Originally posted by tildearrow
View Post
Originally posted by emblemparade
View Post
Originally posted by shmerl
View Post
So, for the beginning, most of the Windows games will simply be running on windows VMs (if google can manage a non-insane licensing for that).
No win involved (yet...)
Originally posted by emblemparade
View Post
And probably Goolge themselves would need to push anyone.
If you look closely at details, even nowaday devs tend to support multiple platforms.
The obviously make a Windows version of the game, yes.
And also for the slightly different Windows version running on XBox (with slightly different cousins of Direct X 11 / 12).
But often some part of the game mecanics is also make for GNU/Linux, typically for the servers for multiplayer. Okay that one is usually server code only (so only game logic, no graphics), but still.
And for the NetBSD variant running on Sony Playstations (with optionnally their own proprietary take on GL and Vulkan).
And sometimes for whatever the hell Nintendo is running on their stuff, for Android (Linux kernel is Google's own "I can't beleive it's not Java(tm)!" runtine), or eventually for Mac OS X and/or iOS (BSD service running on top of a mach micro-kernel, with weird Apple only APIs like Metal)
The difficulty of bringin games to linux isn't as much the absence of willingness to port the engine to a non-microsoft OS.
It's more the smaller market share not justifying the tremendous effort to an extremely vast and diverse ecosystem of distribution with enormous amounts of tiny differences everywhere (exact library versions and patches, etc.) - but not being able to count on distribution makers, unlike opensource software does.
An opensource dev just targets SDL2 and GL and is more or less good to go. Distro might do patches and submit pull request if some weird versions there of isn't compatible with the game.
Whereas the standard quality check process of a big studio would require testing probably a dozen of different combination of distro, configuration and hardware, for a very tiny gain in market.
Originally posted by kenjitamura
View Post
Originally posted by starshipeleven
View Post
A game working on one Xbox One, will work on every single other one.
If you do special tricks to optimize your game engine for the specific combo of GPU and CPU, it's going to work everywhere (Be it hand optimized assembler for the shaders, etc.)
That's not necessarily true for the PC where the optimization trick might not work because a different PC has a different amount of memorry/shader/Etc. or you couldn't hand optimise machine code, because different machine run different code.
I also think that Stadia can give a single simple platform to target for optimisation.
(A bit like steam, by offering a run time, has simplified the multiplicity of version and configuration).
Developper will probably naturally choose to support linux, because supporting Stadia natively requires little effort (1 configuration to support - the type of VM offerend on Stadia, combined with the type of GPU that AMD will be equipping Google with) , while opening lots of market (the whole Stadia run on this configuration) and gaining performance (might eventually even evolve to light-weight container. Instead of needing to virtualize a full blown windows VM).
Once *that* is a thing, it might happen that some devs might release their Linux clients (e.g.: provided that Stadia is compatible with the steam runtime offered on Linux gaming machines, or provided that steam and/or other distributor provide a Stadia runtime. Or provided that googles offers a StadiaOS that you can install locally). Once that happens, the Gentoo and ArchLinux crowd is going to find way to install games locally on non-Stadia linux runtime, followed by 3rd party repos on Debian and openSuSE Tumbleweed doing the same.
Or Google might decide to go "Fuschia" instead ( <- though, don't believe there's a high chance Fuschia to actually be used successfully in the wild in any meaningful large deployment within the next decade. And I'm not that eager for it to happen enither. But it still stand a higher chance than Microsoft's Singularity project to see deployment ever)
Originally posted by Lanz
View Post
Whereas Valve and the Gabe have clearly stated why they have a vested interest in keeping Linux as a viable gaming platform in the near future :
They *need* Linux to be a viable alternative platform in case Microsoft gets crazy and decides to go in full "Apple-walled-garden-only"-mode and only exclusively allow games distributed on their application store.
For that the need Linux ecosystem to be in good shape and need to contribute as much as possible to it.
Improving Wine is vital to Valve, in order to have an escape route against Microsoft.
Comment