It's an impressive technical achievement, no question, but we're bedeviled by elusive crashes that seem to be exposed by some nuance of the Proton (WINE) emulation layer (at least 2% of our game crashes and at least 25% of our launcher crashes which is disproportionate given that only half a percent of our players are using some version of WINE).
So I wonder about the metric for "playable" -- is a 4x greater chance of crashing when using your Steam Deck vs your PC still "playable"? And I wonder if players, who can't tell if a crash is related to emulation or a bug in the game, will blame the game developer when it isn't really fair to hold them accountable.
Depends on how much it's crashing in the first place. The difference between a 1/2500 event and a 1/10000 event is probably not going to be noticed, but the difference between a 1/2 and a 1/8 event probably is.
> but we're bedeviled by elusive crashes that seem to be exposed by some nuance of the Proton (WINE) emulation layer (at least 2% of our game crashes and at least 25% of our launcher crashes which is disproportionate given that only half a percent of our players are using some version of WINE).
Do Windows games work on Linux? No, there's little Valve can do to change that, bless their hearts. But it's a very Reddit Social Media friendly statistic. It doesn't square with the reality from people like you, who actually make games with an audience.
And anyway, if you want to build for Linux, use Unity, so Proton is kind of a moot point, and also, you'll be doing that to sell on Switch anyway, porting your game like Vampire Survivor, Slay the Spire, etc. did, if you care about that audience. Vulkan will be fully deprecated before Proton can substantively reimplement enough Windows for non-middleware games. And anyway, Vulkan is awful, it's poorly supported by all the vendors, including Valve's own, but again, it doesn't matter.
The bigger picture: Steam already sells so many copies of games to so many people who spend 0 minutes playing those games. The idea of playing a game is enough to sell it. So in a sense, does it matter if your game doesn't work for the people who bought it? Like in the sense of objective reality?
RGB? People buy that stuff. It wouldn't be on everything if it didn't sell.
The Steam Deck is a cozy idea of gaming people like to buy. They could also blow $300 on a video card overpowered for playing Overwatch and League. These trends don't make sense in an objective reality, but the product is already psychological for so many people.
Brad Wardell, pariah and guy who was, ultimately, totally utterly right, figured this out ages ago:
> You focus on people who buy your stuff... One of the jokes... is how "ugly" WindowBlinds skins are (though there are plenty of awesome ones too). But the thing is, the people who buy WindowBlinds tend to like a different style of skin than the people who would never buy it in the first place.(https://forums.sinsofasolarempire.com/303512/piracy-pc-gamin... 2008)
And anyway, who cares? At the end of the day, you go and buy Windows on your handheld. Would people spend $15 on an OEM license of Windows for the Deck to have their games not have this problem? Probably. It is good for the planet that Valve is trying to make WINE emulation work, but it's not doing anything for the consumer - Unity is, middlewares are - and it's not doing anything for the game developer - more games would sell if it was just Windows - so really, pass your judgement in an objective reality. I really don't see the point of DXVK.
So I wonder about the metric for "playable" -- is a 4x greater chance of crashing when using your Steam Deck vs your PC still "playable"? And I wonder if players, who can't tell if a crash is related to emulation or a bug in the game, will blame the game developer when it isn't really fair to hold them accountable.