Very few are these days with the enormous amount of resources Valve has been pouring into Proton. The only truly Windows-only games I have encountered were online multiplayer with anticheat and that was almost always because the developer chose not to support Linux (e.g. EAC, Rust) and not for lack of support for Linux.
Perhaps that wouldn't happen if more people switched...
This hasn't been the case for at least half a decade. Most games run perfectly fine if not better on Linux. The only exceptions are purposely broken remote rootkit "anticheat" gsmes which no sane person would want on their system anyway.
With the exception of multiplayer games with invasive anti-cheat measures, a very large number of games works very well on Linux now with Wine/Proton. When using Steam this mostly works just out of the box without fiddling.
I guess for that you could have a Windows machine just for gaming. No need to deal with the trade-offs of that come with using a console (if that matters to you).
Not having a windows machine is one of the benefits of not having a windows machine! Using a console for gaming is cheaper, more secure, and more reliable then using a windows machine. The only downside is i don't have access to the wide ocean of steam and gog releases, but it's a fair trade as far as I'm concerned.
For semis this is irrelevant, as the GVWR is pretty much the same so EV mostly means you get limited on payload weight: in the US the feds increase GVWR limit by 2000 lbs (from 80k to 82k) for EV semis, but not every state has legalised that, and it's barely noticeable (EU has a much more aggressive bye, from 40 to 44 tonnes).
The inconvenient truth might be that the other models score higher than OLMO because they aren't restricted to purely "open and accessible" training data. Who knows what private or ethically dubious data went into training Mistral or llama, for example.
Exactly. If we really wanted to benchmark the various models on the merits of their individual implementations, we should be comparing them all on the same open dataset.
Yeah. Reasoning models like r1 tend to be good for architecting changes and less optimal for actually writing code. So this allows the best of both worlds.
Apple's marketing team must be so good if they can convince HN users that an iPad is a suitable replacement for a "real" computer. No one who knows what a terminal or root access even means should be fooled by this.
reply