SDL3 Improves Steam Controller Support, Now Enabled By Default

[phoronix]

Phoronix: SDL3 Improves Steam Controller Support, Now Enabled By Default

A number of Steam Controller improvements have been merged for SDL, this widely-used hardware/software abstraction layer that is common to cross platform games. Among the latest Steam Controller improvements in SDL are enabling the support by default...

https://www.phoronix.com/news/SDL-Steam-Controller-Default

[WebMac]

Will this work flawlessly with Wayland? At least the current support for the Steam controller under Wayland (even in Steam games) is some kind of broken. You can use it, but the cursor that's visible on the screen won't move. Although the real cursor (which isn't visible) DOES move. So you have no idea where you are going with the cursor when using the touch pad on the Steam Controller. I hope this will get fixed some day...

[chocolate]

Interesting device but horrible everyday controller with no technical support at all.
Super loud and clicky buttons worthy of the worst second player hand-me-downs you could possibly imagine.

Bought one on eBay out of curiosity. Turns out the Steam client lets you change the behavior of the touch circles but then doesn't save those settings anywhere, the new firmware maybe fixes this and maybe it doesn't, but you need to upgrade it via Windows from a batch script that a Valve developer threw out the window, since upgrading the firmware via Steam has been discontinued.
When you do, you gain the much requested bluetooth functionality, but the wi-fi dongle won't work anymore. The closest thing to an official guide will tell you to just pair the dongle again via Steam, which won't work either. So people online will tell you to download an old Steam client to retry the firmware upgrade and pairing from there, and similar follies.

Finally, there's an open-ish, abandoned, alternative firmware with tons of markdown text explaining what the developer had for breakfast, and little else.

Cool that it still gets some support here and there, but stupid of Valve to completely let it go after clearing their stock. Strange that they couldn't release the firmware as open source themselves; I can't imagine there being any third-party IP except for the radio stuff.

[risho]

I absolutely love the steam controller, but it was very clearly a skunk works project that was at best half baked on release. It is made with cheap materials and strangely designed, and you can see it's final form in the steam deck. One of the biggest problems is that the steam controller is like half computer interface and half traditional controller which makes it not that great at either. I think if they made a new steam controller today that iterated and took inspiration from the steam deck design that it could easily be the best desktop computer controller on the market.

[varikonniemi]

Interesting device but horrible everyday controller with no technical support at all.
Super loud and clicky buttons worthy of the worst second player hand-me-downs you could possibly imagine.

Bought one on eBay out of curiosity. Turns out the Steam client lets you change the behavior of the touch circles but then doesn't save those settings anywhere, the new firmware maybe fixes this and maybe it doesn't, but you need to upgrade it via Windows from a batch script that a Valve developer threw out the window, since upgrading the firmware via Steam has been discontinued.
When you do, you gain the much requested bluetooth functionality, but the wi-fi dongle won't work anymore. The closest thing to an official guide will tell you to just pair the dongle again via Steam, which won't work either. So people online will tell you to download an old Steam client to retry the firmware upgrade and pairing from there, and similar follies.

Finally, there's an open-ish, abandoned, alternative firmware with tons of markdown text explaining what the developer had for breakfast, and little else.

Cool that it still gets some support here and there, but stupid of Valve to completely let it go after clearing their stock. Strange that they couldn't release the firmware as open source themselves; I can't imagine there being any third-party IP except for the radio stuff.

Valve should release a steam controller 2 that fixes all this shit. With launch day Linux support. Otherwise it's hard to convince people that they support Linux gaming, instead of only Linux gaming on the steam deck.

[risho]

if you are questioning whether valve cares about gaming on linux then you just havent been paying attention. they are in fact SOLELY responsible for the success that gaming has seen on linux and they have been doing this for years before the steam deck even existed. wine existed since the 90s on linux and was barely usable for decades. once valve showed up wine went from barely usable to basically on par with windows in like 2 years. on top of that the steam deck isn't some random proprietary appliance that happens to be running linux, it is literally arch linux with a full kde desktop with flatpak and distrobox giving it access to anything that you could want to do on linux that doesn't require a kernel module.

[johanb]

Too little and too late, I have since replaced my two Steam Controllers with two PS5 controllers, I'm sure most Steam Controllers are collecting dust nowadays.

[Eudyptula]

The Steam Controller is really so close to being what a controller should be. I feel like its lack of success and mainstream appeal is due to it not being fully cooked yet.

While the controller itself is actually great, the setup and configuration was kind of a hassle and a mess. They shouldn't reduce the customizability, but the UI and UX should've been improved. Disclaimer, last time I used it was several years ago, so things might've improved (though, from videos I've watched on it, it doesn't seem to have changed too much).

The controller itself is pretty decent, but the trackpads should be both larger and more precise. Some of the buttons don't have the best feel.

But with the improvements they've made for the controller components integrated into Steam Deck and some additional development, a Steam Controller 2 should be able to finally convince people. I've been waiting for a Steam Controller 2 since... Well, since I got my Steam Controller.

It's nice to see that the Steam Controller related stuff is getting some support and improvements, though. I just really think Valve should put some resources into a second generation controller. Not the trackpad-less one they just licensed. There's a mountain of joystick-based controllers on the market, we don't need more of those! 8BitDo seem to have made the best joystick-controller, so if Valve takes some inspiration from what makes the 8BitDo controllers great and incorporate the Steam Controller magic in addition, then we'll finally have the "final controller". With large dual trackpads.

I fear that Valve is looking too far ahead and is too focused on VR and AR in such a way that they ignore what they deem the past or current. Not to say they shouldn't explore that stuff (speaking of, I hope the Index 2 rumours are true), but they can do both. They don't have to abandon the past and traditional input devices just because there are interesting next-gen avenues to explore.

Anyway, perhaps I should dust off my Steam Controller and start using it again (and also start occationally gaming again).

Too little and too late, I have since replaced my two Steam Controllers with two PS5 controllers, I'm sure most Steam Controllers are collecting dust nowadays.

Nothing is ever too late. And there will be people who want to continue to use their Steam Controller. And just because Valve sort of messed up the release of the Steam Controller, it doesn't mean they can't release a Steam Controller 2 and do it properly this time. Their hardware and product designing has improved a lot since then and they've gained a lot of experience I'm sure would be highly valuable for a second generation controller.

No offence to your controllers of choice, but the PS5 controllers surely isn't something that could be considered the ultimate controller or anything close to it, in my opinion. And it doesn't replace the trackpad idea (that just needs to be made well enough to actuall convince people instead of going half the distance leaving a sort of unfinished feel over the product).

[Eudyptula]

Interesting device but horrible everyday controller with no technical support at all.
Super loud and clicky buttons worthy of the worst second player hand-me-downs you could possibly imagine.

Bought one on eBay out of curiosity. Turns out the Steam client lets you change the behavior of the touch circles but then doesn't save those settings anywhere, the new firmware maybe fixes this and maybe it doesn't, but you need to upgrade it via Windows from a batch script that a Valve developer threw out the window, since upgrading the firmware via Steam has been discontinued.
When you do, you gain the much requested bluetooth functionality, but the wi-fi dongle won't work anymore. The closest thing to an official guide will tell you to just pair the dongle again via Steam, which won't work either. So people online will tell you to download an old Steam client to retry the firmware upgrade and pairing from there, and similar follies.

Finally, there's an open-ish, abandoned, alternative firmware with tons of markdown text explaining what the developer had for breakfast, and little else.

Cool that it still gets some support here and there, but stupid of Valve to completely let it go after clearing their stock. Strange that they couldn't release the firmware as open source themselves; I can't imagine there being any third-party IP except for the radio stuff.

Yeah, the click feel isn't something to rave about. But it says something when so many people are passionate about this controller despite some of these shortcomings. I just hate the thought of this controller philosophy being abandoned and disregarded both by the masses as well as Valve because people don't have imagination enough to deem it a good idea despite its flaws. The idea is great, the execution is so-so. Valve can do better, especially now. Are they afraid of releasing a second one because of how the first one was received? Either way, they should be the fearless Valve that puts out great ideas and concepts and do it properly, as to avoid good ideas being shut down because of bad execution.

Settings should be stored in the controller itself. I realize not all configs could be stored in the controller. But then again, why not? If it has a microSD card slot, the user can store configurations for as many games and applications they want. The driver or Steam client could be the mediator here and apply the settings for the game or application. There already is an interface for it on the desktop OS. This way, you'd have all you configs no matter where you were using it and it wouldn't affect default or plug-and-play usage situations as it would be application-specific. On Linux, can't you simply link it to the Wayland application? And use keyboard mapping translation in the configuration so that you wouldn't even need a translation or configuration interface? And for the more advanced features that cannot be mapped to keyboard strokes, either use the controller interface driver and the Steam Client and controller driver along with necessary games or application support when needed. My understanding is that it's too dependent and not openly supported enough. For instance, can the Steam Controller interface with libinput? I might be overthinking all of this, but I think it makes sense to support universal interfaces to ensure good plug and play experience and thus greatly expand the use-case of the device instead of having to be aligned with specific interfaces, clients, drivers and explicit application support.

Thanks for the link to the open Steam Controller firmware. It may be abandoned, but I'm sure there's useful stuff there if someone where to pick it up again.

Then again, I really don't understand why Valve hasn't open sourced the firmware. As far as I know, people have been asking for it... They released the CAD models for it etc. but not the firmware? If they want to see such a project succeed, a great way is to at least give the community the possibility of improving it. Especially when the product was sort of half-done to begin with.


If anyone from Valve read this, please, open the Steam Controller and make a Steam Controller 2 happen. (Yes, I'm being a hopeful optimist, but you never know. Giving up never solves anything.)

[marlock]

i know SDL3 handles per-app support for controllers to be useful and this is something else, but taking a slight tangent: whatever happened to SC Controller?

obviously supporting Steam Controller in the kernel is better, but it did work fine as a userland stopgap solution and it also offered a nice set of system-wide controller interaction tweaking capabilities... not as featured as steam but FOSS and more reliable

kozec's fork seems dead but not sure if any other forks have lived on to become a polished alternative

https://github.com/kozec/sc-controller