<unnamed>

Public Key

npub1asqy0409lrzczthp56pw5l46cspy6ws74y3mj2alpmd8gnml088qxwzd6m

Profile Code

nprofile1qqswcqz86hjl33vp9ms6dqh206avgqjd8g02jgae9wlsakn5falhnnspz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsxznfeq

Publishing to

relay.primal.net

relay.damus.io

nos.lol

Last Notes

2026-05-15 16:31:15 GMT

🟧 anyway I'm fucking exhausted but to re-summarise for people who might want to share again:
https://github.com/rust-lang/rfcs/pull/3959
our policy now fully bans LLMs from having any "creative control" over the project. it carves out trivial usage for things like accessibility and to not litigate shit like copilot autocomplete, even though we personally think it's a shit technology that should not be used
disclosure now exists as a means to openly invite LLM users to tell on themselves, effectively. more diplomatically, it exists to justify and hold accountable people who might have used an LLM to collect data or some shit to make sure they only did it for that shit and didn't let it write any code or prose. if people don't disclose anything and an LLM was obviously involved, the result is to assume that it was involved too much and close the PR.
there's plenty to be criticised about this praxis, but it feels like the best way to ensure that people don't get exhausted trying to litigate whether LLMs were used and instead just ask people to be honest. hopefully absolutely canning most of the usage of LLMs will mean people just stop using them except for real accessibility concerns when there's no alternative.
I am extremely frustrated with jyn, the only other serious policy proposer, for effectively throwing us under the bus and citing a fucked-up fashy quote we vented about in an entire blog post: https://txt.ltdk.xyz/testing-the-limits-of-kindness/
(to be clear, jyn has since apologised and was just tired and trying to deal with the situation. everyone involved has just been exhausted trying to do stuff and it means everyone's at each other's throats.)
TC, the other technical policy proposer, should not be taken seriously. he knows what he's doing and I hate him for it, and his policy is not a serious one. it's why I'm excluding him from the discussion here.
I just want people to understand just how absolutely one-sided this argument is despite it draining all of our energy. basically the entire project minus a handful of people agrees that LLMs are trash and should be banned entirely, but that handful of people are some of the ones with the most influence and so things are substantially more difficult. it means that despite all the private remarks that thank us for our policy work most of the private discussions feel like incredibly one-sided dogpiles on us being rude for even daring to call one guy's opinion slightly fascist by accident.
if you wanna help with the discussion, I encourage you to comment on our RFC or show support directly. just don't just show up and make a mess, please, because all you're doing is playing into the handful of hands that are making work hard for the rest of us.

2026-04-01 21:31:02 GMT

- reply

@nprofile…v4j3 only 4?

2026-03-31 18:31:50 GMT

- reply

@nprofile…qj99 this is honestly one point of the LLM grift people don't really focus on
like, right now, people think that these LLMs are good because they respond within reasonable timescales. "Claude might spend ten hours on a problem, which is around how much time I might have spent on it!"
except Claude didn't take ten hours. Claude is multiple concurrent processors running at once. you have to multiply the number of hours by the number of running processors to be reasonable.
even if we pretend Claude isn't fanning out to multiple servers, even like, a single GPU is hundreds of processing units. so now your ten hours is thousands of hours of time spent on the problem
and like, brute-forcing shit takes time. that's why it's like this, and they just hide it behind multiple computers and pretend that it's not brute-forcing and just thinking

2026-02-06 15:47:00 GMT

- reply

@nprofile…qjyt @nprofile…pfuj funnily enough, Rust is the language that would have your back in that case: they offer the most variants of division via built-in methods (even though the default is truncating), but in particular they offer euclidean division, which is specifically "round whichever way makes the modulus positive" which would be what you want here for array indices

2026-02-06 15:18:54 GMT

- reply

@nprofile…pfuj fwiw Python is particularly unique in that // is actually flooring division, e.g. -1 // 2 is -1, not 0
in most languages today, integer division is truncating, not flooring

2026-02-06 03:39:52 GMT

- reply

@nprofile…uptt main issue with the simple layout imho is that it doesn't allow viewing notifications and feed at the same time. would use it instead of the advanced view if not for that; we don't need fifty billion feed columns, just the ability to read notifications without losing spot on regular feed

2025-10-20 16:43:55 GMT

- reply

@nprofile…x33s so they actually explicitly mentioned that it's not even a superellipse, which is quite interesting

2025-10-20 08:28:05 GMT

I do wonder if the Iwerlipse is actually a new shape… 🤔
https://youtube.com/watch?v=_yx0eutBwII

2025-08-11 23:55:24 GMT

- reply

@nprofile…6h3y good
we need fewer shithead hunters like him

2025-07-25 08:01:40 GMT

- reply

@nprofile…hr6z doesn't python just have an out of memory error when this happens or does it actually crash
it's still upsetting to me as a rust dev that python just treats _ as a regular variable name instead of meaning, discard this

2025-07-24 17:30:59 GMT

I'm a bit privy to more of an inside view into payment processing than most people, even though I have barely any, because I have programmed credit card systems before. and while you should not ignorantly trust all information on the matter, hopefully you at least glean a bit of insight on this, and helps you understand reading other resources on the subject like the Wikipedia page on credit card processing.
there are a lot of issues with the credit card system, and the first big one is that it has way more intermediate parties trying to scoop money out of the process than you realise. in addition to the payer, the payment system (visa/mastercard), and the payee, there's also acquiring banks as well.
essentially, an acquiring bank is responsible for holding the money directly after a transaction takes place, and they pay out to the actual company who received the transaction less frequently. effectively:
the issuing bank gives a person a card, and holds the money from people paying for things
the acquiring bank takes the money paid via card
visa/mastercard just facilitate this process (and, of course, take their cut)
acquiring banks will pay back to regular issuing banks for companies at a much slower cadence
these acquiring banks operate like regular banks, not taking any money but instead relying on interest from held money. one of the biggest things they facilitate is also chargebacks, a loadbearing aspect of the credit card system.
basically, when you call your card company saying your card was stolen, and that someone made purchases on your behalf? or that a company charged you when they shouldn't have? that's a chargeback, and they are very bad for acquiring banks.
chargebacks sometimes cost up to 40$ for the company who offered the sale, per transaction, meaning they can be quite expensive. even if the company ultimately proves that the purchase was genuine and gets to un-refund the money, they don't get that fee back. acquiring banks will only do business with companies that have very low chargeback rates, so, it's in a company's best interest to avoid anything that might cause this even if customers are lying, saying that purchases were fraudulent even when they weren't
which leads into why payment processors are so aggressive on lewd content: it is the most likely to be charged back, because of an existing puritanical environment. if your wealthy, conservative account holder calls you and said someone bought a bunch of porn with their card, they won't take it well if you accuse them of buying all that porn themself, even if they likely did. so, banks generally agree to charge back, which ultimately hurts the person selling it: if they get too many, the literal bank that receives the money for them might decide to close the account
so, any business that actually sells lewd content effectively has to game the system, doing what amounts to effectively money laundering: constant shuffling transactions between a large number of banks to avoid any one bank receiving too many chargebacks. they also have to work with rather sketchy companies to do this, since all the "respectable" ones like Stripe will refuse to work with you. this is effectively just seen as the price of doing business for large companies selling lewd content, and usually, the payment processors just look the other way, but for individual creators, this is an impossible barrier to entry
a big part of why we have this absolute mess of payment processing is because the system is optimised for wealthy individuals who flaunt cash wildly but hate to lose it. despite the security theatre of chip readers and various other things credit cards offer, ultimately, typing in a single number to pay for things is an inherently insecure system, and the system is designed with chargebacks in mind to deal with that, instead of being more secure.
for example, did you know that address verification on credit cards is entirely optional? you literally have the option to choose whether you refuse transactions that type the address wrong, or just let them go through anyway. the stated purpose of this is to let stuff like subscriptions work even if your address changes, but you can also consider it as people openly admitting that they are okay taking the money if your card is stolen, and will only get upset if you actually try to work and get it back. the "security" code on the back of the card also falls under one of these optional checks, but I think we can all understand that a three-digit PIN is just security theatre.
I'm not saying that chargebacks are inherently bad; the idea of everyday people being able to get their money back in the event of fraud is a good thing, given our broken payment system. but maybe the system is completely broken and should be replaced.
I don't think that "all-in-one" processors like PayPal are good, where everything takes place in temporary "bank accounts," but at least it has the right idea. people should be able to split their money into a much larger number of pools so that if any one pool is compromised, that doesn't mean they go broke. I just don't want more people to hold onto money, and I don't want individuals to have to micromanage this when they barely have any money to begin with
I dunno what the solution should be. but hopefully, understanding the problem helps a little bit

2025-07-15 06:32:39 GMT

- reply

@nprofile…hr6z if you think about it, the universe is just one big tunnel that connects us all

2025-06-25 19:37:08 GMT

- reply

@nprofile…cgs9 @nprofile…2kg7 not any more!

2025-05-25 17:51:03 GMT

- reply

@nprofile…m8mr you're really just digging your feet in for no reason
the issue isn't the technical limitations being unclear; everyone has understood that aspect. the issue is that there's an unwillingness to work on some compromise instead of just performing this basic analysis and then rejecting the idea wholesale without thinking of any workarounds
for example, Google Play Services already requires multiple OS packages to operate. you install the base OS (GPLv2) and then install the proprietary Google Apps on top before the initial boot. it doesn't seem unreasonable to have a third GPLv3 package containing TTS or other useful software that can't be included in the base image for licencing or security reasons
like, this sort of stubbornness is endemic to FOSS, and while I understand that work doesn't magically get done without actual time and resources being put in, you shouldn't be completely rejecting important work just because of a few key issues that you refuse to find compromises for. you dug yourself into the GPLv2 hole but you don't need to make it your grave

2025-04-23 23:54:15 GMT

- reply

@nprofile…ucsf I thought the smoke was supposed to be blue or pink
how will we tell the gender of our new pope now

2025-04-01 17:40:19 GMT

- reply

@nprofile…6h3y I explicitly tagged this post as "wrong" to clarify that it's a joke

2025-04-01 17:33:27 GMT

- reply

@nprofile…6h3y it changes the language the link is announced in
so, for example, hreflang=fr makes it say "linque" instead of "link"

2025-03-29 15:35:15 GMT

fellow white people,
if you find yourself saying "I don't know what woke meant before" or "I'm referring to the modern definition of woke"
how about you: don't use that word, then

2024-12-30 22:51:01 GMT

- reply

@nprofile…ucsf I mean, it's important to remember that the bots are intentionally making bad moves since, if they chose the best move according to stockfish, they'd just beat human players every time
it might be that they're choosing, say, the Nth worst move according to stockfish, and when you get past a certain point in the endgame there are less than N decent moves, and the moves chosen by the bots thus always end up as terrible self-sacrificing moves
of course, chess.com's code is all closed, so, I can't really know what they're doing here, but maybe if the strategy is actually this simple you could test it, since stockfish is open source

2024-12-23 23:07:45 GMT

NileRed the paper and actually did the work of figuring out how to do the thing after lots of research and preparation
NileBlue right past all those steps and probably failed, but also he has no right to succeed under those conditions anyway

2024-12-20 22:14:36 GMT

- reply

@nprofile…6h3y I'm curious about why the interactive caveat is here since I am imagining some really cursed CSS that tries to restyle buttons by replacing them with ::after contents or something
it seems wild that people still don't want to just write semantic HTML and tweak the appearance after

2024-11-01 00:59:14 GMT

obsidian is pissing me off enough I want to switch to something else but there really aren't many options
I guess I'll keep it, but be upset about it

2024-09-29 11:44:19 GMT

so, fun fact, the full name of MRI is NMRI, nuclear magnetic resonance imaging, and it's the same NMR that chemists use to get a better idea of the molecular structure of things
but the N got cut off from the name because people were terrified of nuclear technology at the time even though it's completely unrelated to nuclear fission
also kinda explains a bit more why human MRI scans require so much extra software to get a good reading, since you're effectively just measuring the molecular structure of everything in an area of the body at once and it's really hard to sift through the noise in that to figure out what you actually care about
it's nuclear magnetic resonance because you're literally measuring how much the nuclei of various atoms react to a fucking massive magnetic field, and it can tell you stuff about even the structure of relatively complex molecules because, as you'd expect, the position of atoms relative to each other can dramatically change the results

2024-09-20 16:50:18 GMT

- reply

@npub17nm…dm57 I just keep all my TOTP codes in bitwarden tbqh

2024-08-11 14:41:11 GMT

- reply

@npub1ek8…t2en if the alt text is forced to have a limit it should be at a bare minimum the amount of legible text the image can contain and it's not even close to that
it's bad

2024-07-10 20:48:04 GMT

btw instead of "algorithm" when referring to the machine-learning shit that websites use to optimise ad revenue, you can instead call them "the business requirements of some chud" and it is clearer and more accurate
there is no algorithm. only the business requirements of some chud

2024-06-28 17:39:57 GMT

the misogyny that thinks that cis women are icky for menstruating is the same misogyny that thinks that trans women are icky for "not having women's bodies" by any possible metric
it's all just people being the epitome of childish for thinking that common body types are icky

2024-06-05 12:12:24 GMT

- reply

@npub1zyl…t54n ah yes, nav role=navigation
surely that makes it twice as navigable

2024-03-22 15:58:56 GMT

- reply

@npub1wu5…qx7r just to summarise this in the likely event the video goes down:
each Switch cartridge contains a unique certificate to verify its uniqueness, and this is used during online play and for other situations like redeeming points for the eshop
people have been able to copy these certificates from cartridges for a while now, but because no one had the ability to actually use them, there hasn't been an incentive to do so
now, there is hardware that is cheaper than the used price of most games to emulate an exact copy of a cartridge, meaning that scalpers are highly incentivised to either dump a game and return it, or sell dumped copies of a game secondhand
in both cases, Nintendo can and will notice that multiple copies of a cartridge are being used online, which could be used to either ban specifically that cartridge or any system using it that isn't permanently offline
returns for opened games have already not really been allowed due to the eshop points thing, but now more than ever, this is going to make selling and buying used games less viable and potentially risky

2024-03-12 21:08:40 GMT

- reply

@npub1zyl…t54n honestly I've found the most important thing both for tea and coffee is having a properly sealed container to keep them fresh
I have glass jars with silicone seals and my earl grey still smells as pungent as when I bought it, compared to when it's just left in the opened tin and loses its smell almost within a few days

2024-03-12 20:52:59 GMT

- reply

@npub1zyl…t54n presumably you would expect the shots to be much heavier than the beans because of the water? unless I'm reading this wrong

2023-07-15 01:31:49 GMT

- reply

@npub1znz…8l64 honestly one thing I keep telling myself (and failing to do enough) is that I need to be okay with writing shit code because shit code is better than no code
most of the time I just think "I could do better" and don't do anything when that's just a bad mentality
so, imho, you're doing well because you're actually making something, and the finer details don't matter nearly as much

2023-06-09 00:38:18 GMT

- reply

@npub1psy…pasy the biggest problem with this is that launching a rocket to mine space uses way more fossil fuels than it would take to otherwise mine those on earth
the biggest concern with mining on earth is that people are exploiting the people doing the mining and we're using them at an unreasonable rate, not nearly as much that we're super low on said resources and have to look elsewhere