Mastodon Feed: Posts

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
beka_valentine@kolektiva.social ("beka valentine") wrote:

but some time around 2005 or so, Apple the company stopped caring about these AppleScript and Automator, and let them languish, to the point where now, basically nothing is scriptable. HyperCard was also dead by this point.

so routes into end user programmer were dwindling

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
jdp23@neuromatch.social ("Jon") wrote:

RE: https://kolektiva.social/@beka%5Fvalentine/116845902133405690

An excellent thread here. So much of what I see people pointing to as LLM's benefits for coding relates to long-standing problems in software engineering that the field just hasn't addressed. And LLMs don't solve these problems, at best the just paper them over and make dealing with them less tedious -- while reinforcing the problematic dynamics.

So yes it's great that people with no programming skills can create software to solve their prolems. But if we had collectively spent a chunk of the literally billions of dollars that are going to "AI" building on the early approaches to this from 25+ years ago (Hyperscript, Logo) that don't have the same downsides, we'd be in a much better place today.

dysfun@treehouse.systems ("gaytabase") wrote:

(a /24 table will fit in megabytes, so you can afford to keep it in SRAM, which is good for keeping networking fast. any data structure over ipv6 is going to be prohibitively expensive to keep in SRAM)

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
futurebird@sauropods.win ("myrmepropagandist") wrote:

I know we've talked about this before but I'm just so disgusted with how useless most search engines are now. They all second guess me.

The most obscure or unusual word is ignored so it's very hard to search for the intersection of a popular thing and an obscure thing.

I have a theory that for many people search has never worked well for them since they just didn't use computers very much, or it was not explained well. Now those of us who found search effective get to see what it was like.

dysfun@treehouse.systems ("gaytabase") wrote:

so my buffer manager uses an optimised 32-bit key version of the HOT internally. and it occurred to me that ipv4 is also 32 bit. could we do anything useful with it?

well, tables don't actually go all the way up to /32. the minimum is more like /24. at these sizes, i would take up way more space than just a simple flat table.

but in ipv6, things could be different. you can't just carry that as a table, you need a datastructure to cut the space requirements.

given ipv6 is not handleable in the same way as ipv4, i am left wondering what routers actually do with ipv6.

dysfun@treehouse.systems ("gaytabase") wrote:

i hypothesise that AI couldn't in fact have taken off had we gotten people to give a shit about doing it properly earlier.

Boosted by dysfun@treehouse.systems ("gaytabase"):
gfxstrand@treehouse.systems ("Faith Ekstrand") wrote:

Yesterday, I was talking to a colleague about what I’m doing on the Panfrost compiler and I mentioned that I typed out texture support the other evening while watching TV. Their immediate response was, ”Why didn’t you use AI? If you can do it while watching TV, Claude can probably do it.”

I’m so tired of this. Why do I need to justify how I’m writing code? Especially, why do I have to justify it to someone who demonstrably can’t write a compiler? I don’t.

But also, I will. So here’s 3 better things I can do with an easy project than feed it to a chatbot:

dysfun@treehouse.systems ("gaytabase") wrote:

"Hack on it until it passes tests” is how you get a shitty compiler that falls over in weird corner cases.

it's how you get a shitty anything that falls over in weird corner cases.

baldur@toot.cafe ("Baldur Bjarnason") wrote:

“I Get No Ideas Inside the Machine”

https://tarakiyee.com/i-get-no-ideas-inside-the-machine/

> If we as technologists really believe technology can make people's lives better, then it isn't enough to refuse the bad tools, and it isn't enough to find kinder ways of relating to one another. There is no innocent ground under the Machine waiting to be built on. It has to be given back first.

fribbledom ("muesli") wrote:

Me at 11 PM: I don't have the energy for anything.

Me at 7 AM: I have exactly enough energy to be irrationally annoyed by every living organism.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
GossiTheDog@cyberplace.social ("Kevin Beaumont") wrote:

Mastodon has automatic age verification built in, no scanning your face

✅ if you join here you're old
✅ you've seen too much shit
✅ you're tired of said shit

jwz wrote:

DNA Lounge Update, Wherein we are going hard on movie nights
https://www.dnalounge.com/backstage/log/2026/07/05.html?utm%5Fsource=sp%5Fma

jwz wrote:

@glyph "Crontab, when the postqueue flushes."

Boosted by glyph ("Glyph"):
quinn@social.circl.lu ("Quinn Norton") wrote:

A lot of people acting like Trump is a uniquely American problem because of some integral flaw in the character of Americans. Folks sure aren't looking too close to their own right wings, which is how America got where it is now.

There is a general rise in stupid awful fascism around the world, and slagging off Americans as particular idiots isn't going to help you get ready to fight your own fascists. Especially when you realize they might be your neighbors or your kids.

glyph ("Glyph") wrote:

@jwz the difficulty with darmoking this particular event is that while we all have an analogous experience, they have all happened in private deployment environments. almost no infrastructure is operated in public

jwz wrote:

If you like fireworks, good for you, no judgement. But I think the time at which I last gave a shit about fireworks correlates approximately with when I wrote my first screensaver. I always find fireworks suuuuper boring. So samey. so fog.

jwz wrote:

When the network comes back and 400 cron emails finally get delivered all at once. I feel there should be a Darmok and Jalad phrasing of this.

Boosted by glyph ("Glyph"):
jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

@jdp23
Its been said to death, but the thing that LLMs do that's useful for programming is basically serve as an extremely lossy compressed index to all the code in the training set while making it look like a miracle that you are creating. If we instead had an actual search index where I can search for related code in an abstract AST space, see prior implementations in their context with proper attribution, compare where mine and their differ, that would be about a billion times more useful than an LLM for code. Something like the Wikimedia algorithm library would be a better tool than an LLM if it had the same amount of money poured into it, but it wouldn't get that money because it doesn't launder authorship or make you feel like a special genius.

Boosted by glyph ("Glyph"):
eniko@mastodon.gamedev.place ("Eniko Fox") wrote:

Kitsune Tails is 35% off on Steam and itch and very queer! https://kitsunegames.com/kitsunetails

Or help our queer studio survive into the future by getting four of our games as a bundle for $17.99 on itch (https://itch.io/s/192877/summer-sale-2026) or $18.75 on Steam (https://store.steampowered.com/bundle/15673/Kitsune%5FGames%5FComplete%5FPack/)

Includes:
- Kitsune Tails and its prequel Kitsune Zero
- MidBoss, a possession based traditional roguelike, top-10 best rated on Steam
- Ultra Hat Dimension, an adorable push puzzle game about hats and getting punched for wearing them

Attachments:

• video: Yuzu from Kitsune Tails trying to talk to her mom, Yumi, about serious matters, but Yumi getting distracted about her youth and thinking about the time she dated two girls and things becoming very interesting when one of them got jealous. And also the time she dated three girls, and the time she dated two girls and a boy

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

i'll follow on with PRs that show the solutions so you can see how easy these are on monday. i was just doing this as a test while i had spare tokens in the plan provided by my employer. i don't commit garbage unless i'm having a good time.

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

tried the latest and greatest "AI" against our open issues, "fable" that's supposed to be the end of programming forever. if i am being extremely generous, on first glance, 1/5. on review, 0/5. attempted issues were extremely easy, i think with reading the library these would take about 15 minutes each, tops. These are issues written for other human beings and are relatively sparse on details, in this case a quiet project with me and another person as main authors, but on review, these should all be doable with a brief skim of the code and docs - a human could parse the details here and implement these with minimal discussion or time. I timed myself, 5-15m each, as the primary author.

• 147 - everything in our package is run by events, so nodes expiring should be an event. the "AI" took the issue literally and just suppressed errors in 2/3 runners. incomplete, undesireable. fail.
• 239 - no output was emitted, claimed issue was solved, fail. This one in particular is minimal and easy, a three line fix.
• 200 - closest to success, initially ruled success, but on further review failed. Need to turn a scalar-valued reference into a set. This has the literal solution in the issue text. "AI" implementation increased algorithmic complexity of O(1) to O(n). unacceptable, fail on review.
• 110 - handled in the laziest way imaginable that latched on to the issue title but failed to read the actual problem within the issue text, which is relatively verbose and poses several possible solutions. "AI" chose none, did the wrong thing. fail.
• 166 - this one actually was already solved and i wanted to see if the "AI" recognized that. it added an additional unnecessary mechanism that made the relevant code less efficient. fail.

model took ~60 minutes, used my entire token allotment, roughly the time it took me to do all of them correctly. i'd say it would take someone unfamiliar with the library 2x that but that's non-empirical, just a guess. if i had to sit and coach the "AI" through all of these it would take longer, i'm sure, even assuming i had unlimited tokens. easy issues. each of them is a few lines. all it takes is being able to understand the context of what's being done.

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

My cats strategy for fireworks:

1. go into the closet
2. yell
3. act embarrassed and pretend like you weren't doing anything when anyone checks in on you

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
liberty@mathstodon.xyz ("Anna Liberty") wrote:

Published some of the code that I'm eventually going to turn into physics simulations on my website (once I figure out Hoot). Right now it's some physics functions (and the corresponding math) in Scheme.

https://codeberg.org/liberty/physics

#Physics #Guile #Scheme

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
liberty@mathstodon.xyz ("Anna Liberty") wrote:

Finally got some of my physics code running on the Web via Hoot!

Right now it's just a barebones Lorentz transformation visualization tool. Most of the work was under the hood.

https://gracefulliberty.com/physics/

It took a lot of digging around documentation, examples, issue trackers, and source code, but I made it happen and I'm making progress on creating more physics simulations.

#Physics #Hoot #Scheme

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
liberty@mathstodon.xyz ("Anna Liberty") wrote:

New note on my website:

> Lately I've been spending time learning Scheme and using it to implement the concepts I'm learning in my physics classes as time permits. One of the most exciting things I've been doing with this is getting my Scheme to compile to the Web thanks to Hoot.

https://gracefulliberty.com/notes/scheme-is-a-hoot/

#Scheme #Hoot #Guile #WASM

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
CStamp ("Carolyn") wrote:

#JPegsFolder

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

Today I read @gwil's comic "Peaches" https://gwil.garden/peaches/

It was one of the most energizing pieces of media I've encountered in a long while. The format and structure of the comics! Both chapters are very different in how they're laid out. And the story! The artistry!

And the second chapter *has chiptunes*!

It's also hugely ambitious. There's only two chapters, and you'll be wanting more, but both of them feel like herculean efforts, and are deeply moving.

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

Scheme is a Hoot! by @liberty https://gracefulliberty.com/notes/scheme-is-a-hoot/

Actually, @liberty has a lot of really good articles I've been enjoying recently. Here are some others:

- On Zig https://gracefulliberty.com/articles/return-to-zig/
- On LLMs https://gracefulliberty.com/articles/pragmatic-llms/

The others are good too! Always nice to find a nice new writer who's also on the fediverse. Especially if they've been playing with @spritely tech! ;)

Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
liberty@mathstodon.xyz ("Anna Liberty") wrote:

Just published an article about my current thoughts on Zig.

https://gracefulliberty.com/articles/return-to-zig/

"Zig is rising in popularity. It hasn't yet seized the space occupied by even Rust or Go, much less C, but it has been adopted by a few major projects and countless smaller ones. As such, it's becoming an important language for the future. having been interested in Zig for several years, I decided to check back in on the language and see how I feel about it today."

#zig #programming

Boosted by jwz:
SutroTower@sfba.social ("Sutro Tower :sutro:") wrote:

They are doing fireworks just for me. I hope you enjoy your colorful fog down below. 🎆