"Developers were kind of stuck between the equation of ‘Do I want to do a bunch of work to port my game? But there’s not an audience for it that justifies this work yet.’ So it was a little bit of a bootstrapping problem.
While Proton is an amazing technology I think it will also be a crutch and an excuse to not make native ports. Proton brought gaming to Linux, but it will also hold games on Linux back for decades.
Performance and playability on Linux are effectively solved by Proton; therefore, the effort required to maintain a native Linux build is an unnecessary and inefficient use of development resources, especially for smaller studios. “Holding games back” on Linux feels like a semantic distinction if we’re moving to a world where every PC game is playable on Linux.
(this assumption relies on Linux marketshare growing and the remaining games that don’t support Proton due to anti-cheat software eventually are pressured to support playing on Linux, even if they don’t build a native linux port. I think we’re well on our way to that future and it’s probably just a matter of time.)
Relying on Proton just offloads development/optimization to the community instead of the actual developers getting paid to develop it. Sure it’s cool the game runs better. But like pay people to do that.
I guess my question is why pay people to solve a problem two ways?
We have an increasingly functional way to play on two platforms with a single build. I’d love to have both for completeness, but as long as Proton is actively being worked on, I feel like that’s good enough and will certainly not hold back gaming on Linux for years to come.
If it gets a proper market share and can’t be ignored, developers will be more inclined to take care of it themselves. A lot of games come already working on it.
If there’s a safe subset of the Windows API defined as a virtual machine for gaming on x86-64 machines, and running with next to no overhead on Linux, does it really matter that the binaries are PE-COFF . EXE files assuming their world to have DirectX rendering and drive letters rather than POSIX?
As SteamOS is Linux, it’s a matter of the Steam Machine actually competing with PS and Xbox.
If it sells enough, it’d be a logical step to start porting directly to it instead of relying on Proton. But then again, logic isn’t always developers or publishers forte.
However, all it’ll take is one studio to find an angle to leverage native linux support in a way that shows its value. What value? I dont know, Im not in that space. But there is a big push from valve here to show devs that this isnt the token effort is was last time, and any effort they put into such a project won’t be wasted.
We just need the killer app, and valve is holding open the space for someone to make it, for now.
We just need the killer app, and valve is holding open the space for someone to make it, for now.
The killer app is already there and it’s Steam and it’s massive library.
Sure, previously Steam has had Steam Link or having a computer connected to your TV, but frankly it just was never going to be a mainstream option. Too much finickiness, it locks up your PC in the other room, sound and controllers are wonky, etc. Local compute on the hardware under your TV is what console gamers are comfortable with and having a PC that isn’t giant and butt-ugly in the living room is a huge hurdle. This hardware, assuming it delivers, is priced right, is a potential console killer.
Agreed. Saying that just because an application layer that allows it to run on Linux exists will prevent developers from simply exporting the game and making the few critical changes necessary in order to make it work on Linux natively are two entirely different things.
The developers that export their games to run on Linux natively will likely have a higher sales rate for their Linux users, as us, the users, will know that the developer cares enough to do the proper export, and therefore we are likely to have some level of support should something go awry.
Whereas with proton, we have to rely on community notes in order to solve incompatibility issues, which can be a major hurdle for the less technologically inclined, AKA the majority of computer users
While Proton is an amazing technology I think it will also be a crutch and an excuse to not make native ports. Proton brought gaming to Linux, but it will also hold games on Linux back for decades.
I half agree. The more we get the masses away from windows and macOS the more likely they’ll be included in the dev process.
Performance and playability on Linux are effectively solved by Proton; therefore, the effort required to maintain a native Linux build is an unnecessary and inefficient use of development resources, especially for smaller studios. “Holding games back” on Linux feels like a semantic distinction if we’re moving to a world where every PC game is playable on Linux.
(this assumption relies on Linux marketshare growing and the remaining games that don’t support Proton due to anti-cheat software eventually are pressured to support playing on Linux, even if they don’t build a native linux port. I think we’re well on our way to that future and it’s probably just a matter of time.)
Relying on Proton just offloads development/optimization to the community instead of the actual developers getting paid to develop it. Sure it’s cool the game runs better. But like pay people to do that.
Valve does, they literally have teams dedicated to the work.
Also, do you guys (who can afford to) not have re-occuring donations going out to your favorite open source projects every month?
I guess my question is why pay people to solve a problem two ways?
We have an increasingly functional way to play on two platforms with a single build. I’d love to have both for completeness, but as long as Proton is actively being worked on, I feel like that’s good enough and will certainly not hold back gaming on Linux for years to come.
If it gets a proper market share and can’t be ignored, developers will be more inclined to take care of it themselves. A lot of games come already working on it.
It’s a crutch until it isn’t. I’m guessing that the majority of Machine buyers won’t install windows on it.
If there’s a safe subset of the Windows API defined as a virtual machine for gaming on x86-64 machines, and running with next to no overhead on Linux, does it really matter that the binaries are PE-COFF . EXE files assuming their world to have DirectX rendering and drive letters rather than POSIX?
As SteamOS is Linux, it’s a matter of the Steam Machine actually competing with PS and Xbox. If it sells enough, it’d be a logical step to start porting directly to it instead of relying on Proton. But then again, logic isn’t always developers or publishers forte.
Im upvoting as this is a valid concern imo.
However, all it’ll take is one studio to find an angle to leverage native linux support in a way that shows its value. What value? I dont know, Im not in that space. But there is a big push from valve here to show devs that this isnt the token effort is was last time, and any effort they put into such a project won’t be wasted.
We just need the killer app, and valve is holding open the space for someone to make it, for now.
The killer app is already there and it’s Steam and it’s massive library.
Sure, previously Steam has had Steam Link or having a computer connected to your TV, but frankly it just was never going to be a mainstream option. Too much finickiness, it locks up your PC in the other room, sound and controllers are wonky, etc. Local compute on the hardware under your TV is what console gamers are comfortable with and having a PC that isn’t giant and butt-ugly in the living room is a huge hurdle. This hardware, assuming it delivers, is priced right, is a potential console killer.
Linux ports are more common now than they ever have been. I think you might be off base with this take.
Agreed. Saying that just because an application layer that allows it to run on Linux exists will prevent developers from simply exporting the game and making the few critical changes necessary in order to make it work on Linux natively are two entirely different things.
The developers that export their games to run on Linux natively will likely have a higher sales rate for their Linux users, as us, the users, will know that the developer cares enough to do the proper export, and therefore we are likely to have some level of support should something go awry.
Whereas with proton, we have to rely on community notes in order to solve incompatibility issues, which can be a major hurdle for the less technologically inclined, AKA the majority of computer users