The "steam box stuff" led directly to the "cool hardware stuff". The Steam Box never went away. The software and ecosystem was worked on continuously for many years. (I worked on it for a few years, early on.) Now it just has *real hardware*.
At the time I started on it (2012-2014 timeframe), the x86 hardware to pull it off right just didn't exist. We looked everywhere. We talked intensely with AMD, NVidia, Intel, etc. - everybody. Times have changed.
The CEO of NVidia and GabeN had a private meeting, hands were shaken and all the sudden we had resources at our disposal that weren't available before. So it's not just Valve who is behind this effort - it's way bigger than that.
At the time, I was partially motivated to give the game industry a backup in case Microsoft execs decided to clamp down. They visited Valve and threatened to do this by squeezing Steam out. Looking back, BillG's Microsoft is ancient history so this worry seems totally silly now.
At this point I wouldn't be surprised at all if Microsoft open source developers contributed to Proton/WINE, the Linux kernel, or key Linux libraries. Everything has changed.
You don't really think GabeN would announce Steam Box and it would just die on the vine and disappear entirely? Imagine how embarrassing that would be. No, this is a very long term effort.
I enjoy it when most others who don't actually understand the PC or video game business laugh at the effort. They have no idea - their mental models are too small.
This goes way deeper than even video games. Some extremely senior ex-MS developers, the types who go home to large estates, contributed to the project in a coding capacity. They didn't trust Windows at all - they feared it.
I found this super ironic and worrying, that some of the key architects of the core parts of Windows itself didn't trust it.
It's a miracle that Gabe was able to set aside some developers inside of Valve, and give them the bandwidth to finish it. Without him it wouldn't have went anywhere.
The lack of trust in Windows seemed to be related to very serious security concerns. They didn't believe Windows was secure or even ever truly securable. You couldn't for example ever do banking on it, or anything truly private. Only an open source OS (Linux) was viable.
If you look at the data, the top installed/most used apps are historically video games. If what most customers are actually using runs outside of Windows entirely, there is much less of a reason or need to use Windows at all. Deprecating Windows because viable.
Either Microsoft can respond by ramping up a new version of Direct3D/DirectX (eventually forcing another round of complex and time consuming Proton updates), or they can start contributing to Linux and Proton and adjust to the new reality.
The "lets ramp D3D to break Proton/WINE" thing doesn't scale well. I would just spend these resources putting Microsoft developers on Proton/Wine or Linux itself. I like Windows for the apps, but Windows itself is late 80's tech.
If Microsoft doesn't adjust an entire large software ecosystem that they don't understand, control or influence at all will be built around them. The smart choice is to dump developers into Linux and WINE/Proton.
It would be interesting to see what BillG would decide in this scenario. Would he force through a D3D update to break Proton? Would he try to move away everybody from Win32 into something else to slow the WINE guys down? All the wrong paths for the 2020's.
Now, you can sit back inside of Microsoft and think this is all BS. The very engineers who built Windows and Microsoft, and used your API's for years (some decades) started and built this project - for these reasons. You can adapt or be made effectively obsolete.
A lot of thought - years of it, has gone into this. I have my popcorn out and I'm eager to see what happens.
BillG isn't there now, so I think they'll adapt or try to live in both camps until it's obvious that Windows is hopelessly unsafe and dead tech.
If the public knew what the architects know, they would be terrified of using Windows for personal computing.
Generated by Thread Navigator
Press ⌘ + S to quick-export