frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

PEP 829 – Package Startup Configuration Files

https://peps.python.org/pep-0829/
1•zahlman•19h ago

Comments

zahlman•19h ago
tl;dr: a new feature is coming in Python 3.15 whereby the `import` lines (which allow arbitrary code execution by design) in .pth files will be skipped whenever a corresponding .start file is found in the same folder; this file specifies one or more entry points to code in the environment (which will run after paths have been collected and added from the .pth files). This is intended to separate responsibilities and make the environment easier to audit. However, it doesn't protect against the sort of malice we've seen with Shai-Hulud type attacks.

There is a deprecation plan whereby the `import` lines in .pth files will become silently ignored (even if there is no .start file) in 3.18, and then start emitting a warning in 3.20. This seems entirely backwards to me, but whatever. (The natural approach, IMO, would be to start warning now while still processing them, and later remove the code entirely such that the import lines then become invalid paths.)

Show HN: Fable,Kimi,GPT5.6 FFA naval war,airstrikes and radio diplomacy [video]

https://www.youtube.com/watch?v=nFxRtjgKtY0
1•masterchef2209•2m ago•0 comments

LA to Farnborough UK in Under 9h Max Speed Mach 0.93

https://bombardier.com/en/media/news/bombardiers-global-8000-worlds-fastest-civil-aircraft-sets-n...
2•r2sk5t•2m ago•0 comments

Why Microsoft's $80B Xbox Bet Backfired

https://www.wsj.com/video/why-microsoft-80-billion-xbox-bet-backfired/1CE7B4ED-CFBE-4DD8-A2AB-CC9...
2•1vuio0pswjnm7•3m ago•0 comments

Cue AI

https://deepmind.google/models/gemma/gemmaverse/cue-ai/
1•logickkk1•5m ago•0 comments

The /Goal Function Made Loop Engineering Click for Me

https://bhart.org/articles/the-goal-function-made-loop-engineering-click-for-me
1•yruzin•7m ago•0 comments

Tell HN: Old Reddit Now Requires Login

3•quincepie•7m ago•1 comments

Show HN: Lagotto Meter – how well do AI agents understand your website?

https://lake8.dev/lagotto-meter/
1•tommy2970•8m ago•0 comments

The Supply Deals Feeding the AI Frenzy Are No Sure Thing

https://www.wsj.com/finance/the-massive-supply-deals-feeding-the-ai-frenzy-are-no-sure-thing-3a89...
1•1vuio0pswjnm7•9m ago•0 comments

FCC plans ID mandate that could block anonymous use of prepaid burner phones

https://arstechnica.com/tech-policy/2026/06/fcc-plans-id-mandate-that-could-block-anonymous-use-o...
2•dynm•10m ago•0 comments

State of In-App Subscriptions

https://adapty.io/state-of-in-app-subscriptions/
1•kamphey•10m ago•0 comments

Reverse-engineering is cheap now

https://simonwillison.net/2026/Jul/20/cheap-reverse-engineering/
1•ingve•10m ago•0 comments

Make Your AI Agent Durable in 8 Minutes [video]

https://www.youtube.com/watch?v=3nhRaJ3eOS8
3•aldersondev•11m ago•1 comments

Using GPT Codex, DeepSeek V4 and Kimi K3 on a Real OSS Project

https://github.com/ikaruscareer/SafeAI
1•ikaruscareer•12m ago•1 comments

NixOS is more complicated than you think but OpenCode fixed it

https://grigio.org/nixos-is-more-complicated-than-you-think-but-opencode-fixed-it/
1•grigio•14m ago•0 comments

Every Microsecond Matters:Achieving Near SpeedOfLight Latency in GPU Collectives

https://arxiv.org/abs/2607.16100
1•matt_d•17m ago•0 comments

Kimi K3 just fixed 15 critical security bugs that Codex and Fable refused

https://twitter.com/DavidSacks/status/2078984980588531855
3•nailer•17m ago•0 comments

Chunk of sea ice has not refrozen in West Antarctica, satellite images show

https://abcnews.com/International/massive-chunk-sea-ice-refrozen-west-antarctica-satellite/story?...
3•littlexsparkee•18m ago•0 comments

Unsloth: Introducing AMD support

https://github.com/unslothai/unsloth/releases/tag/v0.1.50-beta
1•wertyk•18m ago•0 comments

Head of US Commerce Dept.'s AI safety arm resigns

https://thehill.com/policy/technology/5978770-chris-fall-caisi-resigns/
4•classichasclass•19m ago•0 comments

Show HN: Apollo 11 Tranquility Base Archive with Analyses

https://ulix.app/ulixcortex/apollo11/
2•bethanyhunt•21m ago•0 comments

GraphRAG is overkill. This is what we do instead for knowledge graphs

https://www.minimumviablefounder.com/p/why-ai-company-brains-fail
2•hackfather•23m ago•1 comments

The UX of Agent Tools

https://sys0.substack.com/publish/post/207820620
1•shenli3514•24m ago•0 comments

EURO6000 locomotives approved to run in Belgium and Luxembourg

https://www.railadvent.co.uk/2026/07/euro6000-locomotives-approved-to-run-in-belgium-and-luxembou...
2•elDutie•26m ago•0 comments

Bringing agent-ready simulation into blender [video[

https://www.youtube.com/watch?v=XdtQbMXHDjQ
1•me_bx•27m ago•0 comments

Mister - Recreates classic computers, video game consoles, and arcade machines

https://mister-devel.github.io/MkDocs_MiSTer/
3•gregsadetsky•28m ago•0 comments

What counts as a replication?: I'm confused about how we validate science now

https://chillphysicsenjoyer.substack.com/p/what-counts-as-a-replication
3•crescit_eundo•30m ago•0 comments

Meta-Agent in a Coding Harness

https://ampcode.com/news/meet-puck
1•feifan•31m ago•0 comments

AI Red Teaming: Securing Agentic AI Systems (video)

https://alice.io/webinars/it-takes-ai-to-break-ai-the-case-for-ai-red-teaming
1•gk1•32m ago•0 comments

Digitally Entertained with Zero Money

https://folkmoss.bearblog.dev/digital-entertainment-with-zero-money/
1•ExMachina73•33m ago•0 comments

AI solves 20 year old conjecture in graph theory

https://twitter.com/marsxiang_/status/2079276968047153337
10•never_giveup•36m ago•2 comments