News for hackers. Stuff that matters.

moss_dog

Re:My Agent Setup on Thursday August 13, 2026 @04:21PM (#49291327)
Thanks for the write-up! I'm interested in doing something similar. How are you sandboxing the agents?
Re:1worldflag: A blue dot on a transparent background on Monday June 08, 2026 @12:57AM (#48441460)
What do you mean by "wrong"?

I reckon the actual color doesn't matter as long as it's blue-ish. There are no branding guidelines or anything. Even the size of the circle is unspecified, which IMO is a feature, not a bug. It's accessible :)

EDIT: I'm incorrect, there is a spec. Whelp! I still like my imagined more-flexible version of the flag.

Re:Magic the Gathering format: Fun 40 (2025) on Thursday May 21, 2026 @12:16PM (#48225187)
Yes, Tiny Leaders is awesome!

My own lil custom 40 card format is actually explicitly designed to be backwards compatible with any existing commander deck! You can just take a deck, draft from it (or just split it down the middle after a good shuffle), give one of the decks to a friend, and start playing!

Some more info here (if a little self-promo is OK, as a treat): https://scry.fish/microEDH

Re:Magic the Gathering format: Fun 40 (2025) on Thursday May 21, 2026 @11:16AM (#48224225)
Agreed. Playing the same pair of decks with a friend over and over is so fun -- you really start to learn the decks, and the early game becomes extremely tactical.
Re:Magic the Gathering format: Fun 40 (2025) on Thursday May 21, 2026 @11:10AM (#48224086)
This is cool. 40 card decks are great! In addition to being physically ergonomic, I find 40 cards to be a fun deck-building size. With about 17 lands, you get 23 choices to make of what to include outside of that, which feels like a sweet spot between deck-building expressiveness and decision fatigue.

As an aside, I'm convinced that a big reason WotC (and FLGS) are pushing commander so hard is because 100 card decks means you get to sell more cards.

Re:Book Review: There Is No Antimemetics Division on Monday April 06, 2026 @12:24PM (#47662997)
I think this comment is unnecessarily harsh.

To anyone confused (like me), the commenters above had opposite recommendations despite having similar opinions of the book.

Re:Show HN: A game where you build a GPU on Sunday April 05, 2026 @02:28AM (#47646666)
Wow, I didn't realize Zachtronics made Infiniminer. Fascinating!
It's tagged nonfiction.
Great read! Thanks for sharing.
Fantastic, thanks for sharing! I hadn't heard of this before. Very entertaining video!
Great explanation, thanks!
Re:Why isn't LA repaving streets? on Thursday February 26, 2026 @01:09AM (#47162478)
It's interesting that you frame building bike lanes and reducing car speeds as "punishing drivers", seemingly ignoring that such changes enable micro-mobility users to commute to work safely.
Re:A sane but bull case on Clawdbot / OpenClaw on Wednesday February 04, 2026 @10:03AM (#46886673)
I'm generally of the opinion that capitalization is not necessary in many cases, such as at the start of sentences. That's what punctuation is for :)
Re:Flying Around the World in under 80 Days on Tuesday February 03, 2026 @01:35AM (#46867327)
Delightful read, thanks for sharing. The website design is very pleasant, too!

Also FYI the link to "Aves Æternæ" is broken.

I prefer xcancel in part because Twitter doesn't let you view replies etc when not logged in.
Re:Exactitude in Science – Borges (1946) [pdf] on Monday January 26, 2026 @04:04PM (#46771511)
Maybe Eco's "On the Impossibility of Drawing a Map of the Empire on a Scale of 1 to 1"?
Re:Exactitude in Science – Borges (1946) [pdf] on Monday January 26, 2026 @01:08PM (#46769240)
While we're Borges-posting, I recommend "Tlön, Uqbar, Orbis Tertius"

https://sites.evergreen.edu/politicalshakespeares/wp-content...

What's that?
Re:How I Left YouTube on Wednesday December 24, 2025 @06:55PM (#46380640)
IIRC HN typically removes the "How" from article titles like these, presumably to avoid clickbait titles.
Re:Qntm's Power Tower Toy on Wednesday December 24, 2025 @06:51PM (#46380609)
One of my favorites!
I was looking into this earlier -- presumably you'd also need to allowlist Claude itself (whatever endpoints it hits to run inference etc). VM firewall gets a little trickier with Claude's web search tool, too.

The solution I landed on recently was to locally modify the Chrome devtools MCP to launch the browser instance with strict network restrictions. I believe the implementation used `--host-resolver-rules`, blocking all URLs by default with an environment variable to control the allowlist (which, in hindsight, Claude can easily work around if it needs to -- I should probably just hard-code the allowlist).

Thank you for the links / info! I'm looking forward to digging into this.
I'd love to be able to lock down the browser to only allow certain URLs (e.g. localhost) so I can give Claude (and other tools) carte blanche to use browser automation (rather than manually approving each command). Is this something on your radar / roadmap?
I use Navidrome [1] + an rclone'd S3 bucket on a Digital Ocean droplet. Works quite well for my purposes!

[1] https://github.com/navidrome/navidrome

Re:Meta Segment Anything Model Audio on Thursday December 18, 2025 @10:42PM (#46322065)
This is incredible! I wouldn't have thought it was possible to cleanly separate tracks like that. I wonder to what extent the model is filling in gaps, akin to Samsung's "ultra zoom" moon.
Any notable examples you can share?
Re:Ivan Sutherland Sketchpad Demo 1963 [video] on Saturday December 06, 2025 @06:08PM (#46177432)
"it's real nightmare material" (11:18)
Re:Apple Releases Open Weights Video Model on Tuesday December 02, 2025 @09:03AM (#46121254)
Thanks for the recommendation, just downloaded a few episodes!;
Re:Europe is scaling back its landmark privacy and AI laws on Wednesday November 19, 2025 @04:21PM (#45985385)
I wish there was a link to the source of this information in the article! I'd like to read the updated version of these laws (if they're public).

Older »