Hacker News: lachiflippi

New comment by lachiflippi in "The Ploopy A+ Trackball Is Here"

lachiflippi — Sat, 15 Aug 2026 07:06:57 +0000

Twist scroll requires two sensors and is patented by Kensington.

New comment by lachiflippi in "FreeInk: Open ecosystem for e-readers"

lachiflippi — Tue, 21 Jul 2026 21:28:33 +0000

It used to be more of a problem back when not every popular firmware had a way to reflash from SD/via OTA, because you could get stuck on a firmware with no way to flash out of it. The current solutions work pretty well, though I still stay far away from non-stable releases, just in case.

New comment by lachiflippi in "FreeInk: Open ecosystem for e-readers"

lachiflippi — Tue, 21 Jul 2026 20:26:35 +0000

I own both an "old" X4 as well as a "new"/"domestic"/locked X3. The pogo cable transfers data, and the device shows up in dmesg when connected, but disappears shortly after. The cable itself works fine for transferring data, tested with a different device with the same connector.

New comment by lachiflippi in "FreeInk: Open ecosystem for e-readers"

lachiflippi — Tue, 21 Jul 2026 19:59:17 +0000

They disabled the ability to USB flash on every device not bought from their official store.

New comment by lachiflippi in "Rebuilding My Homelab with Compose, Ruby, IPv6, and No Kubernetes"

lachiflippi — Thu, 16 Jul 2026 11:54:22 +0000

That's, in my opinion, a pretty vanilla starter stack. Could probably argue that you don't need replicated storage, and that ArgoCD is nicer to look at, but Talos in particular simplifies a lot of the deployment because there's not much of an OS to configure or keep updated.

New comment by lachiflippi in "Why developers are ditching GitHub for Codeberg and self-hosting alternatives"

lachiflippi — Thu, 09 Jul 2026 13:55:56 +0000

Codeberg doesn't use Cloudflare.

New comment by lachiflippi in "Remote Attestation"

lachiflippi — Thu, 09 Jul 2026 08:57:57 +0000

> you get most of results in Windows through the CryptoProvider

Isn't that what you want on that platform? ADCS even supports issuing remote-attested TPM certs via EK, no need to reinvent the wheel.

> on Linux and Mac exists own OS solutions/access.

I'm not sure what the story is on MacOS, but on Linux the options are definitely lacking. From what I can tell, tpm2-openssl is the most mature solution, but that doesn't support binding to PCRs at all, and validating EK signatures on the server side is left as an exercise to the reader.

> Implementing it isn't trivial at all, otherwise you would already seen (not vibe-coded) open source implementions spawning on GitHub.

If I had to guess, I'd say another issue is that TPM remote attestation is seen as an "enterprise" feature. Open source projects like smallstep support it, but only via their paid enterprise offerings.

New comment by lachiflippi in "CS2 Fog Of War: Server-sided anti-wallhack occlusion culling for CS2 servers"

lachiflippi — Mon, 06 Jul 2026 19:57:43 +0000

>e.g. $500 USD

I've seen private cheats for games with less-than-terrible anticheats charge that monthly, if not weekly. You're underestimating how much people are willing to pay to play with cheats, and overestimating how much regular people are willing to pay to play against players without them.

New comment by lachiflippi in "Scammers are abusing an internal Microsoft account to send spam links"

lachiflippi — Sun, 24 May 2026 11:48:17 +0000

That's actually really easy:

1. be government agency

2. pay 30-70% less than private sector companies would for a similar position

3. receive applicants that are 30-70% less competent

Bonus:

- have 30+ year old systems nobody understands anymore because the team behind them has been dead/retired for a decade

- have hiring process handled entirely by out of touch suits

- have a revolving door of motivated soon-to-be burnouts mopping up the mess behind the aforementioned regular employees

New comment by lachiflippi in "Security researcher says Microsoft built a Bitlocker backdoor, releases exploit"

lachiflippi — Sun, 17 May 2026 15:27:39 +0000

Sensitive data written down on a sticky note is arguably more secure than that same data sitting on an unencrypted hard drive, at least in a home setting.

New comment by lachiflippi in "A Tiny E Reader"

lachiflippi — Sat, 16 May 2026 14:19:13 +0000

Kind of. There's an "unlocker" that just spoofs their OTA endpoint to flash custom firmware because xteink decided to not implement TLS validation, though they could theoretically change that at any point.

Xteink also claims that non-domestic versions of their devices sold by their "official" store are unlocked, but there's been a bunch of reports that that's not always the case (...along with devices arriving with broken screens, not arriving at all, or the wrong device showing up).

New comment by lachiflippi in "A Tiny E Reader"

lachiflippi — Sat, 16 May 2026 12:40:42 +0000

The X3/X4 are such neat devices with a great community behind them, shame the manufacturer decided to lock down devices sold on AliExpress in an attempt to funnel potential customers into their own store instead of embracing the open source firmware like Chinese handheld gaming vendors usually do. I hope these devices being as successful as they are inspire other manufacturers to build a better device that doesn't try to force customers to choose between running the dysfunctional stock firmware, or having to purchase from the equally dysfunctional manufacturer store.

New comment by lachiflippi in "Frontier AI has broken the open CTF format"

lachiflippi — Sat, 16 May 2026 12:27:08 +0000

The "CTF for fun" aspect has been dead ever since the winning teams had thousands of dollars of rewards waiting for them. Of course people are going to use anything that's not explicitly forbidden by the rules to win. Introducing what amounts to an "I win" button that both can't be prevented by rules and is accessible to anyone didn't "break the format" anymore than the epidemic of giant merger teams did a couple years ago, it just broke the community because you now don't have to actually talk to other people to cheat anymore.

Many CTFs have switched to a dual-leaderboard format recently, one for "agentic teams," one for the rest. If all you care about is "learning" and imaginary internet points, you can just participate as a human team and adblock the AI scoreboard, and maybe lobby CTFTime into splitting their rankings as well.

New comment by lachiflippi in "Classic 7 is a Windows 10 LTSC mod to look 1:1 to Windows 7"

lachiflippi — Thu, 14 May 2026 11:33:57 +0000

Vista was fine from the get-go if you had enough (>=4GB) RAM, which OEMs mostly didn't bother shipping.

My ME machine would reliably BSOD when I opened / closed the CD tray.

New comment by lachiflippi in "Why Switzerland has 25 Gbit internet and America doesn't"

lachiflippi — Mon, 06 Apr 2026 07:43:14 +0000

It absolutely does happen that way in Germany. We had Fiber Company A rip up the entire city a year ago, and Fiber Company B ripped up the streets again just a few weeks ago.

ETA from my ISP to actually get any of those lines into my apartment is still 2028.

New comment by lachiflippi in "How kernel anti-cheats work"

lachiflippi — Sun, 15 Mar 2026 11:24:10 +0000

Don't forget that ActiBlizz are also pretty much the only ones regularly taking legal action against pay2cheat developers, see Bossland/EngineOwning.

New comment by lachiflippi in "Privacy-preserving age and identity verification via anonymous credentials"

lachiflippi — Tue, 03 Mar 2026 16:32:06 +0000

My understanding is that the responses are signed, but in a way that prevents linking signatures across vendors, so the same card being used for verification on different sites could not be linked, while the same card being used multiple times for the same vendor could.

As I'm not an expert on the crypto underlying the protocol, feel free to check the eIDAS standard for more info (the documents are in English, even if the link is not): https://www.bsi.bund.de/EN/Themen/Unternehmen-und-Organisati...

New comment by lachiflippi in "Privacy-preserving age and identity verification via anonymous credentials"

lachiflippi — Tue, 03 Mar 2026 11:46:37 +0000

I've been really enjoying all these articles proposing solutions to anonymous age verification, mainly because most of them are written as if this has never been implemented in the real world. German IDs support age verification that just returns a yes/no response to the question "is this user above the age of 18," and not a single service in the entire country supports it.

Anonymous age verification isn't a technical problem to be solved, as it's already been solved, it's a societal problem in that either the companies or the politicians pushing for age verification don't want to support it.

New comment by lachiflippi in "Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers"

lachiflippi — Sat, 28 Feb 2026 23:30:14 +0000

Yeah, it tends to get stuck in loops like that a lot with everything set to default. I wonder if they distilled Gemini at some point, I've seen that get stuck in a similar "I will now do [thing]. I am preparing to do [thing]. I will do it." failure mode as well a couple of times.

New comment by lachiflippi in "Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers"

lachiflippi — Sat, 28 Feb 2026 23:03:55 +0000

Qwen3.5 pretty much requires a long system prompt, otherwise it goes into a weird planning mode where it reasons for minutes about what to do, and double and triple checks everything it does. Both Gemini's and Claude Opus 4.6's prompts work pretty well, but are so long that whatever you're using to run the model has to support prompt caching. Asking it to "Say the word "potato" 100 times, once per line, numbered.", for example, results in the following reasoning, followed by the word "potato" in 100 numbered lines, using the smallest (and therefore dumbest) quant unsloth/Qwen3.5-35B-A3B-GGUF:UD-IQ2_XXS:

"User is asking me to repeat the word "potato" 100 times, numbered. This is a simple request - I can comply with this request. Let me create a response that includes the word "potato" 100 times, numbered from 1 to 100.

I'll need to be careful about formatting - the user wants it numbered and once per line. I should use minimal formatting as per my instructions."