So I’m talking about playing previously Windows-only games on Linux, e.g. via proton.
I don’t know about the libraries etc that are used - is it possible for Microsoft to use some legal voodoo, for example, to suddenly end it all, and make the use of their libraries illegal (if they belong to Microsoft in the first place)?
Or could there be other ways of interference?
Proton is built on top of wine for windows compatibility. The wine project has been very careful to independent build its compatible versions of libraries. There should be no Microsoft code in wine.
I think should is the biggest risk here. With the source code leaks for Windows XP and others, I imagine it’d be quite tempting to reuse some of Microsoft’s code for the more obscure API calls that aren’t implemented yet. The Wine project itself does its very best to avoid doing that, but one lying contributor can throw a wrench into the works.
The people behind Wine are quite vigilant, so I don’t think Microsoft will find any of its (closed source) code in the project.
Microsoft knows that if they start tampering with that they will get into all kind of shit antitrust wise. Proton is a pretty small project from their perspective, so it’s really not worth the risk and/or public backlash.
These days, for sure, but back in the day Microsoft tried to sue Linux into non-existence. You never know if Ballmer ever returns and declares war on Linux again.
Sure, but Microsoft has since contributed a lot to Linux and other open source projects. That’s not me saying “oh they’ve changed!”, that’s me saying they’ve made it significantly harder on themselves to bring legal action against because they’ve publicly endorsed and supported the project for so long.
Whatever legal arguments they tried in the past that failed are even weaker now.
There are techniques to insulate the codebase. For example, you can have one person read the actual leaked code, explain the data structures and algorithms at a high level to a developer, then have the developer implement that logic themselves based only on what they understood from the explanation. I believe this is known as clean-room reverse engineering.
Even then you can still have someone read the source and write a spec for a second programmer to write a library. The programmer never saw the source code but it was still useful. Still legal to do this. If someone dumped original source into the projector could be similarly checked for duplication without breaking the law.
That’s true, but white room reverse engineering requires two people to do what one person could do by just stealing the code. Plus, the person reading the source code would be “burned”, they can’t work on normal implementation after browsing the source.
That’s what LLMs are for