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
1

Another tidbit from the report I'm writing about :

The 52 issues filled correspond roughly to 30% of all the crashes (issues with "type-crash" label) and 2% of all issues (features requests, bugs and invalid issues) reported in the issue tracker during the six months period covered by the report.

Hits and new issues don't appear at a steady pace. It seems that there are long periods of no or nearly no new issues, followed by rapid finding of new results.

A bar graph of issues filed by week number, with X axis ranging from week 44 to week 19 and Y axis ranging from zero to 10 issues. The distribution looks bi-modal, with a clump of high numbers of issues per week from week 44 to 50, zero issues for weeks 51 to 11 (except by two issues in week 6), and a smaller clump of medium numbers of issues from week 12 to 16.
0
1
0

We’re rolling out Account Status on Threads to help you understand what’s going on with your posts and profile, directly in the app.

🔍 More transparency: See actions we’ve taken on any of your posts or replies that violate our Community Standards.

⚠️ More control: Let us know if you think we’ve made a mistake. You’ll get a notification once we review your report.
#New on Threads

An image of an Account Status page showing a menu titled Actions we've taken that features Removed Content, What can't be recommended, Content lowered in feed, and Features you can't use.
0
0
0
0
1

‘벗방’ 했으면 가짜 레즈비언인가···그 인생을 물어뜯을 자격 있나 [이진송의 아니 근데] www.khan.co.kr/article/2025... "성녀/창녀 이분법은 교묘하고 집요하다. 가부장제는 여성들이 언제든 발만 삐끗해도 ‘창녀’로 떨어질 수 있는 수많은 장치를 마련해놓고 ‘네가 창녀가 아니면 괜찮다고’ 여성을 안심 시킨다. 하지만 정말 그럴까? ‘아무리 힘들어도’ 성산업에는 혹하지 않고, 그래서 보호받아 마땅한 여성이 되면, 타락한 여성을 마음껏 욕해도 될까?"

[이진송의 아니 근데] ‘벗방’ 했으면 가짜 레즈비언인...

0
1
1
0

ルーツのある人は全て自国人とみなす。トランプ政権の日本ムーブ再び。

「初となる米国出身の教皇の誕生を受けて、トランプ米大統領は「米国にとって大変な栄誉」と歓迎した」

「プレボスト氏はイリノイ州シカゴ出身。南米で長らく宣教活動を行い、ペルーのチクラヨで司祭を務めた。2015年にはペルー国籍を取得している」

cnn.co.jp/world/35232750.html

0

Mistral AIのLe Chatまあまあいいのでは、という気がしてきた。Mistral AIはLLMを作ってるフランスのスタートアップ。

モデル名はだいたいMistralなんとかって名前なんだけど、チャットAIのサービス名がLe Chatで、Web検索機能・ニュース機能・画像生成ありで、プロジェクト機能みたいな機能がある。んで、プランは無料の次が$14.99(税抜)で基本的には無制限という感じ。

ニュース検索ちょっとやってみたら日本のニュースも出てきたし、$15なのはいいかも。iOSのアプリはあるけどmacOSのアプリはないようなので、そこが少し弱いか。LLM自体の能力はどうなんだろうね、無料で軽く触ってみた分には悪くなかったけども。

mistral.ai/products/le-chat

0
0
0

遊戲中有個稍有份量角色,似乎在玩家間有著強烈的好惡之分,尤其是年輕的族群。我現在理解為在遊戲中,人們希望能成為一個自己期望的樣子,而這個角色被塑形得太像現實中的自己,因而產生反感。畢竟對有些人而言,遊戲是逃離這個社會的唯一出口啊。 :bunny_hopping:

0
0
0

Mistral AIのLe Chatまあまあいいのでは、という気がしてきた。Mistral AIはLLMを作ってるフランスのスタートアップ。

モデル名はだいたいMistralなんとかって名前なんだけど、チャットAIのサービス名がLe Chatで、Web検索機能・ニュース機能・画像生成ありで、プロジェクト機能みたいな機能がある。んで、プランは無料の次が$14.99(税抜)で基本的には無制限という感じ。

ニュース検索ちょっとやってみたら日本のニュースも出てきたし、$15なのはいいかも。iOSのアプリはあるけどmacOSのアプリはないようなので、そこが少し弱いか。LLM自体の能力はどうなんだろうね、無料で軽く触ってみた分には悪くなかったけども。

mistral.ai/products/le-chat

0

It's been a while, but our Thursday Thunderbird Tips are back! And our first post is all about Thunderbird keyboard shortcuts. If you're already using them, awesome! If not, we’ve got a guide to get you started! 🤔 ⌨️

support.mozilla.org/kb/keyboar

0
0
0
0
0
0
0

In Which the Author Complains about Google's Circle to Search Feature: A Diatribe on its Peculiar Exclusivity, and a Note on Google Lens

so android has a feature called "circle to search" where you can take a screenshot of photo and circle things to highlight them, and google will tell you about what you're looking at. for example, take a photo of a cable and circle it to learn that it's a HDMI cable. cool cool cool.

except this feature is only supported on specific allowlisted models. if your phone isn't on the list, it doesn't work. and google seemingly doesn't publish the list. i found an article that extracted a list of devices that are "ready" for circle to search from google play console, and my phone's on the list, but it doesn't have the feature. (yes, my phone is up to date, including google play system updates.)

but you know what i can do on my phone? i can take a photo or screenshot, share it to google lens, and draw a box around an item to learn about what it is. so there's an extra step involved in having to share the image, and i can only draw boxes and not circles... but it works fine. i can do exactly the same thing that circle to search does, it's just clunkier and more obtuse.

why. why does google deem my phone unworthy of drawing a circle. why is my phone in the "ready for circle to search" list despite not having the feature enabled. why is there an allowlist of devices.

see, if this was exclusive to the latest pixels, i'd understand. they're trying to get you to upgrade your old pixel or swap to their ecosystem for this cool feature. it's a unique selling point. whatever. but it's not pixel exclusive! if i upgrade my overpriced samsung to an even more overpriced samsung, google doesn't get anything from me they don't already have! what's the motive?? i could even downgrade to an old galaxy S and still get this feature!

maybe i'm expecting too much from the company that sells a version of Android TV called Google TV while also renaming the Play Movies & TV app to Google TV which has nothing to do with YouTube TV, a streaming subscription for YouTube that lets you watch broadcast television on your Google TV or Android TV or your Google Pixel (Powered by Android) or your Android device with Google Play Services; YouTube, of course, being an app where you can purchase movies and TV shows.

0

gen AI

I'm tempted to play around with an AI code generation "agent" like Aider or Claude Code. But that feels like eating the forbidden fruit. It's easier to stand against something on principle if you haven't already been lured in by its immediate benefits. I've already crossed that line with LLM use in general, including asking an LLM to do one-off code generation tasks (though never for my open-source work). Maybe I can at least resist the temptation to try the next level.

0

‼️Hey! Just launched 29 beta 1
Apart from bug fixes (yay), some new things👇

- New iconset
- Layout of polls improved and a "show results" button added
- Support for the "blur" filter action
- Quality of the image viewer improved
- An additional dialog will prevent accidentally dismissing the media caption dialog
- Boosts and favorites now confirmed via a bottom sheet, not a drop down
- Support for Android 16
- Improved localizations

First on Google Play, later on F-droid!

0
0
0
0


“2022年に射殺されたアリゾナ州のクリス・ペルキーさんが、AIによって生成された映像を通じて裁判で加害者に向けて意見陳述を行った。録音や写真を基に遺族がAIを用いて作成したもので、許しを語る内容だった。判事はその意義を評価したが、専門家の間では今後のAI活用に対する懸念もある。”

(アリゾナで射殺された男性がAIで「復活」、裁判で意見陳述)
◆射殺されたアリゾナ州の男性、裁判で被害者として意見陳述 AIで動画生成
bbc.com/japanese/articles/cx2q

0

일본비평 '부부의 씨(氏) 통일에 관한 일본인의 규범의식: 사회변화에 따른 법·제도와 자유·권리의 간극' 보는 중. 1956년 심의에서 어떤 교수가 부부별성으로 바꾸자는 제안을 냈더니 위원회 사람들이 성씨는 기호에 불과하기 때문에 바꿀 필요가 없다고 했는데 그러자 그 교수가 성씨가 기호에 불과하면 아내쪽을 따르는 부부동성을 하자고 했더니 그 사람들이 무슨 소리냐고 반발했다고... ijs.snu.ac.kr/publications/kor

0
0
1
0

[完了][OpenSearchなどを使わず簡易な方法でembeddingモデルを動かして評価する]
 
<動機>
セマンティック検索について試行錯誤したいがFessやOpenSearchの機能や仕様を調べるコストが高い。シンプルなPythonスクリプトからembeddingモデルを直接使って、どういう使い方をすればうまくいくのかを探りたい

<結論>
シンプルなRAGツールを作った。multilingual-e5-large-instは充分な精度が出た。チャンクの切り方などに気をつけるのがいい。LangChainがやはり便利。ローカルで完結するRAGは容易に実現可能。エンジニアでなくとも使えるようにできるとなおよい。
gist.github.com/g-fukurowl/81c

0
0

대통령 탄핵 선고 수업이 “편향됐다”는 서울시의원…교실은 정치적 무균실이어야 할까요? [뉴스물음표] www.khan.co.kr/article/2025... "인천 계양구에 사는 고등학교 1학년 A양은 “고등학교 때까지 정치·사회 이야기를 하지 않다가 갑자기 어른이 된다고 해서 생각이 다른 사람과 건강하게 대화를 나눌 순 없을 것”이라며 “학교에서 이번처럼 실생활에 맞닿은 토론을 하는 기회가 생기면 도움이 될 것 같다. 잘못된 논리를 고치는 방법을 모르는 채로 어른이 되는 게 더 두렵다”고 했습니다."

대통령 탄핵 선고 수업이 “편향됐다”는 서울시의원…교실...

0

転職してから厳密なクライアントワークからは少し遠ざかっていたから、ちょっとこの辺の世の中の雰囲気がわからないんだけど、こういうことしちゃうクライアントって結構いるのかな…?ディストピア…

こんなことするなら、もうはじめからデザイナーに依頼せずに自分で生成AIで作ればいいんじゃないか?と思ってしまった。

> (1) デザインの提案したら、その案をチャットGPTに投げて作った変更点がクライアントさんから送られてきたんだけど 作風を盗まれてしまうので、生成AIにデザインを投げるのはやめてほしいと伝えたことが、結果的に気分を害されてしまって取引そのものがうまくいかなくなってしまいました これデザインをしている方、どう考えますか? 取引する前に、事前にAIに投げるのは禁止という感じでアナウンスすべきなのか もう仕方がないとあきらめるべきなのか

threads.com/@nishikuradesign/p

1
1
0
0
0
0
1
0

この先よく出てきそうな話だなと思い所感を述べます。

ツールとしてAIを使うとき、AI(ここではLLM)らは選択肢を出すことはできるが決断はしないというのが重要であると思う。

それを人間が「AIの言っていることは絶対、もしくは自分よりベター(この時点の優劣はAIが提示したものと言うファジーなものだが)決断を下せる」と無意識を信じてしまうクライアントは、

AIに使われているのと同義のような気がする。なぜなら製作者へ情報を渡すだけの役になるから…。

RE: https://www.threads.com/@nishikuradesign/post/DJWF0RhRCWB

0
0