BSD0 doesnt

Does that page even say which RISC-V CPUs are being used that are slow? I couldn't see it, which seems a bit of pointless complaining.

> RISC-V builders have four or eight cores with 8, 16 or 32 GB of RAM (depending on a board).

Which boards are used specifically should not matter much. There's not much available.

Except for the Milk-V Pioneer, which has 64 cores and 128GB ram. But that's an older architecture and it's expensive.

Intriguing work! Does it panic on any bad inputs? That's better than memory unsafety of libxml2, but still a DoS concern for some servers.

No it's not, it's 6pm!

> For example, std::time::Instant is implemented on the GPU using a device timer

The code is running on the gpu there. It looks like remote calls are only for "IO", the compiled stdlib is generally running on gpu. (Going just from the post, haven't looked at any details)

Which is a generally valid implementation of IO. For instance on the Nintendo Wii, the support processor ran its own little microkernel OS and exposed an IO API that looked like a remote filesystem (including plan 9 esque network sockets as filesystem devices).

I'm surprised this article doesn't provide a bigger list of calls that run on the gpu and further examples of what needs some cpu interop.

Flip on the pedantic switch. We have std::fs, std::time, some of std::io, and std::net(!). While the `libc` calls go to the host, all the `std` code in-between runs on the GPU.

It looks like it is only applied for PTY sessions, which most computer-computer connections wouldn't be using.

https://github.com/openssh/openssh-portable/blob/d7950aca8ea...

Fs.com are a well regarded company. The OP should contact them and see if there's a recall on those cables or something. Sounds similar to what happened with some mains power cables sold by a hardware store in Australia.

They offer to provide you with personalized service and you block them? What do you do when you walk into a high end watch or clothing store?

FS provides account managers and they are very useful if you are working on a project as they can guide you and help you.

Is it too much to ask to just be able to buy things without being chased around by someone?

I have a Newegg business account (and maybe a few more for other pc stores, I'd have to check), but I literally never have them trying to get me to buy things. /shrug

fs.com sells primarily b2b. Account managers are normal and customary.

Just decline politely. They are not a bad company in that regard.

I have my issues with FS product itself, as it can be spotty sometimes in terms of compatibility and repeatability (e.g. getting the same optics firmware every time over the course of a couple years) but they typically handle exceptions and problems quite well. Via your account manager contact.

Weird reason to hate a company, imo. Would be far different if they continued to spam you with phone calls/e-mails after you declined the request.

Sounds like both parties dodged a bullet.

Right, but sometimes your account manager's job is to manage a giant RMA as a result of their manufacturing department f'ing up. So while I am greatful she did a good job, I would have preferred to not need an account manager at all...

People in the industry would know that QNX has been around since the 90s (or 80s?) as a very solid embedded GUI platform. They're a company that doesn't need to prove their credentials.

I'd agree using qnx.software rather than qnx.com is kinda dumb though.

I thought that QNX was acquired by / is still developed by BlackBerry?

Seems you are right. See how complicated it is? :)

Indeed. And it has happened before that well-known brands change owners and suddenly push new products with certain risks attached for the user. That's all I wanted to point out, and for QNX it would be a very easy way to build trust by having this kind of information (or imprint) on their website.

So it's an elaborate front by South Korea instead.

You're joking but the north korea intrusions are factually established and an ongoing security risk.

That seems worth suggesting to the acme working group mailing list, if it hasn't already been discussed there.

You'd get those benefits from traditional dependencies if you copy them in and never update. Is an AI dependency going to have the equivalent of "upstream fixes"?

Probably? LLMs will train on fixes, then if you run the code through the LLM again to fix it.