What is Hackers' Pub?

Hackers' Pub is a place for software engineers to share their knowledge and experience with each other. It's also an ActivityPub-enabled social network, so you can follow your favorite hackers in the fediverse and get their latest posts in your feed.

0
0
0
0
0
0
1
0
0
Bram Meehan @BramMeehan@ohai.social
Public
Shared by Rafaelito 🦤

Who knows if this’ll make it to anyone new to Fedi, but…

My advice: upload a photo and fill out that profile, just to show there's a human there. Follow anyone and anything that seems moderately interesting (and prune as you go).

Then just read and absorb it all for a couple weeks. Get a sense of what it's all about, how the culture and no algorithm work. This has some trappings of corporate social media sites, but it's very not — figuring out exactly what that means was key to my experience.

0
0
0
0
0
0
2
0
1
2
0
0
0
0
0
1
0
0
0
0
0
1
0
0
Simon Ser @emersion@hachyderm.io
Public

Today I've investigated migrating drmdb.emersion.fr to SQLite. It wasn't as simple as I hoped.

drmdb collects JSON blobs representing supported features for GPU display engines. It has ~1k blobs, each between 10 and 200KiB. Operations needed to render a page aren't easily described in something other than application code: specialized comparison functions, deeply nested aggregates, and so on.

Originally drmdb used a very dumb scheme: it stored all of the blobs in a directory. Each time it needed to render a page, it loaded all of the blobs from disk. It worked pretty well until we hit a few hundred blobs. Then it took multiple seconds to load each page.

I put varnish in front of drmdb as a stop-gap, but cache invalidation was annoying to deal with. Users wouldn't see their device in the lists after uploading their data. Pages would still take multiple seconds to generate after being invalidated.

Then I switched to another dumb approach: on startup, read everything from disk into memory, up to a limit of 1k. Startup would take a dozen seconds, but then everything would render instantly.

Then the dataset grew larger. I bumped the 1k memory limit to 4k. It felt wrong, but worked okay. The heaviest page loads in 3s now, second heaviest page loads in 400ms.

I wondered how SQLite would fare for this kind of unusual use-case. Just moving the dataset from the filesystem to SQLite made everything very slow again: well, everything is pulled from disk so that's to be expected.

(Continued below)

DRM database

drmdb.emersion.fr

0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
✧✦Catherine✦✧ @whitequark@treehouse.systems
Public
Shared by 洪 民憙 (Hong Minhee)

hi! i'm cat. i work on open source software and hardware, like:

and a lot more.

i joined mastodon.social back when it was cool, about a decade ago. it stopped being cool a while ago so i figured i'd move over to treehouse.social. here's to a decade more here!

Introduction — Amaranth language & toolchain 0.5.8 documentation

amaranth-lang.org

3
0
0
1
0
0
0
0
0
0
0
0
Bram Meehan @BramMeehan@ohai.social
Public

Who knows if this’ll make it to anyone new to Fedi, but…

My advice: upload a photo and fill out that profile, just to show there's a human there. Follow anyone and anything that seems moderately interesting (and prune as you go).

Then just read and absorb it all for a couple weeks. Get a sense of what it's all about, how the culture and no algorithm work. This has some trappings of corporate social media sites, but it's very not — figuring out exactly what that means was key to my experience.

0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
Chao-c' @xChaos@f.cz
Public

RE: graphics.social/@metin/1158723

"Pirating" AI prompts sounds funny. They should think about themself rather as photographers making photographs of public spaces - they can't really expect to copyright one exact shot from certain angle.

This is more complex problem in the FOSS ecosystem already. Lot of companies use free software and their admins get paid well, sometimes, but the programmer's work is somehow expected to be free or paid by hourly fee, because the profit of the companies using free software is not intended to be shared. Somehow same situations was there with DJs, being famous for mixing nameless tracks.

The important question, if "sloperators" expect themself to be well paid professionals or even famous artists in the future. I think this future will never be. There are no "famous cell phone users". The AI model companies will have famous brands, some developers will get famous and well paid. But I think that average sloperator can expect probably as much attention, as eg. dish washer or coffee machine operator. Anybody else will be easily able to achieve the same result, and actually, I expect "open prompt" culture, where prompts generating certain result will be shared together with the result, much like open source is available for binaries.

Very few people will be able to created prompts "better then someone else", so much of the current effort of many people will soon look the same like putting photos of your fast food on Instagram. There will be nothing unique and competitive about the output.

0
0
0
1
0
0
いかる:role_illustrator::skeb::irai_bosyuu_tyuu: @ikaru@misskey.io
Public

ただ、１度くらい完コピ人間を目指してもいいのかなとは。技術的に無理だってのがわしの出した結論だけども。

（言い方悪いけど）誤認させるほどまで真似れたらクオリティ爆上がりは間違いないのよね。そこ目指してもいいのかな…とかとか考えてるとこですな​:thinknyan:​

1
0
0
1
0
0

Cinema on the Streets: 17 Murals That Look Like Movie Stills

STREET ART UTOPIA @streetartutopia@streetartutopia.com

Ever walked down a street and felt like you were suddenly inside a Hollywood blockbuster? These 17 masterpieces turn ordinary walls into extraordinary cinematic scenes. From towering monsters peeking around corners to intimate, emotional portraits that feel like frozen film frames, street art is evolving into a form of public storytelling. We’ve collected 17 incredible artworks that blur the line between reality and the silver screen, proving that the world truly is a stage. More: 14 […]

Read more →
0
0
0
2
0
0
0
0
0
Joshua Barretto @jsbarretto@social.coop
Public

What would a language model trained on all of the literature present in 1500 look like? No doubt it would spout a lot about God and the geocentric model.

Would we expect it to ever develop the heliocentric model, Newtonian mechanics, or Ricardian economics?

Exponential growth is self-similar at all scales, so suggesting that LLMs can push the envelope of 21st century knowledge and beyond is akin to claiming that our hypothesised 1500s GPT could do the same in its own time.

0
0
0
いかる:role_illustrator::skeb::irai_bosyuu_tyuu: @ikaru@misskey.io
Public

でも夢見がちょっとアレで。前にも悩んでた「おてほんのコピー人間になる」か「それが嫌なので大きく参考にし続ける」の２択を夢の中で迷ってたんですね。

コピー人間になればクオリティは爆上がりするのでこれ１択に思えるけど、わしがなりたいのは完コピ人間ではないので。でも結局答えが出ないという。うーむ…​:meow_umm:​

1
0
0
1
0
0
1
0
0
Rory Mir @falsemirror@masto.nyc
Public

Let's mandate they store even *more* sensitive info. About children maybe?
cybersecuritynews.com/instagra

Instagram Data Leak Exposes Sensitive Info of 17.5M Accounts [Updated]

A significant security breach has compromised approximately 17.5 million Instagram user accounts, exposing sensitive personal information.

cybersecuritynews.com · Cyber Security News

0
0
0
1
0
0
Cheeseness @Cheeseness@mastodon.social
Public

@mcc This is a feeling I've been having for a few years, and I struggle to see how the normalisation of unaccountable LLM platforms and users won't spell the end of the Free Software movement/type of F/OSS culture I care about, and that's heartbreaking on so many levels.

Cheeseness @Cheeseness@mastodon.social
Public
Shared by mcc

@mcc It makes me reluctant when I think about sharing my work. It makes me reluctant when I think about accepting contributions. Even for closed source projects, it makes me wary of collaboration because I feel like the level of micromanagement I'd need to have confidence that my work wasn't being tainted just isn't the sort of working relationship I'd like to have with others.

I don't eat fish because I'm not interested in carefully chewing for bones all the time. This feels like that.

0
0
0
mcc @mcc@mastodon.social
Public

LLM advocates and their enablers will hide the fact that they are posting stolen code— they brag about this, their most widely-effective argument for LLM code being allowed in OSS is "if you tell us we can't, we'll do it and not tell you"— and attempts to track which open source projects use stolen code will be harassed off the Internet. I do not want to be part of this open source community, so the only options I seem to have are to wall myself off from it, or stop doing open source altogether.

Cheeseness @Cheeseness@mastodon.social
Public
Shared by mcc

@mcc This is a feeling I've been having for a few years, and I struggle to see how the normalisation of unaccountable LLM platforms and users won't spell the end of the Free Software movement/type of F/OSS culture I care about, and that's heartbreaking on so many levels.

0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
2
0
0
0
0
0