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.

1
0
0

Due to me not understanding how to create Linux from Scratch...Damn, reproducing a NixOS install via command line is like magic (well documented magic, naturally). It's entertaining watching this man do this for 29 minutes and some change!

The wild thing is, that if I tried, I could do this too! LOL If I felt confident enough to create a WHOLE ASS script and learn Nix OS's computer whispering language.

youtu.be/2QjzI5dXwDY

0
0
2
0
1
1
1
0

「ゆるい繋がり」は、相応のモデレーションと集団的成熟がないと、すぐに「ガバガバ責任感とギチギチ視点による繋がり」に変質しますね。

私も20代はよく失敗しました。Discordサーバーとか、オフラインとか

1
0

「ブリュッセルはオープンソース推進で欧州を大手テックから引き離す計画
/Call for Evidenceは、FOSSを米国の依存を打破する手段として提示している 」:

「欧州委員会はオープンソースに関する新たな協議を開始し、欧州の開発者コミュニティが米国の巨大テクノロジー企業のプラットフォームを支える以上のことを成し遂げるという野心を示した。

今週発表された 「証拠の提出を求める声明」の中で 、ブリュッセルは、EUが非欧州の技術サプライヤー(つまり米国の巨大IT企業)に依存していることが戦略的負債となり、選択肢を制限し、競争力を弱め、クラウドサービスから重要インフラに至るまであらゆる分野でサプライチェーンリスクを生み出していると述べています。1月6日から2月3日まで行われるこの 協議は 、「欧州オープンデジタルエコシステム」に関する正式な戦略策定に向けた初期段階であり、オープンソースを「あれば良い」ものではなく、中核インフラとして扱うことを目指しています。 」

theregister.com/2026/01/11/eu_

0
0
1
0
1
0
1

Extermination in the Gaza Strip is no longer carried out through bombing alone by Israel. Alongside the bombs, a quieter and more deadly policy is unfolding: the management of starvation as a tool of slow death, implemented through careful control of the quality of food and aid.

Strict restrictions have been imposed on the entry of therapeutic food, nutritional supplements, baby milk, vitamins, and essential medicines.

thecanary.co/global/world-anal

🕎 🇵🇸 ☮️

0
0

I want to join the call for justice for Renee Nicole Good, the woman who was murdered today by ICE thugs in Minneapolis...

But true justice would mean not being shot and killed by fascists for no reason in the first place. There's nothing just about this situation and there never will be.

I don't ask for justice.

I want accountability for every crime committed by ICE and for the corrupt politicians who directed them.

0
0
0
1
0
1
0
0

I dunno, opposing fascism is dangerous. Anyone who tells you otherwise doesn't understand what fascism is. The line between what is legal, and illegal, shifts considerably under any fascist regime over time, and their justice systems (funny word there) are both punitive and retroactive. They are probably already going to beat the shit out of you and throw you in prison on a long enough timeline for shit you've already done.

Opposing fascism is dangerous. It's also necessary. Just how it be.

0
1
0
0

“윤석열, ‘사형’ 훈장으로 여길 것”…서울대 로스쿨 교수 경고 n.news.naver.com/mnews/articl... 한 교수는 12·3 내란사태와 윤 전 대통령을 강하게 비판해 온 형사법 전문가로, 대표적인 사형제 폐지론자다. 사형이 윤 전 대통령에게 ‘순교자 서사’를 부여하는 부작용을 낳을 수 있다는 점도 문제점으로 꼽힌다. “다만 법정-실질 최종형이 무기징역 미만으로 내려갈 때는 분노 경악할 이유가 충분하다고 생각한다"

“윤석열, ‘사형’ 훈장으로 여길 것”…서울대 로스쿨 ...

0
0
1
0
0
0
0
0
0

Today I pushed a small fix for mtw: the MediaTek MT7601U USB WiFi driver in FreeBSD.

The issue: driver worked perfectly on first boot. Connect the adapter, it initializes, you're online. Great. Then you reboot your machine and... nothing. "timeout waiting for MCU to initialize." The only fix was to physically unplug the adapter and plug it back in, or do a full cold boot (power off completely).

The root cause was a state machine transition problem. When you do a warm reboot, USB devices don't lose power - they just sit there, happily retaining whatever state they were in. The MT7601U's microcontroller still had its "ready" flag set from the previous session. So when the driver came back up and checked "is firmware already running?", the device said "yep!" and the driver said "cool, nothing to do" and returned early.

Except the device wasn't actually ready. It was in a Schrödinger state - simultaneously "ready" and "not ready" until you observe it by trying to actually probe it, at which point the wavefunction collapses into "definitely broken." Technically powered, firmware loaded, flags set, but not actually functional.

The fix was straightforward: if we detect the MCU is already marked ready at attach time, assume it's stale state from a previous session, reset the MCU, clear the ready flag, and do a proper initialization from scratch. Force the cat to be alive.

Also added USB re-enumeration on attach and bumped some timeouts because these devices can be slow after a reset.

What's funny about driver development is that progress isn't linear. You find and fix a bug in hardware revision 1 and 2. Revision 3 ships and works perfectly - they must have fixed it in silicon, you think. Then revision 6 manufactured a year later shows up with the exact same bug. Different fab? Cost reduction? Recycled design from a different product line? Who knows.

You end up maintaining a mental map of "this chip from this date range has this quirk, but the same chip from that date range doesn't, except if it's from this OEM who used a different reference design."

And the datasheets - when they exist - are always for the "ideal" version of the hardware that exists only in the minds of the engineers who wrote them.

Anyway, if you're running FreeBSD with an MT7601U adapter and wondered why you had to unplug it after every reboot, this one's for you.

0
1
1
0

Good news:
My Zaurus blog post got on OSNews! Cool!
Bad news:
The RSS on my site is broken!

For some reason, Zola (the site gen I use) doesn't seem to build the RSS feed when I'm on OpenBSD. It works fine on Linux. I don't have time to figure that out right now, but I will ASAP.

Here's a short blog post about my move: goldfishdev.net/posts/movingth

Special thanks to @OpenBSDAmsOpenBSD Amsterdam and httpd(8)!

0
0
2
1
0
0
2