bpye

If you have a separate WiFi AP running your own router on your *nix of choice is quite achievable. I use NixOS since I use it everywhere else, if I didn't I'd probably go with OpenBSD.

They'd also need to MPTCP right?

Out of interest - what models are you using for this? GLM?

From experience OpenAI and Anthropic have pretty heavy safeguards in recent (public) models.

We got very close in XZ Utils.

https://en.wikipedia.org/wiki/XZ_Utils_backdoor

> With the assumption that any code base scanned in its entirety by the latest frontier cybersecurity focused models is secure...

The models that many organisations and most individuals don't have access to?

I like some of the 2-in-1 devices that come with a pen - Surface Pro type devices or ones with a 360 degree hinge.

But with a conventional laptop I've only ever accidentally used the touch screen.

R/w access to your production database is still a problem, though yes less bad than world read/write.

And also relies on the target being patched against any Linux EoP...

I don't think a PoC was publicly available for Heartbleed before patches? I may be misremembering.

> You can tell with some of the more recent models that they have been fine-tuned in such a way to produce a lot of descriptions that sound impressive but have no real meaning. In terms of production of semantic text, the recent models are significantly worse than a year ago.

I frequently find this in LLM generated comments, both when using an LLM myself or when reviewing code that someone else used an LLM for. I find it is often the most obvious indication that nobody really read through what was generated.

That's device specific however. Pixels, yes, but say a Samsung device? Not any more - they nixed bootloader unlocking.

Until iOS lets me run a web browser that isn't WebKit, without geographic restrictions, I'm not interested. It's an absurd restriction.

Being able to have a fully functional version of KDE Connect was also be a requirement.

They're still using Xe based iGPUs, so 3D still matters.

Would save states suffice? Because that already exists.

For video decode if it supports the codec you want you're set, but video encode has definitely improved on more modern GPUs. Older GPUs work but you'll likely get worse quality for a given bitrate.

The 9070 (XT) is great on Linux if you're happy to give up RT. It works but there is a significant perf gap vs Windows.

I do like my Tangara, though I use an ancient Sansa Clip more just because it's so much smaller.

I've been using KDE for the last few years, it's been the most familiar experience.

CAD and video games are the areas where I struggle to ditch Windows. Many games work well under Proton and I have had luck with running Windows with an assigned GPU - but that does sometimes break for seemingly no reason.

I am quite happy to give Mubi my money each month, even if I don't use it a _huge_ amount. I have seen some great films on the platform, and I really do want to see them be successful.

Some content has overlap with conventional streaming services, but then films like "Do not expect much from the end of the world" probably wouldn't see a streaming release elsewhere?

I'm very interested in Valve's approach, shifting some of the weight to the back.

> Can't imagine how restrictive it might be on 70 x 66.

Ouch, that's a tough sell. Maybe useful for AR type stuff, but I wouldn't want to go from a Quest/Index/Frame/etc to that for VR.

IOMMU branding at least makes a little bit of sense. AMDs and Intels virtualisation ISA are completely different, including their IOMMU.

> sufficient tricks (that Intel GPUs lack)

Isn't it just being able to shift the window of GPU memory visible to the CPU?

I patched my old Haswell system to add an NVMe DXE so I could boot from a modern SSD. I currently have a Lets Note laptop that definitely has UEFI, but pretty much pretends to be BIOS - apparently it's possible to modify to expose UEFI features but I have yet to try...

Blackberry might not be, but RIM are still around, QNX is still in use.

Nvidia's kernel module is open source these days, it's the usermode stack that's still closed.

> Conversion of legacy nontrivial C++ code bases into Rust (or anything else for that matter) feels like it should be one of the "Millenium problems" for AGI.

Whilst I have no doubt that LLMs will be useful here, I still have reservations about validation. I think experience tells us that test coverage is generally insufficient to ensure functional equivalence, and not all components are well specified.

Are there DP-HDMI adapters that work with HD audio bitstreaming and VRR? Different use case, but I've not had much luck finding good options.

I pay just for individual with its 60GB, it really hasn't been an issue.

What do you use on Android instead of Google Maps, OsmAnd or something else?