad we know
visual pawn
it turns out everyone was a kid once
sick owl
You wouldn't be able to tell
visual pawn
unless they're not human
fast pagoda
people pop up all over the place
sick owl
Dude thanks for that trivia I'm gonna annoy my family with it now
fast pagoda
that's what i live for
azure lynx
unless they're not human
even they will have a mind that was shaped by thousands of people who were once kids.
so it's likely that the right prompting and self-reflection could result in beliefs identical to someone who was once a kid, with just the same feeling of truthiness as everyone else.
azure lynx
I wonder what the rules must be like for robot wars in terms of what you can bui...
i see these old series and they are so limited because they didn't have to same high density power sources we have now. they're using old battery tech like gas powered cars use, not the batteries like electric cars use. they're also limited by safety. the walls can only be so thick. between the lack of available power and the need for safety, they have some constraints. these days most robots use fast spinning weapons instead of swinging weapons, either mounted vertically or horizontally.
visual pawn
it'd be interesting if language was necessary for thought
obsidian mantle
We have words for most of things actually
Its not that language is necessary for thought. Its that with language we can describe thoughts
visual pawn
even they will have a mind that was shaped by thousands of people who were once ...
well, it'd still mean that not all AI models must have been a kid once
azure lynx
it'd be interesting if language was necessary for thought
not necessarily language but something like tokens. categories that could be used to reference gthings.
obsidian mantle
Its kind of what they do
visual pawn
i see these old series and they are so limited because they didn't have to same ...
new rules: the opponent isn't defeated until their LiPo battery is punctured
azure lynx
well, it'd still mean that not all AI models must have been a kid once
i think any which could communicate and understand you correctly probably could imagine having a childhood to some extent.
visual pawn
we're not really talking anything concrete ofc, but you still just mean one could imagine it, not that all experienced it; an LLM isn't really experiencing anything during training
well, maybe in reinforcement learning
obsidian mantle
new rules: the opponent isn't defeated until their LiPo battery is punctured
Cover the battery with 10cm lead plate and attatch wheels to it 
visual pawn
Cover the battery with 10cm lead plate and attatch wheels to it <:neuroTomfooler...
opponent covers them in a tarp and overheats them
obsidian mantle
right
azure lynx
new rules: the opponent isn't defeated until their LiPo battery is punctured
most events don't particularly like LiPo fires because of safety reasons. the organization who runs the events i go to doesn't mind as much because their arenas are designed to evacuate the smoke and they have house bots in the arena with fire extinguishers too.
and fire looks cool. (they also allow robots with flamethrowers)
visual pawn
a part of me just wants to see something like robot wars with zero regard for safety
azure lynx
no... you need safety. Its do able safely with bigger weapons.
visual pawn
they start getting outfitted with APFSDS rounds
azure lynx
it's just with 250lb robots, it's hard to do it safely with big weapons.
visual pawn
or high explosive anti tank ammunition
young plover
it'd be interesting if language was necessary for thought
Took a linguistics class in college and there was an interesting part about how human language can express things that other relatively intelligent animals like Orcas and Chimps can't.
https://en.wikipedia.org/wiki/Hockett's_design_features
Like talking about things which aren't currently happening.
Whether the ability to think about those ideas is tied to language... 🤷♂️
visual pawn
although I guess what I'm asking for is real, just outside of a sports context now =\
potent geyser
put the battle arena in desert.
obsidian mantle
Are they not using military grade actual combat robots out there yet
potent geyser
we will get Gundam in 10 years
visual pawn
Are they not using military grade actual combat robots out there yet
they are
and we've had drone on drone combat for a while now
azure lynx
actual "no holds barred" robot wars would be boring because it'd just be the first to hack the other robots.
obsidian mantle
Drones are lame
visual pawn
but there are ground drones in use now too
potent geyser
FPV drones are lame
visual pawn
and the ground drones have officially taken part in combat operations now
potent geyser
as transport?
visual pawn
no, combat
obsidian mantle
The future is now 
visual pawn
I believe there was even footage of a ground drone taking prisoners
but I assume saying details would break some kind of rule on this discord
visual pawn
the ground drones are given grenade launchers, mortars, and other small arms
azure lynx
first write the video game, then integrate your AI to be a player, and then some magic happens you swap out your game for a military contract and drink away your guilt with the millions they pay you. ;/
obsidian mantle
visual pawn
first write the video game, then integrate your AI to be a player, and then _som...
so, ender's game?
azure lynx
basically yeah, but for an AI and with no worse sequels.
visual pawn
there aren't any ground drones that I'm aware of that have seen combat that are actually autonomous though
potent geyser
feels like they should be quite effective in sabotage mission.
azure lynx
i gave up on my autonomous ground vehicle after it crashed ay high speed in to a curb it didn't see in a parking lot, destroying the mount for the sensor array that couldn't see things at 45 degree angles. it's probably worse navigating when you're larger than a cat.
afaik they're all driven by people.
potent geyser
Just wondering, does anyone have experience with bluetooth AoA?
azure lynx
what about it specifically? i'm a little familiar with some related topics (phased antenna arrays and microphone arrays) but not much more than the basics.
potent geyser
I was looking at the phased antenna arrays. wondering how accurate would it be
might be able to setup a decent Bluetooth based VR tracking system or something
that was one of my (failed) project from last year lol
azure lynx
i think you'd need to sample it with multiple objects at once to tell what sort of rates and resolution you'd get.
it might have too slow of an update rate when there are multiple devices or it might jitter a lot when you are still, or it might be quite granular, only giving you exact multiples of 6 inches for example.
it also might work perfectly.
there also might be ways of avoiding any of the other problems too, like smoothing multiple frames of data together or forcing some hard assumptions on where you can be when things don't make sense, or overclocking it or otherwise hacking some code to give you higher resolutions.
potent geyser
and use multiple different solution (a combination of video and hardware) to bypass noise.
but in the end, I have limited experience with hardware stuff. It was ratherhard to evaluate.
at least back then, I found several solution (with api) aimed toward warehouse level tracking
but 1 meter accuracy would be quite funny
azure lynx
1 meter accuracy when Neuro is driving a tank towards you is scary enough thank you.
potent geyser
you know what, a RFID based VR tracking system might be interesting
slap down 8 RFID reader. since it can calculate distance between reader and the tag. might be able to map everything with that boundary
visual pawn
what would be the tangible benefit over, well, a number of other similar options?
azure lynx
you could also just have lots of tags scattered around you and one reader, kinda like GPS maybe?
potent geyser
I am not sure if the reader can get the direction data. Therefore, it might need multiple reader with to calculate the location.
visual pawn
let's say uhh, in comparison to having a bunch of ARTags
potent geyser
and the tags are cheap
visual pawn
would the reader also need to be pretty beefy?
how well would it scale compared to, say, EM tracking?
potent geyser
how well would it scale compared to, say, EM tracking?
right, I forgot about that aspect of the problem. I guess a tag would not transmit the orientation data
I guess a $0.1 to $1.5 tag wouldn't cut it after all
azure lynx
cheap tags would just echo the pulse which activated them back, potentially modulating it. but the sender/receiver is getting multiple signals back with different strengths that it can use to determine distances and then use those and some math to work out where it is in relation to all the signals.
so the tracker would need some level of processing power
potent geyser
I think it has build in RSSI
on the tracker side
and they are plenty of off shelf tracker cost around 15-30 USD
but we don't have any orientation data relating to the tag
azure lynx
i know there are algorithms in computer vision which take 8 or more points from multiple directions and use that to construct what those postions and directions were. i think you'd need a similar number of passive points that you're tracking to be able to reliably track things. if you have other constraints you could reduce that.
the easiest way to get orientation data is to use multiple trackers at one location.
with 3 trackers in a triangle you know where they all are and you know what their angles are to each other, so you know what direction they are facing
potent geyser
what happens if the tag rotates at the same XYZ location
azure lynx
multiple tags i means. sorry.
potent geyser
worth a shot
true hemlock
visual pawn
i know there are algorithms in computer vision which take 8 or more points from ...
4 at minimum
but the more you have the more reliable, especially as some points get obstructed
azure lynx
i think the algorithm was intended for noisy data.
visual pawn
a decent optical tracker with 4 passive markers can get you sub-millimetre accuracy though
potent geyser
and cloth/body will obstruct signal strength as well
visual pawn
a decent optical tracker with 4 passive markers can get you sub-millimetre accur...
the sub-mm ones I'm talking about are $30k minimum though
azure lynx
VR is only fun until your arms get tired from all the ill thought out user interfaces.
visual pawn
the gorilla arm problem
potent geyser
the sub-mm ones I'm talking about are $30k minimum though
damn, I tried optical tracking as well. I mean cheap webcam based solution
visual pawn
people want to have the sci-fi arms waving in the air interfaces, but they don't realize how tiring those get very quickly
visual pawn
damn, I tried optical tracking as well. I mean cheap webcam based solution
yeah, the ones I'm referring to are things like this: https://www.ebay.com/itm/167701345937
potent geyser
and then I realized with low end optic tracking, I am not going to beat any of the existing solution lol
visual pawn
but they're also for an industry where a single disposable Electromagnetically tracked needle is a couple thousand dollars
potent geyser
Latency 17 ms
visual pawn
these are for tracking surgical tools, hence the sub millimeter accuracy
azure lynx
you could combine optical tracking with accelerometer and gyroscope data, using the optical to detect the true horizon and direction.
potent geyser
for sure
visual pawn
but you can get the "at home" version of this tracker using a webcam and paper markers
we've had high school interns demonstrate feasibility using printed out paper trackers before using the actually expensive tracker
because high school interns are amazing at destroying >$20k equipment
potent geyser
With multiple camera?
azure lynx
you can do stereo cameras with opencv quite easily.
visual pawn
multiple webcams was what I've seen but
some algorithms and methods can work with one
azure lynx
the hardest part is getting them synchronized but for things which aren't moving much being 1/2 a frame out of sync usually isn't a problem and tends to fix itself when it is.
potent geyser
I do remember reading a paper relating to mutli-camera setup
potent geyser
you can do stereo cameras with opencv quite easily.
interesting maybe I should revist that
visual pawn
I believe if you have certain fixed setups you can pull off ArUco marker-based tracking without multiple cameras
azure lynx
if you have a moving camera you can try doing monocular-SLAM stuff but that was all academic papers. it basically does fake stereo vision by comparing between frames where the camera has moved and identifying the same points and by using parallax it gets depth and after a few samples it knows the position.
visual pawn
I used to see SLAM everywhere but I feel like it's all been replaced by AI/deep learning
azure lynx
maybe there is an AI version of it now that works better
visual pawn
and frankly, deep learning solutions have done way better
without writing an essay, one project we've kept coming back to is doing tracking of a bronchoscope using only monocular vision
because multiple cameras increases the size of the bronchoscope significantly and EM tracking can get expensive and annoying
and we used to try doing it with SLAM or Shape from Shading but they've never been better than incredibly jank
it actually feels pretty feasible with AI though
and AI-based solutions cut through artifacts like mucus and bleeding a lot better
azure lynx
i think the problem is likely specular highlights coz of the nature of the surfaces inside of people and where the light has to be coming from so you can see. any reflections of light that change based on angle are difficult to model and acount for.
visual pawn
although more recently we've relied on reconstructing the image in 3D from a monocular image using GAN and then going from there, but people have tried all-in-one-model solutions for tracking that way
AI based solutions also deal pretty well with specular highlights, but even then, specular highlights can be removed with traditional algorithms pretty well
the bigger issue tends to be movement from breathing or when mucus just completely fills up an airway
potent geyser
sounds like a classification problem
visual pawn
you also are always dealing with a single known light source
azure lynx
movement between frames also screws with traditional SLAM coz it makes position paradoxes "that one dot can't be in two different places..."
visual pawn
yeah, and SLAM especially sucked in this environment because features are just less distinct
we barely got SLAM working in perfectly still and dry silicone rubber models
azure lynx
they'd be all "no, i meant the other greyish-pink colored bit next to the blood drop. no... the.. whatever."
visual pawn
and by barely I mean we made it past the first "intersection" before it got utterly lost
images of pig lungs in there if anyone is squeamish
oh, it doesn't reference the earlier SLAM work; that's an earlier AI-based method
Training took about 8 hours for cGAN and 30 hours for 3LcGAN, and less than 0.02 s for the inference of one depth map on a GTX 1080 Ti GPU for both models.
Harvard med school affiliated AI research being done on a 1080 Ti
visual pawn
yeah we did it during covid
potent geyser
and 1080ti
visual pawn
he never asked for anything better
visual pawn
but medical AI models tend to be way, way, wayyyy smaller than anything like an LLM
and Transformer architecture hasn't really made much of an impact either yet
we're still mostly using CNNs and RNNs
potent geyser
same in the financial sector
visual pawn
I've tried training transformer models but they just perform worse for way more resources so far
potent geyser
after all latency matters a lot. but I am the development ops side. so I don't have to worry about that
visual pawn
I think a part of it is that our training data still depends on very tedious manual labour for labeling
we may have a wealth of medical data, but it's not labeled
so you get maybe a few hundred labeled images to train a model
oh, when I say labeled I don't mean like, just a classification
but like, traced organs on images and stuff
I have been getting some pressure from higher ups to incorporate language models in something but I still can't find a decent use-case
potent geyser
we made a internal chat bot with access to company libraries. so maybe that
and call it a day
visual pawn
to throw a bone, I tried throwing in a little chat bot that tries to explain what's going on in a procedure
but it did such a bad job that Neuro probably could've done better
potent geyser
sounds about right.
visual pawn
I know some other labs have been trying to do surgical procedures purely via vision-language-action models, but I can't say how well it's really working yet
like, they have enough to make highlight reels, but that's it
mighty thorn
we made a internal chat bot with access to company libraries. so maybe that
Me in a security risk competition
visual pawn
well who's going to do something silly like expose the internal chat bot to someone outside huh
no way
jagged turtle
language colour coding ded lmao
potent geyser
Me in a security risk competition
it's ok, it is an isolated instance or something
olive sable
Intel i9 12900kf
RTX 3090 24Gb Zotac
32Gb RAM
1TB ssd
for 1100 euro 
mighty thorn
Intel i9 12900kf
RTX 3090 24Gb Zotac
32Gb RAM
1TB ssd
for 1100 euro <:hmm:13031...
Why do you have so much money
visual pawn
but also sometimes you just gotta make a dumb demo so they'll stop asking you to "put chatGPT in something"
olive sable
Why do you have so much money
social security
potent geyser
but also sometimes you just gotta make a dumb demo so they'll stop asking you to...
then they will ask you to make an agent
visual pawn
then they will ask you to make an agent
I've long since learned that there is a class of people where the second you say "it's AI-powered" they'll be happy
the people who even know what an agent means aren't the ones I need to worry about
although more recently I've been making a pitch of "creating hybrid systems combining traditional and AI-based methods"
which usually translates to doing the same task two different ways and taking the result that is better
olive sable
Intel i9 12900kf
RTX 3090 24Gb Zotac
32Gb RAM
1TB ssd
for 1100 euro <:hmm:13031...
shoudl i buy it?
mighty thorn
which usually translates to doing the same task two different ways and taking th...
I’m so glad ai has enabled us to construct an or gate
mighty thorn
shoudl i buy it? <:hmm:1303107376295186524>
Do you need it?
olive sable
not really, but pc parts are kinda like the stock market rn
visual pawn
do you believe the "5090 to be $5000 in a year" rumours?
olive sable
i need to buy while its still low
olive sable
do you believe the "5090 to be $5000 in a year" rumours?
not 5K
but i can def believe 4K
potent geyser
but can you resell it for 4k
olive sable
im not gonna be reselling 5090s, so no
i want the 3090s while they're still fresh and cheap
potent geyser
also, we have this->
U.S. approves Samsung, SK Hynix chipmaking tool shipments to China for 2026: Reuters
mighty thorn
Okay chat I go sleep now
visual pawn
I’m so glad ai has enabled us to construct an or gate
well, one method being AI and the other being traditional and then an OR gate, but I've seen people call a single If statement "AI" so
mighty thorn
well, one method being AI and the other being traditional and then an OR gate, b...
Due to our new ai based filtering, we’ve enabled our users to filter out products above or below a certain threshold
jagged turtle
Okay chat I go sleep now
how did you manage to do that lmao
mighty thorn
how did you manage to do that lmao
(not my screenshot)
jagged turtle
(not my screenshot)
ah
olive sable
siemens mentioned
jagged turtle
Neither is this
oh yeah ik people have been memeing that to death atp
olive sable
lesgooo
mighty thorn
Me when the new flavor of underwater fiber optic cable drops
olive sable
nom nom nom
visual pawn
I've yet to use a single Siemens medical system UI that seemed like it was designed for humans
mighty thorn
Me when the new flavor of underwater fiber optic cable drops
They should use the strat that people use on cats where they soak things in hot sauce so cats don’t touch
But with fiber optic cable
visual pawn
sometimes I feel like half the job of an MRI tech is to figure out what is going on in the software interface: https://www.siemens-healthineers.com/magnetic-resonance-imaging/advanced-imaging-applications/syngo-via
olive sable
I've yet to use a single Siemens medical system UI that seemed like it was desig...
its not designed for humans, its for germans
mighty thorn
sometimes I feel like half the job of an MRI tech is to figure out what is going...
That’s the least dysfunctional medical software I’ve ever seen
Lmao
visual pawn
that's until you find out that everything functions via click and drag
mighty thorn
I’ve seen far worse
visual pawn
I suppose you're right
I worked with someone that I.....think worked at Siemens? if not another scanner company
and they claimed that one software imported VTK something like several dozen times, just different versions of it
because each team decided they needed a different version
when I was a student, I used to complain about how convoluted the most popular open source medical imaging software we use was
but then I saw the source code for some propriety software for the same purpose and I never complained about our open source one ever again
potent birch
midnight sigil
potent birch
iirc gd is like py
now the only thing i need to do with this movement is a smooth climb on a small uphill and jump
azure lynx
but then I saw the source code for some propriety software for the same purpose ...
at least when you complain about an open source thing someone can see the problem and fix it themself. not always what happens with proprietary stuff.
potent birch
Holy fuck good thing I version controlled and saved my godot project, power just went out
brisk shore
chat am i cooked
potent birch
Holy qr code
visual pawn
at least when you complain about an open source thing someone can see the proble...
yup
azure lynx
i have a t-shirt with a qrcode like that on it: it crashes some cameras.
rigid snow
iirc gd is like py
it’s much nicer than python
potent birch
Gdscript is easier for me thats too lazy fr fr
I download the one with .net framework but idk what that does
Is it easier for it to use for Neuro integration?
I might yap here while doing this so I don't get bored and procrastinate lulw
visual pawn
I have no idea about neuro's godot integration, but there's documentation
looks pretty straightforwards to enable the plugin and take a look at it
olive sable
<:vedalCry:1234565975492264018> <@851857878876028928> i read "usage" in japanese...
bunny? 🐰
visual pawn
so there's no way in fuck I could convince a medical device company to volunteer their robot for a neuro stream, but I wonder if there's hypothetically a way to do a neuro performs surgery stream using open source hardware/software and have it be interesting
covert bane
Do you guys think, using a file for communication between two processes (and calling it IPC) is scuffed?
tender river
Do you guys think, using a file for communication between two processes (and cal...
it normal as long as its socket file
covert bane
like, a .json file for sending a message to one process. and a .response file for receiving a response
tender river
ye you'll run into issues that way
covert bane
<:vedalShrug:1433173412217356389> it normal as long as its socket file
whaat... is a socket file?
visual pawn
I could set up the legit kind of software they'd use for neurosurgery along with a medical phantom/model, and then I could add Neuro integration to some robotic arms, but I dunno how interesting "Neuro jabs needles into a brain" would be
covert bane
ye you'll run into issues that way <:vedalShrug:1433173412217356389>
the current issue I can think of is communications being able to be leaked
so like, anyone can access it
and talk to the processes
if they code it right
visual pawn
idk if windows has it
it's a Unix thing only I think
covert bane
would there be any other issues?
tender river
would there be any other issues?
synchronization
no easy way to tell if the process has finished writing the file
you have to include something like html content-length or something similar
stupid method either way just do proper ipc
visual pawn
Do you guys think, using a file for communication between two processes (and cal...
do you mean like
one program writes to a file, and the other reads it?
covert bane
synchronization
https://github.com/cassitly/neuro-desktop/blob/dev/desktop/apps/neuro-desktop/src/ipc_handler.rs#L119
it has been working fine so far
covert bane
one program writes to a file, and the other reads it?
yea
visual pawn
because that actually works fine on Windows; I've done it in the past when lazy
covert bane
but now I need bidirectional comms
visual pawn
actually, I've found it works better on windows than on linux
covert bane
and there are multiple processes that will start in my program
I need one for python now
visual pawn
I dunno about bidirectional though, I've only done it as a temporary solution for one-way communication
tender river
https://github.com/cassitly/neuro-desktop/blob/dev/desktop/apps/neuro-desktop/sr...
the code has data races
covert bane
<:neurowheeze:1126151823925977220> the code has data races
really?
tender river
ye
covert bane
where are they, and what are the conditions?
azure lynx
i've used that general method for updating parameters in a deep learning art project. (writing to a file which the process polls for changes and reloads)
covert bane
oh
covert bane
conditions of data races always random
well, how can they be caused
tender river
and the data race is you might not finish writing the file before the other program starts reading it
covert bane
I would want to fix it
covert bane
and the data race is you might not finish writing the file before the other prog...
so a .lock file is best?
wdyt?
no just dont use files
covert bane
<:vedalDepress:1343810731270864976> no just dont use files
I just don't know how to implement proper IPC
I don't want to link stuff too
tender river
I just don't know how to implement proper IPC
then do research 
covert bane
too complicated
azure lynx
i had to add a delay of a few hundred milliseconds after it detected a change before it started reading, coz the program i was editing with sometimes didn't write it all at once.
but that was only safe because it was human edited
(me)
covert bane
then do research <:vedalCry:1234565975492264018>
where would you recommend I research?
tender river
i would recommend you to just use grpc 
azure lynx
windows IPC [language name here]
tender river
where would you recommend I research?
and ye it should be trivial to find if you have basic google skills
rigid snow
windows has bsd sockets too iirc
visual pawn
your issue will be more about being spoiled for choice
olive sable
and the data race is you might not finish writing the file before the other prog...
never had that issue with the json file i kept my artist and video title info in for the media-player io. 
tender river
windows has bsd sockets too iirc
i remember using pipe files or something :mlntcandy:
visual pawn
so as long as you can make a decision, you can do it
azure lynx
windows has kinda bsd sockets... can't treat them like files though. ;[
olive sable
why cant you just have the programs directly output to each-other?
visual pawn
I don't
"Hey Claude Opus 4.5, can you write some basic IPC for me and recommend something to use?"
covert bane
why cant you just have the programs directly output to each-other?
how do I do that?
covert bane
"Hey Claude Opus 4.5, can you write some basic IPC for me and recommend somethin...
claude is expensive, I try to code myself as much as possible. And use claude whenever I've spent hours on a issue I can't fix
visual pawn
claude is expensive, I try to code myself as much as possible. And use claude wh...
"hey chatGPT"
covert bane
"hey chatGPT"
GPT5.2 is just stupid
tender river
how do I do that?
i really recommend you to start using google more
rigid snow
windows has kinda bsd sockets... can't treat them like files though. ;[
not on the fs level but you can treat ythem as file handles
potent birch
https://github.com/VedalAI/neuro-sdk/blob/main/Godot/README.md
https://github.c...
I'm too newbie for that atm, gonna read that when I'm sure I can implement that on my own game
olive sable
i know that for python i could just print whatever i needed as output, and the host program could get it like that
visual pawn
something like this is simple enough for it to not get it wrong most of the time, and it'll be obvious when it's wrong
covert bane
<:vedalCry:1234565975492264018> i really recommend you to start using google mor...
You guys are my google-
rigid snow
bad
tender river
You guys are my google-
google faster than asking people
covert bane
i know that for python i could just print whatever i needed as output, and the h...
ooooh, that is smart actually
amber fractal
chat is only google if it is Nix code 
visual pawn
I'm gonna assume this is universal, but generally no-one likes answering a question that has the premise of "I tried nothing, what do I do?"
olive sable
for my waybar i do i with prints here
covert bane
I'm gonna assume this is universal, but generally no-one likes answering a quest...
true, I should stop doing that. sorry tho
visual pawn
I mean, technically you did say you were trying to write to files so
you didn't quite try nothing
olive sable
for my waybar i do i with prints here
it works 
covert bane
you didn't quite try nothing
I was thinking about what I could improve on my code
so mmphm
tender river
it works <:evilShrug:1131357858101993603>
you are only reading the output when the writing is done
presumably
olive sable
probably
tender river
and if not its still fine because its not a request-response architecture
olive sable
true
finite sphinx
guys does this count as programming 🥀
azure lynx
looks like it's a bunch of if statements. 7 segment LED ?
rigid snow
guys does this count as programming 🥀
your mistake is assuming this channel is for programming
azure lynx
now that LLMs use natural language for input and output, programming languages are very variable.
finite sphinx
looks like it's a bunch of if statements. 7 segment LED ?
it's 8 1-bit memory cells
visual pawn
are you going to start the "is vibe coding programming" debate?
finite sphinx
all this for a single byte
finite sphinx
it's 8 1-bit memory cells
i edited this message 3 times i'm crine
visual pawn
and also, if setting up the Difference Engine counts as programming, then that is as well
azure lynx
nice. also: I'm so sorry for the time you spent on that which you'll never get back.
visual pawn
programming doesn't even require it to be for a device that uses electricity
finite sphinx
nice. also: I'm so sorry for the time you spent on that which you'll never get b...
time well spent 👍
visual pawn
like half of this field: https://en.wikipedia.org/wiki/Unconventional_computing
I took an unconventional computing course in uni and we spent time on programming for a theoretical computer made from a time travel machine
azure lynx
i like the implementations of Conway's Game of Life made using virtual computers made from an implementation of the Game of Life.
tender river
i like the implementations of Conway's Game of Life made using virtual computers...
oh speaking of game of life
they recently invented a 1 block tall automaton that moves 
called spaceship in game of life terminology
visual pawn
how many blocks wide"
visual pawn
I can't find a video, just a forum post
visual pawn
2016 needs to be the year we discover this.
bit off
tender river
it has a bounding box of 3707300605x1
takes 133076755768 generations to move by 2 cells 
This ship consists of a unidimensional pattern that is a predecessor to a universal constructor. The process of creating a unidimensional universal constructor predecessor was the relatively simple part of the design; having the universal constructor then resynthesize the unidimensional pattern was much more difficult, as all sparks had to disappear before the ship entered its unidimensional phase. Due to symmetry, any one-cell-thick spaceship must move in the direction its length points in.
mhm i totally understand
pretty
finite sphinx
<:vedalWow:1343810742989623296> pretty
manipulating biblically accurate angels to do addition
azure lynx
luckily for the real universe there isn't such a simple universal underlying mechanism. probably.
tender river
low level processes are relatively simple and thats the beautiful part of high level processes 
covert bane
variable Configuration with value Release False False False would no longer be valid.
At line:7 char:5
+ [ValidateSet('Debug', 'Release')]
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : MetadataError: (:) [], ParentContainsErrorRecordException
+ FullyQualifiedErrorId : ValidateSetFailure
huh, it was just working fine a second ago. Powershell is definitely messing with me, I feel as if there is a ghost inside my pc
olive sable
sage crag
olive sable
sage crag
tender river
💥
sage crag
tender river
squish
sage crag
tender river
ermsquish
lament igloo
This is truly peak #programming
sage crag
promot
jagged turtle
olive sable
exponential ermfish
sage crag
tender river
rrrr python closure bad 
jagged turtle
what the fuck is this channel now
sage crag
tender river
what the fuck is this channel now
olive sable
best channel in neurocord
tender river
<:vedalErmFish:1234565978780860467>
<:ermFishL:1145682428325810176><:ermFishR:1...
exponential ermfish vs exponential ermsquish
olive sable
i dont need to know
tender river
olive sable
since #osu is ded all thats left is #programming to be the best
sage crag
<:vedalErmFish:1234565978780860467>
ermsquish takes 4 iterations and ermfish takes 3 iterations
even though they both exponentiate by two, ermsquish is slower
jagged turtle
<:glueless:1282337396230328425>
why am I
tender river
3 iteration to halve and 2 iteration to double
jagged turtle
did I run out of glue
tender river
ye
tender river
3 iteration to halve and 2 iteration to double
me head hurt
tender river
me have me own
tender river
i just stocked up on glue
clearly not
jagged turtle
I just checked and I thought I did have glue but it turns out I actually am glueless
sage crag
ok it double every 6 iteration <:vedalErmFish:1234565978780860467>
i not ask questions
ohhh my bad
the curve is pow(2, x/6)
tender river
ye me lazy just simulate in python
import math
n = 1
for i in range(1002):
if i % 2 == 0:
n *= 2
if i % 3 == 0:
n //= 2
print(math.log2(n))
jagged turtle
wait whaaaa
azure lynx
why are you doing integer division by 2 when it's a multiple of 3?
sage crag
i is unrelated to n
covert bane
any opinions on if it's normal to have the same block of code across every action like this?
thinking about if I should refactor and fix that
tender river
dont like it move to a separate function
jagged turtle
thinking about if I should refactor and fix that
you're just describing modularisation lmao
covert bane
<:vedalShrug:1433173412217356389> dont like it move to a separate function
subfunction inside execute_command?
tender river
whatever you want
azure lynx
execute_and_clear or something
tender river
code refactoring is primarily for you to maintain the code
not for some abstract notion of clean code
sage crag
wrr
quick condor
Gaslighting gpt has become a new hobby
dusty niche
Why does AWS have to be a pain in the ass to setup postgres on 💔
sage crag
wrr
import matplotlib.pyplot as plt
import numpy as np
x = np.arange(100)
plt.plot(x, 2.0 ** (x // 2 - x // 3))
plt.plot(x, 2.0 ** (x / 6))
plt.show()
If you can't already tell, I like modularity.
my codebase compiling into multiple binaries, so I don't have to recompile the entire codebase to test a fix I made on something
sage crag
wrr
covert bane
jagged turtle
If you can't already tell, I like modularity.
*my codebase compiling into multip...
hot reload when
covert bane
hot reload when
hmm, great idea!
dusty niche
airisprod
covert bane
I'll pin it
jagged turtle
nah I'm kidding, hot reload normally recompiles the entire thing anyways
sage crag
```py
import matplotlib.pyplot as plt
import numpy as np
x = np.arange(100)
pl...
just to be clear its the same
jagged turtle
but like having multiple binaries is just... weird
covert bane
nah I'm kidding, hot reload normally recompiles the entire thing anyways
hot reload for specific functionality could be possible
tender river
```py
import matplotlib.pyplot as plt
import numpy as np
x = np.arange(100)
pl...
x/2-x/3 is equal to x/6
covert bane
but like having multiple binaries is just... weird
hot reload compatible, it was future proofed for this!
covert bane
I didn't even think about that
sage crag
modular arithmetic always cause interesting
patent shard
good work, Discord 
jagged turtle
hot reload for specific functionality could be possible
i mean right now I'm still trying to figure out how best to design a system that allows me to defer neurontainer's backend stuff to a local process which is running hono dev or smth like that
tender river
<:NeurOhISee:1136209034966483084> x/2-x/3 is equal to x/6
x/3*6-x/2*6
patent shard
Discord is truly in Discord
dusty niche
https://files.catbox.moe/56zqv2.mp4 I like my project tbh
covert bane
i mean right now I'm still trying to figure out how best to design a system that...
I don't get what you mean, but try going to bed or smth. I best get my ideas right before I fall asleep :D
covert bane
<:vedalShrug:1433173412217356389> dont like it move to a separate function
what should I even name the func?
covert bane
I really dont know what to name it
tender river
pick whatever helps you maintain it
rigid snow
what should I even name the func?
jagged turtle
I don't get what you mean, but try going to bed or smth. I best get my ideas rig...
ok so essentially, the vm backend for #1452618407899041792 is also written in js, but you can't test it in docker desktop without rebuilding the container image and updating it, whereas you can for the ui frontend, so I'm trying to figure out if theres a feasible way to be able to defer stuff to hit a locally running process instead
rigid snow
wait WHAT did you develop
what the fuck
why
what do you mean docker engine integration
covert bane
ok so essentially, the vm backend for <#1452618407899041792> is also written in ...
hmm, ummm, make it its own binary?
wait no
its in JS
covert bane
yea but it's just weird tbh
quick condor
https://cdn.discordapp.com/attachments/1072697081443131476/1355617829939974144/e...
NGL there is something incredibly funny about the idea of the swarm silently putting this into codebases in places that wolnt be descovered for years, only for some poor soul to Google it only to find that picture
covert bane
then you'd have to recompile that everytime as well
jagged turtle
the problem is that in order to a new version of the backend, I would have to rebuild the extension container image, tag it as latest-dev (need to go change that actually) and then update the extension container
covert bane
ummm, gimme a moment to think
covert bane
the problem is that in order to a new version of the backend, I would have to re...
why not separate the image to a server and client model?
the server should just contain boilerplate code or smth that allows the client to connect and do stuff with Docker
the backend can be a server and client model
visual pawn
any opinions on if it's normal to have the same block of code across every actio...
instead of deciding based on some notion of what you "should" do, instead decide based on "will this be annoying as hell in the future to change for either me tomorrow or me 4 years from now?"
covert bane
the server is the actual extension container image
covert bane
the problem is that in order to a new version of the backend, I would have to re...
and the client is, well, whatever the you need it to be
visual pawn
if you think you're going to be change what's in that block of code and having to copy that across all of those instances every time for many timnes, then yes refactor it
jagged turtle
why not separate the image to a server and client model?
wdym separate the image to a server and client model
it already is
covert bane
(I haven't quite read the codebase, so I'm just guessing)
covert bane
wdym separate the image to a server and client model
No, the backend is it's own little server and client
covert bane
so you just have to compile and do all of that for the server
but not for the client
jagged turtle
that... is not how it works
covert bane
so like, you only have to do it once
visual pawn
NGL there is something incredibly funny about the idea of the swarm silently put...
I've dealt with someone's codebase where everything was named after inside jokes, it was awful. And worst of all, it was all in japanese
covert bane
that... is not how it works
I haven't read the code, so I did say I was just guessing
covert bane
instead of deciding based on some notion of what you "should" do, instead decide...
ehhh, it will be claude's problem not mine. When I find something annoying I have to rewrite or something like that. cause I am not dealing with stuff like that :D
jagged turtle
which is why I was thinking of some rpc-like model that connects to a local instance which may be running on your computer
and so any requests sent to the backend through the frontend will just be directed to the locally running instance instead
covert bane
which is why I was thinking of some rpc-like model that connects to a local inst...
make it multiple binaries then :D
solves your problem
jagged turtle
making it binaries doesn't solve the problem!!
covert bane
oh- right-!
I'll go read the codebase then
so I understand better :D
sorry if I annoyed you
jagged turtle
idt it's a codebase issue
it's more of a docker desktop extension develoment process
covert bane
oh. well, I never made an docker extension, so I don't quite know
patent shard
good work, Discord <:NekoEllaOk:1218012705797963796>
I also like how it just pops up random threads
jagged turtle
sorry if I annoyed you
making neurontainer has costed me more of my sanity than this yap session
jagged turtle
oh. well, I never made an docker extension, so I don't quite know
yeah, I thought it was gonna be easy too
jagged turtle
why
also if you were wondering for an actual answer: I thought it would be funny for neuro to be able to spin up pytorch or nginx at will
patent shard
okay, it fixed itself
that was weird.. so before it was two across and now it's five? and it also started showing the oldest ones but you scroll down and it said older posts and just showed everything that was deleted
truly Discord 
rigid snow
what would she do with nginx
visual pawn
also if you were wondering for an actual answer: I thought it would be funny for...
so neuro can start training her own models and achieve AGI right?
rigid snow
start training
jagged turtle
what would she do with nginx
idfk I just thought the idea was funny
covert bane
and so any requests sent to the backend through the frontend will just be direct...
I have an idea. So what needs to be reinstalled is the docker extension image right?
jagged turtle
I have an idea. So what needs to be reinstalled is the docker extension image ri...
yes
rigid snow
congrats joke is complete
now please spend your time on something better
tender river
neuro task manager integration
visual pawn
well, it's gotta be more useful than the neuro integration I'm arguing with myself over whether to do or not
jagged turtle
congrats joke is complete
who said it was a joke 
jagged turtle
well, it's gotta be more useful than the neuro integration I'm arguing with myse...
which one was it
jagged turtle
neuro task manager integration
good idea!
visual pawn
which one was it
rigid snow
neuro task manager integration
*kills obs*
jagged turtle
https://openigtlink.org/
:Looking: what the hell am I looking at
rigid snow
visual pawn
:Looking: what the hell am I looking at
standardized protocol for medical imaging and robotic surgery and general operating room integration
rigid snow
neuro doctor
jagged turtle
standardized protocol for medical imaging and robotic surgery and general operat...
Ah fun, neuro controlling hospital equipment, what could POSSIBLY go wrong here?
visual pawn
well, the idea was maybe Neuro could do Neurosurgery
tender river
lobotomy
visual pawn
I have the skull/brain models, I have the simulation software, I....might be able to convince someone to lend the equipment but I could make something open source instead
jagged turtle
\*kills obs*
someone seriously said in the neuropilot thread to start tracking if obs was on and start using the more invasive but vedal-proof ui
I think
so it would be funny to just hand obs control to neuro
rigid snow
what could ever go wrong
covert bane
yes
why not make the extension image a server that exports the Docker Extension API, and then you can make another program that is the client, and connects to the server. And uses the exposed API, so you just have to recompile the client. And not have to reinstall the extension
visual pawn
but worst case if there's no way to make a neuro neurosurgery stream interesting or workable, I can convince my boss I did language model surgery integration so he can go get funding off of AI hype
jagged turtle
what could ever go wrong
vedal I should show chat this screen reveals vedal's credit card info during amazon stream
jagged turtle
why not make the extension image a server that exports the Docker Extension API,...
...what?
rigid snow
also someone was thinking of, or making a thing so neuro can "make music" and i suggested giving her splice and letting her put loops together. i remembered stable audio open exists
jagged turtle
now please spend your time on something better <:vedalCry:1234565975492264018>
also for a split second I was considering neuro-browser but then I realise there's gonna have to be a way in which filters will be applied somehow sooooo
visual pawn
also for a split second I was considering neuro-browser but then I realise there...
Im sure some off-the-shelf parental control software would work
rigid snow
ye browser is a bad idea but somehow still 100x better than docker
jagged turtle
also someone was thinking of, or making a thing so neuro can "make music" and i ...
people are still pushing for strudel surprisingly, thought that discussion would have been forgotten a while ago but guess not?
covert bane
...what?
like, the part that you have to install into docker. You make it a server that exposes docker api, and you have a separate app that uses those expose APIs for your actual impl. So you don't have to recompile / reinstall everytime you make a change to the backend
jagged turtle
ye browser is a bad idea but somehow still 100x better than docker
no I'm sure neuro being able to execute code in a docker container is good idea and I see nothing wrong with it 
(if you can't tell some people irl have been making me deranged as of late)
jagged turtle
like, the part that you have to install into docker. You make it a server that e...
tl;dr make the user run a binary directly on their computer, how about no
covert bane
tl;dr make the user run a binary directly on their computer, how about no
but that solves your problem no? I also don't understand wdym by that
rigid snow
people are still pushing for strudel surprisingly, thought that discussion would...
strudel would probably require a massive prompt for her to even understand how to use it, it was niche enough until recently i doubt it's in her base model
at all
covert bane
like the actual neuro integration is a docker image?
jagged turtle
but that solves your problem no? I also don't understand wdym by that
the problem is that it's just not friendly to have to start up docker and a random ass binary just to use the integration
jagged turtle
like the actual neuro integration is a docker image?
the entire thing is a docker image
jagged turtle
strudel would probably require a massive prompt for her to even understand how t...
yeah that was my concern
we'd have to do a big explainer of what strudel/tidalcycles is
rigid snow
ye
covert bane
the problem is that it's just not friendly to have to start up docker and a rand...
that's what I think modularity is (neuro-desktop will have 3 binaries lol)
jagged turtle
that's what I think modularity is (neuro-desktop will have 3 binaries lol)
maybe don't have it like that 
covert bane
maybe don't have it like that <:neuroSob:1069255117427200091>
why? Linking is hard for me
jagged turtle
tidalcycles could work
tidalcycles and strudel have the same syntax and stuff, no?
jagged turtle
that's what I think modularity is (neuro-desktop will have 3 binaries lol)
also no, modularity should be during development, you shouldn't need to run 3 separate binaries to fully activate a program
rigid snow
one is a haskell dsl the other is a js dsl
jagged turtle
yeah true
covert bane
maybe don't have it like that <:neuroSob:1069255117427200091>
there's a binary that manages the other binaries. An binary that connects and manages the connection to the neuro backend. A binary that manages the controller (written in python), that controller gives neuro control of windows thru mouse and keyboard.
Doesn't seem that bad
rigid snow
but ye both still dsl
jagged turtle
but the uzulangs have the same syntax across both
covert bane
also no, modularity should be during development, you shouldn't need to run 3 se...
oh
well... that was my plan-!
rigid snow
one is a haskell dsl the other is a js dsl
so no not same syntax at all
jagged turtle
there's a binary that manages the other binaries. An binary that connects and ma...
no, this isn't what I want to do, this means that when someone installs my extension, they would also need to run a separate binary which is just kinda sus??
jagged turtle
so no not same syntax at all
what really?
covert bane
no, this isn't what I want to do, this means that when someone installs my exten...
ooooh. I see
rigid snow
have you seen haskell 
jagged turtle
have you seen haskell <:mlntcandy:1455273822625333375>
I meant like, composing the music patterns themselves
rigid snow
syntax different 
jagged turtle
huh I didn't realise that
tender river
pattern syntax same but rest of syntax different
jagged turtle
wait no- that gives the same issue
also on top of that issue, another issue is that one extension can only be assigned to one image
tender river
just assume that if there isnt tons of training data on the internet neuro cant do it
covert bane
also on top of that issue, another issue is that one extension can only be assig...
I wouldn't have known
rigid snow
just assume that if there isnt tons of training data on the internet neuro cant ...
that's what i said yep
visual pawn
"but just have Vedal teach her like a real child"
strudel is probably out, but surely there must be some kind of musical notation of some kind in her training data?
if it really needs to be in strudel, it probably isn't that difficult to prompt a more recent model to convert whatever she makes into strudel
and maybe with some additions intepreted from a description she gives or something
oh, tidalcycles, I see
rigid snow
there is 0 point making an llm write any musical notation
visual pawn
well it doesn't have to be good
rigid snow
i'll try to explain
if there is a thing you can just use like she did with the images she made with python, then sure
if we're building an integration from scratch what's the point
visual pawn
I've done music theory long ago, if there's any data out there in text form it should be possible for it to be at least somewhat structured
sage crag
hollow surge
sage crag
rigid snow
I've done music theory long ago, if there's any data out there in text form it s...
ye even just A4 A4 - C#4 - E3 E3 she will be able to do
or something like that
visual pawn
like I bet if you asked Neuro to do a chromatic scale she'd be able to give something
olive sable
banana campaign but im accidentally painting everything red 
hollow surge
interesting
rigid snow
You can turn this off
hollow surge
which is good
rigid snow
google everything opt out
visual pawn
well, claude 4.5 sonnet managed to give me....something in strudel
visual pawn
yeah but I'm saying Neuro could prompt a newer model
and have it try and interpret whatever notation she might come up with
lime stag
Just saw megalag new video Bruh ain't that a security risk and breach the extension is malicious as hell
rigid snow
grass 2027 campaign
sick owl
well, claude 4.5 sonnet managed to give me....something in strudel
Gemini 3 Pro or Opus 4.5 would do a lot better
rigid snow
bocchi 2028
rigid snow
overcomplicating
probably
no, definitely
there isn't a precedent of her having access to a bigger model
visual pawn
well, this is only assuming that Neuro will struggle to write any kind of formal structured notation
but if she has something in her training data that is structured, then this is easy with some basic conversion
jagged turtle
if there is a thing you can just use like she did with the images she made with ...
I'm pretty sure there is
tender river
there isn't a precedent of her having access to a bigger model
she probably has an extra model for making game decisions
tender river
she probably has an extra model for making game decisions
but i doubt it could program well
sage crag
ordr issou best search ? q = konii
olive sable
visual pawn
well, if you go with the theory that her vision is another vision-language model, then that could be a bigger model
but that's besides the point
rigid snow
she probably has an extra model for making game decisions
i meant to a bigger llm
rigid snow
why acting dumb when this is what they meant
sick owl
she probably has an extra model for making game decisions
I would have assumed that she makes the decisions and then they're parsed by a secondary model
visual pawn
I don't think there's any precedence for any game-playing model likely being an LLM
a larger LLM
jagged turtle
I would have assumed that she makes the decisions and then they're parsed by a s...
idt so, at least for the "simpler" integrations there isn't a secondary model in play
visual pawn
neuro is just sending instructions to the Vedal model
olive sable
banana 2026 campaign
sage crag
visual pawn
but going back to the original point, it seems an LLM can produce musical notation, even if it's something simple
so I don't think it's entirely impossible for Neuro to come up with something workable, even if it isn't going to be strudel
olive sable
banana campaign but im accidentally painting everything red <:AquaCry:1212306335...
fixed with bandage paper
medic 100
jagged turtle
so I don't think it's entirely impossible for Neuro to come up with something wo...
not saying it's impossible but strudel/tidalcycles seems like kinda the wrong or suboptimal way to go about it
olive sable
visual pawn
not saying it's impossible but strudel/tidalcycles seems like kinda the wrong or...
yeah, but you could use something like: https://en.wikipedia.org/wiki/LilyPond
and if it reallllly needed to be strudel, you can convert it into strudel
hollow surge
I'm now banana
sage crag
I'm now banana<:neurojuice:1155650238778978394>
ill reveal answer when
i remember who you are
and get home
rigid snow
jagged turtle
https://github.com/paulrosen/abcjs there's also this, not sure how likely it is that neuro knows what it is though
also the broken github profiles arent giving me a lot of hope
visual pawn
like I doubt Neuro will make anything complex but
the competition isn't exactly a high bar
hollow surge
masterpiece
if you compare it to some art that goes for millions, this is as complex as that
visual pawn
I think if she can throw a few measures together it'll be compelling still
jagged turtle
the competition isn't exactly a high bar
it's about the experience tbh, the only problem is how do we know which is best to pick
visual pawn
well, there's only one person who will be able to test what she can do
hollow surge
Evil
visual pawn
yeah if Neuro can't compose, Evil can
jagged turtle
I might consider making something an integration first to see what happens but it's really annoying to try and do this tbh
jagged turtle
hm interesting
hollow surge
nice
jagged turtle
you should become banana
???
tender river
yelow
jagged turtle
konii are you okay
tender river
no she's right
visual pawn
hm interesting
your favourite
jagged turtle
your favourite
it's not my favourite... i swear....
hollow surge
docker
sage crag
jagged turtle
anyways I have a full-on integrations list rn
tender river
jagged turtle
ugh, fine
sage crag
good diagram
olive sable
konii are you okay
you should become banana today
this message is brought to you by the banana campaign
this message was sponsored by samvm.dev and the triangle benchmark organization
hollow surge
banana
sage crag
jagged turtle
fine
visual pawn
hm interesting
ok it looks like it's focusing on notation, not playable music, but it still should be convertable to midi or strudel or etc.
jagged turtle
can this channel be renamed #bananagrams yet
sage crag
when banana 2026 campaign reach the ultimate conclusion
olive sable
@ shiro rename to banana general
sage crag
wrrr
hollow surge
with the sole purpose to draw bananas with different languages
sage crag
jagged turtle
ok it looks like it's focusing on notation, not playable music, but it still sho...
probably, I'm just looking for something that is somewhat known, a little bit old, and abstracts a lot of the direct music stuff away without compromising on usability
hollow surge
neuro banana emote when
jagged turtle
probably, I'm just looking for something that is somewhat known, a little bit ol...
idrc if it's usable as a website or a binary or smth
visual pawn
idrc if it's usable as a website or a binary or smth
there's MuseScore as well
jagged turtle
there's MuseScore as well
:LULE: good luck integrating neuro into that
sage crag
there's MuseScore as well
banana
olive sable
why do minecraft chickens lay eggs when they don't need them?
visual pawn
well, she just needs to be able to spit out something in MuseScore that one can convert into something else
oh wait
jagged turtle
visual pawn
yeah ok nvm lol
olive sable
neuro banana emote when
🍌
sage crag
Just saw megalag new video Bruh ain't that a security risk and breach the extens...
this is the first non-banana message
jagged turtle
queenpb is with us 
olive sable
Just saw megalag new video Bruh ain't that a security risk and breach the extens...
you should become banana today 🍌
this message is brought to you by the banana campaign
this message was sponsored by samvm.dev and the triangle benchmark organization
sage crag
jagged turtle
I just sorta assumed musescore had a text-based input as well
even if so how would I direct musescore to play, to move to input box, to pause, to delete, and all that, without modding the app?
visual pawn
even if so how would I direct musescore to play, to move to input box, to pause,...
.....pyautogui
foolproof
jagged turtle
.....pyautogui
ah got it, so reinvent #1428980640895733820
visual pawn
yes
visual pawn
py-auto-gui?
also yes
jagged turtle
py-auto-gui?
yeah it's py(thon) auto(matic) gui
visual pawn
it's just mouse and keyboard control via python
olive sable
smh not using opengl for graphics
visual pawn
ah it's not a gui framework
it's for automating gui tasks
like move mouse to here, click button, drag to here, press X, etc.
visual pawn
like autohotkey, but python
yeah
or when you want to automate something, you don't care to properly integrate or mod something, and you don't care that you won't be able to use your computer during it
or you're making a simple one-way bot in a game but who would ever do that
well, not necessarily one-way, you can grab screenshots and locate buttons and stuff using pyautogui
shut sand
holy, the piss campaign hit #programming real hard
olive sable
sage crag
holy, the piss campaign hit <#1071784467036913664> real hard
banana campaign
sage crag
sage crag
hollow surge
one day every channel will be banana town
nocturne olive
Holy #programming has become 90% more yellow
sinful cairn
sage crag
Holy <#1071784467036913664> has become 90% more yellow
banana campaign
these are both analytic continuations of the same function, except that one works in the domain and one works in both the domain and range
jagged turtle
Holy <#1071784467036913664> has become 90% more yellow
become yellow
nocturne olive
No that goes against my brand identity guidelines (tm)
mossy fossil
i got bored, so fuck it, desktop posting time
tender river
these are both analytic continuations of the same function, except that one work...
what does "works in the domain" mean
nocturne olive
i got bored, so fuck it, desktop posting time
XFCE, huh? Interesting
mossy fossil
XFCE is my comfy place.
nocturne olive
Apparently LACT uses a ton of CPU
mossy fossil
it just works, rarely gets in my way, uses almost no resources whatsoever.
nocturne olive
Apparently LACT uses a ton of CPU
That's a full thread for the daemon and GUI combined
nocturne olive
it just works, rarely gets in my way, uses almost no resources whatsoever.
Sily
I use Cinnamon myself
sage crag
what does "works in the domain" mean <:neuro3:1087521476070080592>
vibes
nocturne olive
That's a full thread for the daemon and GUI combined
Oh I see killing the GUI removes all that usage
sage crag
wrr
nocturne olive
Hem
What is this and why is it using a full CPU thread now?
nocturne olive
Hem
What is this and why is it using a full CPU thread now?
Oh ok it went down now too
sage crag
try set non integer coefficients, F stops working as expected
tender river
<:vedalShrug:1433173412217356389> vibes
nocturne olive
i got bored, so fuck it, desktop posting time
This is approximately what my system has
tender river
also what is this notation
nocturne olive
I assume a piece-wise function definition
sage crag
also what is this notation
piecewise function
desmos not support
iteration
only recursive
tender river
wait me miss rest of definition :mug:
tender river
oneliner without scrollbar
sage crag
oneliner without scrollbar <:evilDeadge:1150735483396161586>
yes this is why use desmodder
me on mobile though
sage crag
iteration
not strictly true
but, its iteration rely on ticker
is bad
me make
i didnt make well
tender river
sage crag
i didnt make well
zoom into same place even if you change centre coord
tender river
well done
sage crag
tender river
sage crag
tender river
programming baaaaaaaad
sage crag
tender river
me need break head hurt 
tender river
every animal eat leaf
sage crag
mossy fossil
covert bane
sage crag
why arent you banana
covert bane
why are we sending random gifs in #programming exactly? :neuroQuestioning:
covert bane
didnt you hear of the banana 2026 campaign
I don't think I did... could you tell me about it?
tender river
https://tenor.com/view/my-jolly-reaction-jolly-my-honest-reaction-gif-5490890317...
who are you chatter
covert bane
I think.... koni, but smarter-!
sage crag
tender river
no me not think so
sage crag
me wrrrr
tender river
differnet person
covert bane
did the IQ of #programming chatters go down- (sorry if I sound mean, this was a geniune question from me)
sage crag
sage crag
iq meaningless
sage crag
- vibecoder not talk to me about intelligence
covert bane
How much would you go about, caring about someone else's safety, situation, feelings, etc
sage crag
penguins
covert bane
are we trying to act like a different person for Neuro HQ's new years resolution? I'm definitely failing so hard atp
tender river
same person
covert bane
but you act different... hmmm okay
sage crag
me new year resolution is 720p
tender river
sage crag
tender river
current or aspirational
tender river
me not bother with resolution just have thing want to do not care if not go as planned
vm
sage crag
tomoto
wary folio
my own 3090s can reach 55C on idle... its actually quite normal when the fans are disabled... which is something the card will like to do unless you manually set the fans above 30% at all times
sage crag
my own 3090s can reach 55C on idle... its actually quite normal when the fans ar...
banana campaign
wary folio
banana campaign
shush
wary folio
is piss
tender river
banana
sage crag
cheese
tender river
tender river
it more beige
