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.

Initiative Hamburg Werbefrei @hhwerbefrei@bewegung.social
Public

Schon eine voll?

Glückwunsch! Bringe sie gern schon vorbei. Bei uns in der (neben dem ) gibt es dafür:

- Anerkennung
- Andenken: hamburg-werbefrei.de/wp/post/6
- den neuesten Tratsch von der ;-)

Hamburg Werbefrei – Volksentscheid Hamburg Werbefrei

Hamburg Werbefrei – Volksentscheid Hamburg Werbefrei

Damit es zur Volksabstimmung kommt, brauchen wir bis zum 13. Mai 2025 die Unterstützung von 80.000 Menschen.

www.hamburg-werbefrei.de

0
0
0
Initiative Hamburg Werbefrei @hhwerbefrei@bewegung.social
Public

Wenn STROER versucht, die Stadt damit zu erpressen - dann sollte die Stadt die Sache eben wieder selbst in die Hand nehmen.

Anstatt sich von den Werbekonzernen noch den Wortlaut von Verordnungen diktieren zu lassen.

Früher ging es doch auch ... die Stadt hatte eigene Werbeflächen von der Art, die wir beibehalten wollen (Plakatwände und Litfasssäulen) und auch Fahrgastunterstände.

Initiative Hamburg Werbefrei @hhwerbefrei@bewegung.social
Public
Shared by Quincy ⁂

Die Koppelgeschäfte mit "Stadtmöblierern" sind ein Irrweg, denn sie machen die weniger demokratisch:

1) die Verträge gehen über 15 Jahre, also drei

2) manche Details gelten als , obwohl sie unmittelbar die Allgemeinheit betreffen

3) der Einfluss auf die geht sehr weit, z.B. darf politische (!) vor Wahlen und Volksbegehren die "Rechte" von STROER und JCDecaux nicht einschränken - und die Sichtbarkeit der Werbeanlagen nicht behindern (!)

4) die Stadt wird immer weiter über den Tisch gezogen, da man sich mit den privaten Partnern gutstellen will. So haben die Konzerne während der Corona-Krise großzügige Zusatzkonditionen für sich herausverhandelt.

0
0
0
Elena Rossini on GoToSocial ⁂ @elena@aseachange.com
Public

🏕️ my adventures in #selfhosting - day 139 (filmmaker edition) 🎬

Good morning Fedi friends!

I've been quiet over here for a good reason: I'm currently laser-focused on producing a first promo video about the Fediverse. I'm working with two talented interns and today we start filming! More on this here: https://mastodon.social/@_elena/114454286121971517

Any free time I have I spend working on a 10 minute presentation about my self-hosting journey, that I will deliver later this month at a FOSS conference. I was invited by the amazing #YunoHost team to speak along with other inspiring folks about using YunoHost to self-host.

All this to say: it's a very very busy month and I may not have much to say about self-hosting per se. But! I just did my Monday morning ritual AKA backing up my YunoHost setup.

Wishing everyone a great week!

#MySoCalledSudoLife

Three people sit on a couch. On the left: a young man with black hair and a beard wearing a black blazer and white pants; in the middle: a woman with long brown hair (me) wearing a blue and white striped t-shirt and blue jeans, holding a clapper with the text: "Fediverse Promo no.1 by Elena Riyen & Sam - May 2nd 2025"; on the right there is a young man with glasses and light brown hair wearing a red hooded sweatshirt. A bookshelf and various items are visible in the background.

Elena Rossini ⁂ (@_elena@mastodon.social)

Attached: 1 image Good morning Fedi friends! Please join me in giving a warm welcome to my interns Riyen (@patel.riyen@pixelfed.social) & Sam (@samaaberg). They will be working with me through the end of June on videos promoting the Fediverse. This is their #introduction. Sam & Riyen are two talented film students who are #NewHere: a mere 7 days ago they had no idea what the Fediverse was. I have the immense honor of introducing them to it. We are filming this week! Send us good vibes 🎥✨ #EleFediVideos

mastodon.social · Mastodon

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
한겨레 @hanibsky.bsky.social@bsky.brid.gy
Public

또다시 ‘전봉준 트랙터 투쟁단’이 서울을 향합니다. 지난해 12·3 계엄 내란사태 뒤, 지난해 12월 시민과 함께한 ‘남태령 대첩’과 지난 3월 서울 경복궁 트랙터 입성 이후 세번째입니다.

또다시 서울로 가는 ‘전봉준 트랙터’…여성 국회의원·도...

다시 서울 향하는 ‘전봉준 트랙터’…운전대는 여성 도의원들이 잡는다

또다시 ‘전봉준 트랙터 투쟁단’이 서울을 향한다. 전국농민회총연맹(전농) 광주전남연맹은 6일 전남 무안 전남도청 앞길에서 트랙터 투쟁에 나선다. 지난해 12·3 계엄 내란사태 뒤, 지난해 12월 시민과 함께한 ‘남태령 대첩’과 지난 3월 서울 경복궁 트랙터 입성 이후

www.hani.co.kr · 한겨레

0
0
0
0
0
0
1
0
0
Ketan Joshi @ketan@climatejustice.social
Public

1 week ago, Spain was plunged into darkness, and in the days that followed a traditional / social media firestorm cynically exploited the catastrophe: blaming it on Spain's solar power and calling for more fossil fuels and nuclear to 'save' the grid.

So: I decided to look at the longer-term data, and I found that nuclear power in Spain was operating at its lowest levels of generation in at least ten years (if not longer). We have almost no information on why, but one operator blamed power prices being too cheap to run nuclear.

I have been pretty bloody stunned at the total absence of media questions around Spain's nuclear fleet, why so much capcacity was being intentionally withheld at the time of the blackout, and what their stated rationalisations mean for the future of grid stability, power prices and variable renewable energy.

So - here's a hefty, data-driven write-up for you. Enjoy! :)
ketanjoshi.co/2025/05/05/why-n

Why no one is asking questions about Spain’s mysterious missing nukes

Spain’s blackout is being exploited by advocates of fossil fuels and nuclear. But no one is asking why Spanish nukes went abnormally AWOL in the days and weeks before the blackout

ketanjoshi.co · Ketan Joshi

Link author: Ketan Joshi@ketanjoshi.co@ketanjoshi.co

0
0
0
0
0
0
0
0
0
Lambda_ @Lambda_1stLight@qdon.space
Public

이탈리아 브레인랏 확산은 확실히 GPT 무료티어로도 이미지 생성이 가능한게 한몫한듯... 그냥 자기가 생각하는 사물이나 동물 다 융합하고 적당히 '이탈리안 브레인랏 스타일로 만들어줘' 하면 되니까...
심지어 대화 메모리 써써 지금까지의 내 대화 내용을 기반으로 날 이탈리안 브레인랏 스타일로 표현해줘 하고 검색기능 켜놓으면 본인이 이탈리안 브레인랏 검색해서 발라서 만들어줌...

0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
jonny (good kind) @jonny@neuromatch.social
Public

If you want to understand the psychological harm LLMs can do to someone, you have to read conspiracy theory forums. This pattern of the LLM spiraling with you into a private universe of meaning is the overwhelming norm

rollingstone.com/culture/cultu

delusions of grandeur chat gpt ai

People Are Losing Loved Ones to AI-Fueled Spiritual Fantasies

Marriages and families are falling apart as people are sucked into fantasy worlds of spiritual prophecy by AI tools like OpenAI's ChatGPT

www.rollingstone.com · Rolling Stone

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
Elena Rossini ⁂ @_elena@mastodon.social
Public

Good morning Fedi friends!

Please join me in giving a warm welcome to my interns Riyen (@patel.riyenThe Life of Tarzan) & Sam (@samaabergSamuel Aaberg). They will be working with me through the end of June on videos promoting the Fediverse. This is their .

Sam & Riyen are two talented film students who are : a mere 7 days ago they had no idea what the Fediverse was. I have the immense honor of introducing them to it.

We are filming this week! Send us good vibes 🎥✨

Three people sit on a couch. On the left: a young man with black hair and a beard wearing a black blazer and white pants; in the middle: a woman with long brown hair (me) wearing a blue and white striped t-shirt and blue jeans, holding a clapper with the text: "Fediverse Promo no.1 by Elena Riyen & Sam - May 2nd 2025"; on the right there is a young man with glasses and light brown hair wearing a red hooded sweatshirt. A bookshelf and various items are visible in the background.
0
3
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Frau Herbst @thebadfall@troet.cafe
Public

Hab gestern noch ne Talkshow geguckt, da hat ein Gast erzählt, dass sie Zahlen in Farben sieht und da ist es mir wie Schuppen von den Augen gefallen. 🙃Ich auch!! 🤓Aber es war mir gar nicht bewusst. Aber alle Ziffern von 0-9 haben bei mir ne bestimmte Farbe und so bestehen dann alle Zahlen aus diesen Farben. Und ich kann schon immer gut mit Zahlen. Lasse aber auch gerne mal Zahlen. 🤣

0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
那邊那個誰🎃 @Ginger_oo@g0v.social
Public
Shared by MJ+

不斷的練習跟自己對話；不斷的重組自己與這個世界重新連結，卻還是沒有進展。

沒進展也沒關係，可是卻發現自己好像跟這個世界脫節，那種靈魂跟肉體錯開的感覺，一直以為只是過度抽離，直到跟友人聊起近況，他說我可能有輕微解離狀況。

那種對現實不真實感很微妙，感覺自己好像不存在，卻像遊魂。

0
0
0
那邊那個誰🎃 @Ginger_oo@g0v.social
Public

不斷的練習跟自己對話；不斷的重組自己與這個世界重新連結，卻還是沒有進展。

沒進展也沒關係，可是卻發現自己好像跟這個世界脫節，那種靈魂跟肉體錯開的感覺，一直以為只是過度抽離，直到跟友人聊起近況，他說我可能有輕微解離狀況。

那種對現實不真實感很微妙，感覺自己好像不存在，卻像遊魂。

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hacker News 100 @hn100@social.lansky.name
Public

Matrix-vector multiplication implemented in off-the-shelf DRAM for Low-Bit LLMs

Link: arxiv.org/abs/2503.23817
Discussion: news.ycombinator.com/item?id=4

arXiv logo

MVDRAM: Enabling GeMV Execution in Unmodified DRAM for Low-Bit LLM Acceleration

General matrix-vector multiplication (GeMV) remains a critical latency bottleneck in large language model (LLM) inference, even with quantized low-bit models. Processing-Using-DRAM (PUD), an analog in-DRAM computing technique, has the potential to repurpose on-device DRAM as a GeMV engine, offering additional high-throughput processing capabilities to widespread consumer devices without DRAM modifications. However, applying PUD to GeMV operations in the LLM inference pipeline incurs significant overheads $\textit{before}$ and $\textit{after}$ in-DRAM computation, diminishing the benefits of its high-throughput processing capabilities. This paper presents MVDRAM, the first practical system to accelerate GeMV operations for low-bit LLM inference using unmodified DRAM. By leveraging the data sharing patterns and mathematical linearity in GeMV operations, MVDRAM orchestrates the processor and DRAM to eliminate the costs associated with pre-arranging inputs and bit-transposition of outputs required in conventional PUD approaches. Our experimental evaluation with four DDR4 DRAM modules shows that MVDRAM achieves comparable or even better inference speed than the processor-based implementation for GeMV operations in low-bit (under 4-bit) LLM. In particular, MVDRAM achieves up to 7.29$\times$ speedup and 30.5$\times$ energy efficiency for low-bit GeMV operations. For end-to-end LLM inference, MVDRAM achieves 2.18$\times$ and 1.31$\times$ throughput improvements, along with 3.04$\times$ and 2.35$\times$ energy efficiency, for 2-bit and 4-bit quantized low-bit models, respectively. MVDRAM has the potential to redefine the AI hardware landscape by demonstrating the feasibility of standard DRAM as an LLM accelerator.

arxiv.org · arXiv.org

0
0
0