Hacker News: swinglock

New comment by swinglock in "Kagi added a setting for removing paywalled links from search results"

swinglock — Fri, 21 Aug 2026 17:52:17 +0000

https://ounapuu.ee/posts/2025/07/17/kagi/

New comment by swinglock in "Kagi added a setting for removing paywalled links from search results"

swinglock — Fri, 21 Aug 2026 16:03:03 +0000

Have they stopped sponsoring terrorism?

New comment by swinglock in "I close SSH port 22 (and what I use instead)"

swinglock — Sat, 15 Aug 2026 08:06:00 +0000

No, it doesn't. You're arguing against a straw man. The primary motivation was in second paragraph.

New comment by swinglock in "OpenNode – Bitcoin Payment Processor"

swinglock — Wed, 22 Jul 2026 16:31:28 +0000

Wire transfer is better?

New comment by swinglock in "An update on residential proxies and the scraper situation"

swinglock — Sun, 19 Jul 2026 18:20:56 +0000

If it takes seconds to compile it then the browser wins. The PoW only runs for seconds already.

New comment by swinglock in "Dumber Mini: A Nokia-Style Phone with WhatsApp and Maps"

swinglock — Sun, 19 Jul 2026 18:18:51 +0000

"Sorry, this item's currently unavailable in your location."

New comment by swinglock in "An update on residential proxies and the scraper situation"

swinglock — Sat, 11 Jul 2026 08:20:40 +0000

Can you not design a PoW that is most efficient in a browser? Don't brute force hashes like Hashcash/Bitcoin, do something similar to RandomX instead but in JS. Browsers ought to run the fastest JS interpreters already so if interpreting JS becomes the bulk of the work, that attack might not work. Maybe even involve the DOM or whatever else makes sense.

New comment by swinglock in "Resetting Xbox"

swinglock — Wed, 08 Jul 2026 17:17:27 +0000

It was more interesting when there were real tangible improvements. Game graphics has been like smartphones for a good long time now. Performance and feel (low latency and high consistency) still matters though, that is not solved, but it's also never marketed. Quality is not included with a UE license.

New comment by swinglock in "PostgreSQL and the OOM killer: Why we use strict memory overcommit"

swinglock — Fri, 03 Jul 2026 20:38:57 +0000

Windows does memory compression too.

New comment by swinglock in "PostgreSQL and the OOM killer: Why we use strict memory overcommit"

swinglock — Fri, 03 Jul 2026 16:15:45 +0000

I don't think it has overcommit at all, at least that's the default. That would be why you don't have Windows OOM killer stories.

New comment by swinglock in "Steam Machine launches today"

swinglock — Tue, 23 Jun 2026 12:23:22 +0000

Steam Support can probably get it sorted.

New comment by swinglock in "OpenBSD 7.9"

swinglock — Tue, 19 May 2026 18:12:05 +0000

AF_ALG does ring a bell.

New comment by swinglock in "CopyFail was not disclosed to distro developers?"

swinglock — Thu, 30 Apr 2026 20:24:27 +0000

Neither was NixOS.

https://discourse.nixos.org/t/is-nixos-affected-by-copy-fail...

New comment by swinglock in "The Day the Telnet Died"

swinglock — Wed, 11 Feb 2026 05:03:15 +0000

Because it's there.

New comment by swinglock in "Nvidia-smi hangs indefinitely after ~66 days"

swinglock — Sun, 25 Jan 2026 10:09:45 +0000

Something breaking after 49.7 days is a classic. Someone counted milliseconds since start with a 32 bit unsigned int and some code assumed it couldn't wrap.

New comment by swinglock in "A tab hoarder's journey to sanity"

swinglock — Tue, 20 Jan 2026 20:36:45 +0000

There are self-hosted options too. I can't vouch for any of them but many look competent.

New comment by swinglock in "Wine 11.0"

swinglock — Sun, 18 Jan 2026 23:41:54 +0000

Win32 and Wine being a lightweight alternative to HTML and Electrum is a fun idea.

New comment by swinglock in "A tab hoarder's journey to sanity"

swinglock — Wed, 07 Jan 2026 18:35:31 +0000

This is why I used ReadItLater and now Instapaper. It's integrated in my ebook reader too.

New comment by swinglock in "Linux kernel security work"

swinglock — Sat, 03 Jan 2026 23:33:52 +0000

Most users have JS enabled nowadays. Much of the web doesn't work without it. It was just an example.

New comment by swinglock in "Linux kernel security work"

swinglock — Sat, 03 Jan 2026 16:17:40 +0000

No it does not. I've had this happen in legacy systems myself. The checksums of TCP/IP are weak and will let random errors through to L7 if there are enough of them. It's not even CRC and you must bring your own verification if it's critical for your application that the data is correct. TLS does that and more, protecting not only against random corruption but also active attackers. The checks you get for free are to be seen only as an optimization, letting most but not all errors be discarded quick and easy. Just use TLS.