frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Agent Bazaar: Enabling Economic Alignment in Multi-Agent Marketplaces

https://www.chatpaper.ai/dashboard/paper/80941351-e4d5-4566-a458-aa93c4f7dbff
1•doener•2m ago•0 comments

How to Speed Up Phrase Search with Bigram_index

https://manticoresearch.com/blog/how-to-speed-up-phrase-search-with-bigram-index/
1•snikolaev•2m ago•0 comments

The Economics of Human Extinction

https://www.andrewleigh.com/speech_the_economics_of_human_extinction_21_may_2026
1•ajdlinux•4m ago•0 comments

Show HN: The-knowledge-guy – ask, walk, and skim every book on your shelf

https://github.com/vitalysim/the-knowledge-guy
1•vitalysim•4m ago•0 comments

We Backtest High-Frequency Options Spread Strategies with Volatility Timing

https://medium.com/@DolphinDB_Inc/backtesting-application-medium-and-high-frequency-options-sprea...
1•CrazyTomato•6m ago•0 comments

What it was like working on LLMs and security at Meta (2022-2026)

https://joshuasaxe181906.substack.com/p/what-it-was-like-working-on-llms
2•takira•9m ago•0 comments

AltTab Pro

https://alt-tab.app
1•frizlab•10m ago•1 comments

CAISI states open models lag behind the American frontier, with the gap widening

https://www.nist.gov/news-events/news/2026/05/caisi-evaluation-deepseek-v4-pro
1•nilen•12m ago•0 comments

Ask HN: Is anyone using vibe coding harnesses like OmO or Ralph loops at work?

1•choam•15m ago•0 comments

AI red teaming agents change how LLMs get tested

https://www.helpnetsecurity.com/2026/05/21/ai-red-teaming-agents-research/
1•SVI•15m ago•0 comments

AVX-512 and Validating Usage on AMD EPYC

https://www.amd.com/en/blogs/2026/understanding-avx-512---validating-usage-on-amd-epyc-.html
1•tosh•17m ago•0 comments

Google search AI mode, the biggest update

https://www.usatoday.com/story/tech/2026/05/20/google-search-ai-mode-biggest-update/90180507007/
1•aledevv•19m ago•0 comments

Show HN: Gemini Omni Flash access notes and AI video generator

https://omnigemini.video
1•howardV•19m ago•0 comments

Why Optimistic Merging Works Better

http://hintjens.com/blog:1061
2•jimsojim•24m ago•0 comments

Nvidia Rides Blistering Chip Sales to Another Record Quarter

https://www.wsj.com/tech/ai/nvidia-nvda-1st-quarter-earnings-report-2026-stock-c2bb9c1c
2•andsoitis•25m ago•0 comments

Fate: Building an Astro Blog with a Modern React Data Client

https://jsdev.space/meet-fate-react-astro/
1•javatuts•26m ago•0 comments

Twelve Ways to Be Wrong About AI-Assisted Coding

https://third-bit.com/2026/05/20/twelve-ways-to-be-wrong/
2•calcifer•27m ago•0 comments

Can ChatGPT order a jumbo breakfast roll without messing up?

https://www.rte.ie/brainstorm/2026/0520/1574290-chat-gpt-breakfast-roll-irish-english-dialect-phr...
2•austinallegro•30m ago•0 comments

GraphMyCode –Visualize any codebase as an interactive knowledge graph,in-browser

https://graphmycode.com
1•FValero•30m ago•0 comments

How to Build a Tech Startup When You Barely Know What a Boolean Is

https://bwanaerp.medium.com/how-to-build-a-tech-startup-when-you-barely-know-what-a-boolean-is-vi...
1•instarlaxy•31m ago•0 comments

Unix has been changing, but in places where I don't see it

https://utcc.utoronto.ca/~cks/space/blog/unix/UnixChangingQuietly
1•ingve•32m ago•0 comments

Immigrants taking your jobs? [short video]

https://www.youtube.com/shorts/hrej-oY3xbg
1•stiray•32m ago•1 comments

"Don't Be Evil" Was a Lie from the Start (2024)

https://old.reddit.com/r/degoogle/comments/1g5uepg/dont_be_evil_was_a_lie_from_the_start_google/
3•huijzer•37m ago•2 comments

Don't Build a General Purpose API (4 Years Later)

https://max.engineer/server-informed-ui-p2
1•goranmoomin•39m ago•0 comments

The physicists who convinced Fermilab to send Brazil's emails

https://buttondown.com/blog/brazil-fermilab-email
1•maguay•45m ago•0 comments

Economics Lessons from Home Depot

https://www.economist.com/finance-and-economics/2026/05/20/economics-lessons-from-home-depot
1•andsoitis•45m ago•0 comments

Linki – open-source AI SDR for LinkedIn sequences and cold email

https://github.com/moaljumaa/linki
3•moisacow•45m ago•1 comments

How to scrape any website and get structured data with a single API call

https://spidra.io/blog/scrape-any-website-structured-data-single-api-call
2•joelolawanle•47m ago•0 comments

Are TypeScript back end frameworks ready for AI Agents?

https://encore.dev/blog/ai-benchmark
3•dohguy•48m ago•0 comments

Docket – another beautiful, minimal task manager that lives in macOS menu bar

https://github.com/santoru/docket
1•meliot•49m ago•0 comments
Open in hackernews

Passwords are okay, impulsive Internet isn't

https://www.dedoimedo.com/life/passwords-passkeys.html
3•brycewray•1y ago

Comments

palata•1y ago
Hmm... I see a rant against the state of software (bad software, AI diarrhea, ...) and TooBigTech having control over everything. I can agree with that, but it has nothing to do with the "passwords vs passkeys" question.

The rant against passkeys? I don't get it. Just like one can use a password manager controlled by TooBigTech or KeePass, one can use a passkey controlled by TooBigTech or a Yubikey. I find it great to authenticate directly with my Yubikey (over FIDO2) instead of using my Yubikey to decrypt a password and copying it in a form.

And then there is the part that is completely wrong about security. They say that they "can't trust their phone" so they don't want to keep the passkeys there. But that is not correct: if the passkeys are encrypted and the key is stored in a TPM, then that's effectively similar to having a security key (you have to trust the TPM, just as you have to trust the security key of course).

And then there is the nonsense:

> I can set up KeePass Portable on a USB key, run it in Linux via WINE, place it inside an encrypted VeraCrypt container, copy to any which file sharing service, if I want.

If the device where you enter the password is compromised, then the password will be compromised as soon as you enter it on that device. No matter how much you show off with your funny setup with WINE and VeraCrypt. A password manager doesn't protect against that, so passwords can be exfiltrated as they are used. Whereas a FIDO2 authentication requires the passkey every time. E.g. I need to physically touch my Yubikey for it to sign the challenge. It could be MitM, but it is visible ("I touched my Yubikey and it didn't work, what happened?").

Authenticating over FIDO2 with a security key is strictly superior to entering a password in a field, period.