nnevatie

> An LLM is called

Hooray for incidental non-determinism.

> these models are useless on any real world task

I beg to differ. They are not perfect but immensively useful today.

Yeah, no. I definitely wouldn't consider this a solid conclusion. The attempts pasted to the article look...pretty tame.

Because you'd be missing the market and monetization layer that Steam so conveniently provides.

It's a meme.

Indeed. Even Qt isn't native, in the most purist sense.

Yes, naturally. However the PUBG AI feature is based on generative AI, at least partly: https://www.nvidia.com/en-us/geforce/news/pubg-ally-ai-teamm...

They do, to a degree. See e.g. PUBG's latest AI teammate addition.

At this price, ain't two people going to make eye contact, ever.

As explained by a LLM.

Ahh, I should have probably added some context around my hyperbole. I was referring to real-time computer vision - think of e.g. segmenting FHD/UHD video.

No one uses ONNXRuntime (nor the new engine in OpenCV 5) in production. For anything performance-sensitive, one would run models under TensorRT, as an example.

This is one way to rephrase "we don't want your AI slop, thanks.".

That's one way to look at it, certainly. There are several OK options in that space, e.g. Conan (2) and vcpkg.

The language is fine, mostly, nowadays.

The ecosystem isn't fine - just to get a project going requires picking a non-trivial set of tools and approaches, none of which the C++ standard enforces or guides to.

For example, will you manage dependencies via packages? If so, with what? What will you use for building your project? The list goes on and on.

Yeah, it doesn't look too bad on the surface. However, Qt is bad and I'd advice moving away from it if you can. This is based on ~20 years of experience using it in various projects.

The licensing is trying to paint a picture of LGPL being somehow uncertain/evil, the way Qt advocates non-idiomatic C++ practices - I could go on for days, but shall digress.

More visual examples are sorely needed - I could only find a small (toy) example of a dialog, that didn't seem to showcase many of the framework capabilities.

And here I thought I'm dealing with legacy setups!

Notice the demo video from Comanche also shows roll.

Edit: To support roll, the renderer essentially rendered the voxel terrain into a frame buffer and then applied camera transformations that gave the appearance of a fully rotating viewpoint. The terrain itself was not being raycast through a true 3D voxel volume.

> Visual SourceSafe

Ugh, instant flashbacks and not the good kind.

Reads like something that came out of an LLM.

I noticed the slop immediately and ejected it from by brain.

Yes, to a degree. For example, if you look into Eigen, the math library, you'll notice that it's mostly header-only and heavily templated. Writing all that by hand in C would be possible, but incredibly time-consuming unless you'd rely on some pretty incredible macro-magistry.

> WHAT THE FUCK DID YOU DO???

For me, this doesn't require using an AI agent/model, even. Just using Windows and watching it freeze its File Explorer for the nth time does it for me. How did we end up here were the software/OS stack is so shit it can barely be used for the most trivial things, is wildly beyond me.

Yes, this must be it. There's no experience like driving a manual with a two-plus ton vehicle.

> It creates an entirely new torque language controlled by the driver

Oh wow, sounds like some corporation BS if I ever read some. My EV works by pressing the gas pedal and the torque is right there - not sure what revolutionary new invention is required?

swoosh swoosh - that's the sound of things traversing backward in time.

The link doesn't work - could you link it again? I might be basing my above comment on and older version of Highway.

Yes, technically it is of course fine, just as a design I find use of thread_local is more of an afterthought than something I'd prefer.

In my view, thread_local is a bit of a code/design red flag. I didn't read the entire code in this case to see whether the thread_local use is warranted or not, though.