Hacker News: rcxdude

New comment by rcxdude in "Waymo in Toronto"

rcxdude — Sat, 17 Oct 2026 13:55:34 +0000

The average human driver is safe enough that a single anecdotal experience isn't a good measure of safety. Waymo has, from the start, had a very serious focus on making the system safe and reliable in a fully autonomous context, in comparison to a far more 'YOLO' approach from their competitors. Tesla, meanwhile, has been consistently trying to jazz up a driver assist approach into a fully autonomous system, and in general it just hasn't demonstrated that it's actually reliable enough there (despite many claims for years that the reason they require hands on the wheel is just for regulatory reasons, it's actually been quite important for safety. It's quite easy to lie with statistics here and Tesla is quite prone to it).

New comment by rcxdude in "Waymo in Toronto"

rcxdude — Sat, 17 Oct 2026 13:50:29 +0000

Waymo is a mostly seperate group of people from Google, to start with. What's more, on the whole I think the reason they've been winning the self-driving race is that they've actually taken the safety and reliability of the system very seriously, in comparison to a lot of their competitors. Waymo seems to be the only one that both has the necessary level of respect for the problem and the ability to execute on it.

New comment by rcxdude in "Waymo in Toronto"

rcxdude — Sat, 17 Oct 2026 13:44:24 +0000

Enshittification is generally an exploitation of market efficiencies in some form, be it lock-in effects or market dominance in an area with a high barrier to entry. It's an explicit goal of a lot of investment: pick one company in an area, invest so much money into it no-one else can compete, corner the market by offering a better product than anyone else, then once you own the market, turn up the price and turn down the quality to get as much money out of it before someone builds a competitor and your customers can leave.

New comment by rcxdude in "Waymo in Toronto"

rcxdude — Sat, 17 Oct 2026 13:31:58 +0000

I think the lack of any person involved is part of the appeal for a lot of people. The lower cost does also help, I think it will be interesting to see if the notional price advantage from the lack of driver will bear out over time, though.

New comment by rcxdude in "Meta confirms smart glasses were recording without being activated"

rcxdude — Sat, 17 Oct 2026 13:08:22 +0000

The fact that Meta is pushing this is just going to poison the whole product in a huge number of people's minds. I don't think there's any chance that Zuck or Meta are going to regain enough trust to avoid that. (Whatsapp being a good example: despite being far more verifiably private and trustable as a product than the glasses it's still tainted by who owns it now).

New comment by rcxdude in "US Rep Announces Congressional Legislation to Protect Video Game Ownership"

rcxdude — Sat, 17 Oct 2026 13:04:37 +0000

One general difficulty with implementing this is dealing with the case of a company going under. It's hard to compel performance out of a bankrupt organisation, so in a lot of cases it might still fall through. (And if this becomes a thing, you might see video game studios setting up companies to operate each game to exploit this).

New comment by rcxdude in "Anthropic assigned highest junk credit rating in June"

rcxdude — Sat, 17 Oct 2026 10:52:28 +0000

I think it's more "I know this is wrong myself, but also even the LLM can find the problems with it, that's how wrong it is"

New comment by rcxdude in "Anthropic assigned highest junk credit rating in June"

rcxdude — Sat, 17 Oct 2026 10:51:17 +0000

Generally didn't happen, despite the common perception. The bailouts saved the banks, they generally did not care to save the owners or executives of the banks.

New comment by rcxdude in "I never authorized an integration that scans my secrets"

rcxdude — Sat, 17 Oct 2026 10:31:31 +0000

Yeah, I think the situation is that if your password is already weak/known to a level that it should already be considered compromised, a malicious HIPB could surface that compromise a lot more reliably/quickly.

New comment by rcxdude in "AI-assisted PCB design for KiCad 10"

rcxdude — Sat, 17 Oct 2026 10:15:20 +0000

To me the routing is the fun part. It's the main part of PCB design where I can actually get into a flow state, because it stops being a bunch of part searches and starts becoming a neat self-contained and fairly verifiable problem. I've often suspected that the resistance to autorouters is half their UI kinda sucking and half engineers not wanting to cut out the fun part.

New comment by rcxdude in "Last rites for Gentoo's Chromium package"

rcxdude — Fri, 16 Oct 2026 10:28:11 +0000

Regardless it's still a PITA to maintain an ancient distro as a build environment and the alternatives are hacky. It might be the best of the linux libraries but it's still not good.

New comment by rcxdude in "Last rites for Gentoo's Chromium package"

rcxdude — Fri, 16 Oct 2026 10:26:53 +0000

Yeah, that's forwards-compatibility, which is different. Something built targeting a newer version of glibc is unlikely to work with an older one. (and gentoo will by default build against the latest you have installed, so yeah, downgrading has a good chance of breaking near everything).

New comment by rcxdude in "Last rites for Gentoo's Chromium package"

rcxdude — Fri, 16 Oct 2026 08:42:04 +0000

>glibc has an unstable ABI, making back-portability impossible

Of all the libraries to call out for this, glibc is not the one. It's basically the only library on linux that actually has a strong commitment to providing a stable ABI, with its own symbol versioning system to allow it. It is a bit of a PITA to compile for an older version of that ABI, but it's very possible to be compatible with glibc across a wide range of versions and distributions (I don't know why there isn't a flag to do it: the most widely supported option is compiling on a system with the oldest glibc you want to support, and there's some header hacks which claim to be able to avoid that).

It's everything else necessary to interact with the system (especially anything desktop/multimedia) that's wildly unstable.

New comment by rcxdude in "Last rites for Gentoo's Chromium package"

rcxdude — Fri, 16 Oct 2026 08:26:58 +0000

Chromium looks like a build system for a complex integrated project where the engineers have one priority: make the application work as best it can. You will quickly find that when you focus on that then making the build system convenient, especially convenient for distribution packagers with their myriad of own rules and modifications that they want to make on top of your system, is quite a big blocker to the primary goal. You can see similar things with applications like OBS, where they bundle modified libraries to fix problems and then distributions unbundle them and release broken packages, causing no end of grief for users and the developers.

New comment by rcxdude in "Passkeys are not a trap"

rcxdude — Thu, 15 Oct 2026 23:41:52 +0000

AFAIK passkeys are specifically designed to avoid this. Normal passkeys contain no information about the device or software that is storing them (each key is specific to one website), and even with attestation (which I hate), the attestation reveals only what kind of device you are using (the keys are shared across many devices, usually one key per design).

New comment by rcxdude in "Passkeys are not a trap"

rcxdude — Thu, 15 Oct 2026 23:32:58 +0000

Not encrypting by default is a ludicrously unstable position as far as trust 'equilibrium' goes. It would be an astounding feat of coordination to avoid needing it as the internet grow, probably one that would cost far more than https.

New comment by rcxdude in "Passkeys are not a trap"

rcxdude — Thu, 15 Oct 2026 23:29:20 +0000

Paypal's implementation is weird. They have it locked behind some other kind of check so it only seems to prompt to make a passkey sometimes and then it only ever asks for that passkey on that device.

(the UX flows around passkeys are horrendously inconsistent between implementations in general, but even still paypal is a bit of an outlier.)

New comment by rcxdude in "The Double Headed Stack Allocator"

rcxdude — Thu, 15 Oct 2026 08:59:51 +0000

Inconsistent frame times tend to feel worse than just lower frametimes, on the whole. There's a reason a lot of benchmarking has started highlighting the 1% lows.

New comment by rcxdude in "Problems with the Rust Language"

rcxdude — Wed, 14 Oct 2026 23:26:26 +0000

The 'woah, slow down' also just feels kinda... late? Rust is pretty mature by now, it's not really some new unproven thing that needs a bunch of work to be useful.

New comment by rcxdude in "Problems with the Rust Language"

rcxdude — Wed, 14 Oct 2026 23:22:52 +0000

Indeed, and there are multiple projects which provide ways for you to run a full source bootstrap of not just rust but the entire C and C++ toolchain:

https://guix.gnu.org/manual/1.5.0/en/html_node/Full_002dSour...

https://stagex.tools/

(I think nixOS also has some work in this area, but I'm not sure how complete it is)