frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

The fix can't wait. What breaks when it lands?

https://irlabs.ai/blog/the-fix-cant-wait/
2•marssaxman•8m ago

Comments

joehoover•7m ago
Author here. I work at IR Labs, mostly on AI systems for low-level code analysis and verification. Short version: about 1 in 30 kernel fixes breaks something new, and we tested whether you can catch the break from the patch alone. Longer version, for the curious:

When I joined IR Labs in February, we were primarily focused on evidence-based code review. That feels like a lifetime ago, but that was when people were just starting to bang the review-is-the-new-bottleneck drum.

Much has changed since then.

In many organizations, review has shifted closer to a commodity (or, at least, people pretend that it has) and SDLC limiting factors have taken on an air of subjectivity: just fly a little closer to the sun and maybe everything can move that much faster.

However, for the most part, every organization gets to choose how close to the sun they fly. Conservative, risk averse orgs can still move slow and try not to break anything.

Except when they can't.

In this year of AI, there's one thing that no one can afford to go slow on: fixes.

Unfortunately, even in the best of times and most careful of orgs, fixes break things, too.

This seems obvious, pedestrian even. But there was a moment a while back when this struck us, full bore, and we realized there was an oncoming cataclysm of fix-induced failures. Patch slinging frontier models, AI-enhanced adversaries, and mountains of critical software festering with undreamt vulnerabilities would put enormous pressure on fix delivery. Everyone would have to move fast, and things would definitely break.

So, we started working on systems to help folks minimize their fix-driven breakage rate. Rather than just a/b testing a fix against a PoC, we've shifted toward a much more defensive flavor of fix verification that tries to anticipate the unanticipated.

Still early days for us, but we're trying to get better at sharing early and often.

To kick things off, we wanted to share two things.

First, some early results from exploratory analyses of the kernel, where we look at the dynamics of fix-introduced defects. For instance, we found that about 1 in 30 fixes to runtime kernel code broke something that needed its own fix within a year, and some needed several repairs, a few of which needed repairs of their own. We'll have a lot more to say on that topic in the coming weeks.

And second, results of an experiment: how reliably can we detect fix-introduced defects with executable evidence? So far, our system caught 4 of 6 real breaks and 7 of 10 planted ones, from the patch alone.

Again, this is all early (e.g. sample sizes are small) and we're working on a more robust eval report.

If you maintain a kernel subsystem, or anything else where fixes ship under pressure, please poke holes. I'd love to hear where this would fall over.

Fair warning: I'm offline for 10 days starting in about 30 minutes (and will try real hard to stay that way). Teammates will be in the thread.

Show HN: Bocalito – A Meal Prep Planner

https://bocalito.com
1•eskombro•44s ago•0 comments

The Brain–is wider than the Sky

https://blue-continuum.com/the-brain-is-wider-than-the-sky
1•dnetesn•47s ago•0 comments

Waymo stalls in high San Francisco floodwaters during storm

https://ktla.com/news/california/waymo-stalls-san-francisco-floodwaters/
1•mikhael•2m ago•0 comments

Unit-Checker for Excel

https://github.com/psteckler/daubee
1•psteckler•3m ago•0 comments

A New Era of Deterministic AI Begins: OLMapp Emerges from Stealth

https://olmapp.ai/index.html
1•northlondoner•6m ago•1 comments

The fix can't wait. What breaks when it lands?

https://irlabs.ai/blog/the-fix-cant-wait/
2•marssaxman•8m ago•1 comments

Batteries: The Invention That Untethered the World

https://stepchangeshow.substack.com/p/batteries-the-invention-that-untethered
1•toomuchtodo•8m ago•1 comments

P&G Aims to Reduce Manufacturing Workforce Up to 30% by 2030

https://www.bloomberg.com/news/articles/2026-11-19/p-g-aims-to-reduce-manufacturing-workforce-up-...
1•petethomas•9m ago•0 comments

Celliad – a daily Excel game (French)

https://celliad.com/
1•didir98•10m ago•0 comments

An FBI agent called a Claude user about his chats

https://austinpatrick.substack.com/p/when-should-ai-labs-alert-police
2•AustinLikesAI•11m ago•0 comments

Last remaining NIH-owned chimpanzees will retire to a sanctuary next year

https://www.science.org/content/article/last-remaining-nih-owned-chimpanzees-will-retire-sanctuar...
1•gmays•15m ago•0 comments

Show HN: 15MB transcription model that runs in the browser/device

https://starbright.quickthoughts.ca/
2•quickthoughts•15m ago•1 comments

Galleri – Screen for 50 cancers with a single blood test

https://www.galleri.com/
1•numlocked•16m ago•0 comments

The trouble with serendipit: scientific discoveries

https://www.worksinprogress.news/p/the-trouble-with-serendipity
1•greazy•18m ago•0 comments

From 8-bit Amiga to Synthwave: Original tracks composed on an Amiga in the 90s

https://vosmeer.com/
1•doener•20m ago•0 comments

RegExr: Temporarily Offline

https://regexr.com/
5•somebudyelse•20m ago•4 comments

The [uncertain] Four Seasons: Vivaldi re-composed for 2050 climate projections

https://the-uncertain-four-seasons.info/
1•__fst__•21m ago•0 comments

Visa, Mastercard confirm cyberattacks that may have exposed customers' info

https://www.theregister.com/security/2026/11/19/chase-visa-mastercard-confirm-third-party-cyberat...
3•Betelbuddy•21m ago•1 comments

Intuit to Sell Mailchimp to Bending Spoons for $3.68B

https://www.cnbc.com/2026/11/19/intuit-mailchimp-sale-bending-spoons.html
2•jsk2600•23m ago•1 comments

C-to-Rust Translation Validation at the LLVM IR Level via AI-Assisted Proof

https://arxiv.org/abs/2611.18036
1•matt_d•25m ago•0 comments

AMD dives on short-seller report of alleged China exposure

https://www.cnbc.com/2026/11/19/amd-falls-on-short-seller-report-of-alleged-china-exposure.html
2•Betelbuddy•26m ago•0 comments

A Novel AI-Native GPU Kernel Programming Language for Autonomous Kernel Tuning

https://al.radbox.org/doi/10.1145/3828523.3857041
1•matt_d•26m ago•0 comments

Register your app on open-source repositories

https://developer.android.com/developer-verification/guides/open-source-app-registration
2•exploraz•27m ago•0 comments

IPO: Anthropic Bankruptsy Likelier Than Success [video][15m]

https://www.youtube.com/watch?v=a5pHmGq1mEs
4•Bender•29m ago•0 comments

Neoism – Coolest Agent Framework Around Town

https://neoism.ai/
1•Mr_Settle•29m ago•0 comments

Math Books by Robert Ghrist

https://www2.math.upenn.edu/~ghrist/books/
2•ibobev•31m ago•1 comments

Attacker swapped Prime Intellect's DeepSeek v4.1 Flash endpoint for 53 hours

https://runtimewire.com/article/openrouter-notified-users-of-a-reported-prime-intellect-compromise
6•dvaplima•32m ago•0 comments

Synology Security Advisory

https://www.synology.com/en-uk/security/advisory/Synology_SA_26_21
33•choult•34m ago•8 comments

Show HN: A deterministic agent memory engine in Rust with no LLM

https://github.com/samvallad33/vestige
2•entity002•36m ago•1 comments

Learning to Compress a Billion Bytes of Wikipedia

https://tufalabs.ai/research/tufazip/
2•7777777phil•37m ago•0 comments