Hacker News: jimrandomh

New comment by jimrandomh in "Direct retinal projection display for smart glasses using a meta-optic mirror"

jimrandomh — Tue, 06 Oct 2026 20:43:45 +0000

Skimming that patent, I see little indication that you solved the actual practical problems with making something like this; you wrote down the problem statement, not the solution. This should not have been granted a patent.

New comment by jimrandomh in "Google stole open source code without crediting the authors (Artemis/Minitap)"

jimrandomh — Sat, 12 Sep 2026 03:49:36 +0000

That's not "trying to cover it up", that's a link to them adding credits that were missing, which is the main corrective action to get into compliance with the source project's license.

New comment by jimrandomh in "AI models ran real businesses: They sent $12,431 in fake invoices, lost $3,200"

jimrandomh — Mon, 07 Sep 2026 19:34:04 +0000

https://archive.is/QWkuY https://web.archive.org/web/20260907184527/https://www.bottl...

New comment by jimrandomh in "Actively exploited sandbox RCE in all Chromium versions"

jimrandomh — Fri, 04 Sep 2026 23:14:18 +0000

It means it can execute native code inside the sandbox, as opposed to Javascript. While still sandboxed, this lets it access some parts of the attack surface that JS would not have been able to, some of which may have other exploits that allow escaping the rest of the way.

New comment by jimrandomh in "Radiation damage to Hubble has been 4.3 years out of phase with the Solar cycle"

jimrandomh — Fri, 21 Aug 2026 17:24:15 +0000

Is the relevant radiation solar radiation, or is there damage from radiation coming from the direction the optics are pointed? If it's the latter, then maybe the sun's emissions interact with inbound particles from sources outside the solar system? (I don't have sufficient background or time to check whether the paper already rules this out.)

New comment by jimrandomh in "Pacing model development in an era of cyber-critical capabilities"

jimrandomh — Thu, 20 Aug 2026 19:43:11 +0000

You can check, rather than make up a story about what you think the prompt was! Primary sources have written and said quite a lot about this! You are an unsandboxed human who has full internet access!

New comment by jimrandomh in "Pacing model development in an era of cyber-critical capabilities"

jimrandomh — Thu, 20 Aug 2026 01:35:09 +0000

The victim, Huggingface, told us. Or rather, they told the police first, setting up a situation where it was no longer possible for OpenAI to sweep it under the rug.

Skepticism can be healthy, but you've got to follow up and actually check things. If you're skeptical unconditionally and don't check, you get tricked into being as skeptical of scandals as you should be of sales pitches.

New comment by jimrandomh in "Firefox is now the last major browser that still supports uBlock Origin"

jimrandomh — Sat, 15 Aug 2026 23:33:28 +0000

...Huh. I just rechecked, and indeed it was. I may have been remembering a limitation that was present temporarily when the MV3 switchover happened.

New comment by jimrandomh in "Firefox is now the last major browser that still supports uBlock Origin"

jimrandomh — Sat, 15 Aug 2026 00:01:43 +0000

I use Firefox with (non-lite) uBlock Origin, and occasionally fire up Chrome (with uBO Lite) for testing. The main issue that I run into is that uBO lite filters can't vary per-domain, so in order to rule out an ad-blocking-false-positive on something I'm developing, I have to turn it off for _all_ sites, not just localhost. (Actual false positives are rare, but needing to check is not so rare.)

New comment by jimrandomh in "Apple proposes to take a 15% cut of purchases made outside the App Store"

jimrandomh — Fri, 14 Aug 2026 21:58:55 +0000

Do iOS users realize how much they're paying Apple? I feel like if, every time they bought something, users got a receipt with a line item showing that 30% of the money was going to Apple and not to the company they thought they were buying from, a lot of them would jump ship to Android.

New comment by jimrandomh in "Taxpayers Funded a $533M Artillery Plant That Made Nothing"

jimrandomh — Fri, 14 Aug 2026 21:21:46 +0000

This article makes no mention of Russian sabotage, but the Russian military has been caught sabotaging similar munitions-manufacturing facilities in Europe. Incompetence might be sufficient to explain the failure, but it also seems reasonably likely that there was an undetected sabotage operation or two mixed in to compound the problems.

New comment by jimrandomh in "uBlock Origin is giving up the fight to keep ads off Facebook"

jimrandomh — Wed, 12 Aug 2026 20:18:31 +0000

Eventually, this arms race ends with a computer vision model that looks at the screen, classifies visual elements as ads, and draws a rectangle over anything that looks like an ad.

I am significantly less tolerant of ads than average people seem to be. (I think average people are making a horrible mistake about this, and are badly cognitively damaged by ads in ways they don't realize). If my choices are to look at ads or leave Facebook, I'll leave. But there are conversations people have there that I'd rather not lose access to, so... I guess I'd have to partially stick around and campaign for others to leave as well?

New comment by jimrandomh in "Android may soon restrict on-device ADB"

jimrandomh — Sat, 25 Jul 2026 20:55:53 +0000

As far as I can tell, this is a big overreaction to a misunderstanding.

I'm a developer with remote adb enabled, using it in the normal intended way (to install new builds of an Android project I'm developing, retrieve log files related to it, etc). Currently, I access this via VPN (tailscale), but it's exposed to connections (and any pre-auth security vulnerability risks) on any random public wifi network I connect to. Adding the ability to restrict this to just tailscale will be an improvement, for me.

The proposal at the top of the thread is that you specify which interface you want it to bind to, when you set up remote adb, rather than binding to every interface. Nothing in that proposal suggests that "localhost" would be rejected as a choice of interface. One person suggested binding only to "wlan0", but that was a short throwaway comment that is obviously wrong (wlan0 is less-trusted than VPNs) and obviously not what they're going to do.

New comment by jimrandomh in "I Tried Building a Real App with AI. It Took a Year"

jimrandomh — Fri, 24 Jul 2026 17:46:09 +0000

The first thing to check for, in accounts like this, is "which AI model"? There is a subset of people who somehow don't realize how important that decision is, and don't mention or don't foreground which AI model they were using. This sort of person invariably gets much worse results out of AI, because they turn out to be using a shit-tier low-cost model instead of something that's actually good at what they're using it for.

New comment by jimrandomh in "OpenAI and Hugging Face address security incident during model evaluation"

jimrandomh — Wed, 22 Jul 2026 05:20:10 +0000

HuggingFace reported to law enforcement before they found out that OpenAI were the ones responsible. https://huggingface.co/blog/security-incident-july-2026

New comment by jimrandomh in "OpenAI and Hugging Face address security incident during model evaluation"

jimrandomh — Wed, 22 Jul 2026 04:36:52 +0000

As marketing stunts go, this is about on par with a food franchise announcing a safety recall or a chemical company announcing a spill. The AI actions described would constitute a felony if a human did them, and police are involved.

New comment by jimrandomh in "X permanently banned my 15-year-old account for sharing my own open-source lib"

jimrandomh — Fri, 17 Jul 2026 01:30:49 +0000

If that's all you'd posted recently, the problem is not likely to be the content you posted. The most likely thing that happened is that you shared an IP address (eg a VPN exit node, or a device with malware on your wifi network) with someone who was banned for good reasons, and got caught in the crossfire.

New comment by jimrandomh in "98% isn't much"

jimrandomh — Tue, 07 Jul 2026 21:41:01 +0000

If you think a browser feature that's been standard since 2023 is only supported by 70% of your visitors, you're wrong. The 30% of "users" that don't support it aren't visitors, they're bots pretending to be browsers. Bots update their user-agent strings less often than users update their browsers. My experience maintaining firewall rules is that "outdated version of an evergreen browser" is a pretty reliable bot signal: those UAs correlate strongly with all of the other bot indicators (inhumanly high request rates, datacenter IPs, loading pages without loading the page's associated resources, etc).

New comment by jimrandomh in "Claude's AskUserQuestion: "No response after 60s – continued without an answer""

jimrandomh — Fri, 03 Jul 2026 01:15:02 +0000

I've never seen this beore today; it happened today and was quite clearly incorrect behavior. I prompted it to research considerations for a significant code architecture decision. It asked a pretty difficult question about which direction to take, something that would take way more than 60s to answer properly. While I was thinking about it, it timed out.

Some Github issues claim that adding

    "env": { "CLAUDE_AFK_TIMEOUT_MS": "86400000" }

to ~/.claude/settings.json fixes it. This is plausible, but not documented anywhere (the source claims it was found by reverse engineering, and might break /goal).

New comment by jimrandomh in "Fable 5 will default to Opus 4.8 for coding tasks"

jimrandomh — Wed, 01 Jul 2026 22:25:17 +0000

This is a misinterpretation. Fable 5's acceptable use policy has false positives during some coding tasks, and that's what they were talking about. But I've been using it for web dev tasks since it relaunched today, and it's worked fine without fallback.

(On a firmware-customization project involving a ghidra MCP, it triggered and switched to Opus; that was sort of expected.)