Simon McVittie

Steam supports non-Wayland systems, so I am probably going to have to implement something that looks very much like what I proposed here one way or another, regardless of whether...

> libmanette does a bit of check (more less the same thing as udev does for ID_INPUT_JOYSTICK). See is_game_controller function in libmanette-monitor.c. Thanks, I'll take a look! > you can't...

Ah, sorry, I was looking in the wrong place - I assumed libmanette's filtering would be in the Monitor, but it's actually in the Device.

https://gitlab.steamos.cloud/steamrt/steam-runtime-tools/-/merge_requests/159 is a better link for my portal prototype, !158 is just the API. The way the portal works is that the server-side uses an implementation of the SrtInputDeviceMonitor GInterface...

> Just going to mention that it might be a better long term plan to instead adopt an OpenXR input action style API This is an interface in which the...

> if they use SDL1 or some other means of reading the game controllers, they can't see gamepads at all I think this is most likely because SDL1 and those...

> I am sure your response to that revelation is going to be "oh my god please don't do that" Less extreme than that, perhaps, but... You should be aware...

"Oh my god please don't do that" is tempting as a response to using SDL 1 this decade, though :-)

SDL 2 is likely to be the closest thing there is to doing gamepad enumeration the right way (disclaimer: I contributed many of the patches that made this happen, so...

> do we presently have the ability to override the behavior of applications that are doing it the wrong way, to have them do it the "not wrong, but not...