I proposed this a long time ago

Where?

I proposed this a long time ago

Where?

On discord. But my memory was poor. It was more a question and you said I can give it a try. I guess I started to work on this a bit later but then abandonded it because of time.

Thanks

I did a test of the multitouch feature and it is working well. When is it planned to be released?

I don't know. Nobody reviewed this yet.

I did a test of the multitouch feature and it is working well. When is it planned to be released?

I don't know. Nobody reviewed this yet.

Everyone on the team has been very busy the past few months, and haven't had a whole lot of time to put towards Silk.NET, so open PRs such as these have stalled quite a bit. We apologize for the lengthy review/response times, and have no ETA for the next release/when this PR will be merged.

I notice this currently doesn't have a proposal in documentation/proposals. Ideally, this should be done to ensure we can review the API at a high-level and keep a record of our discussion/decisions.

if this ends up being accepted, I am very willing to port mygesture recognition work I'm making for Godot to Silk.

IIRC the only work I have left on it is godot-specific implementation details. It features support for multiple types of gestures occurring simultaneously and felt really good to me at the time I had it running on Godot 3.

Thanks@domportera. It's hopeful to see the positivity on this PR! At the moment this is currently blocked on a proposal needing to be written (as perdocumentation/proposals). This doesn't necessarily have to be done by Pyrdacor, I don't know what their availability is like, so anyone can feel free to do this. This allows the Working Group to bikeshed the API details in a meeting and ensure our usual design processes are followed.

looking at the code in this PR there is actually a ton of overlap in what this addresses vs what mine does, so maybe it's best not to step on anyone's toes

that being said, because of that, I can probably whip up a proposal that aligns pretty closely with the original author's intent. I'm not terribly informed on the state of touch in Silk though, as I've only ever used Silk for desktop imgui stuff atp. I'm assuming it currently looks like "raw" touch events with finger persistence from the SDL's implementation, but not so low level as to have actual sensor information? do other backends have a working touch implementation?

Anything you can get through SDL is at your disposal. I'm assuming GLFW doesn't have the APIs we require.

@domportera Note that there's some interest from the Stride project about sensors. I think it may just be worth writing this proposal atop Multi-Backend Input (i.e. 3.0) or at least evaluating the differences between 2.X and 3.0 to see if there's a way we can have one proposal for both 2.X and 3.0, if not I'd rather we focus on 3.0.

I have copied the feedback I have received from Stride so far on Multi-Backend Input below for reference.

I read over the whole Input proposal (okay, I skimmed some parts of it), as I have decent experience in that area.
It looks pretty good, and I don't see any issues if Stride switched to it, with just a thin wrapper around it.

There are a few things

• It doesn't look like pens and touch screens are supported? Basically pointer devices. Stride doesn't actually have the concept of a Mouse exposed to the user, only a Pointer., with an exception being the virtual input system (where you define an 'action' and set different input device buttons to raise that action).
• Doesn't look gestures are a built-in thing. Stride has them for doing like flicks, long presses, tabs, etc. And supports setting the needed number of fingers and such. (not 100% clear if these are actually connected to devices currently in Stride)
• More of a question. IsInputContext.ConnectionChanged raised when a device is connected/disconnected. Like if a gamepad gets disconnected?
• Stride has the concept of aISensorDevice which is a type ofIInputDevice. These sensors are like an accelerometer, and gyroscope, etc. that give real world motion data about the device. It don't seem like they would be supported in the current proposal.
• Doesn't look like mouse delta is support. It is relatively easy for a user (like Stride) to implement, but might be worth adding?
• I didn't see a way to get or a good way to implement getting the latest input device to receive input. This info is mostly used in games to update UI icons that show icons for inputs (like switching from "Press E to open" to "Press (X) to open")
• I would recommend using the cardinal directions for the gamepad buttons and dpad instead of up/down/left/right, so north/south/east/west. This avoid having to check if "up" is pressed down, which can be a bit confusing.
• I don't seeText input events? But maybe that is supported elsewhere I just didn't see?

damn ok this seems right up my alley, as I have another project (in its ABSOLUTE infancy, called omnio, made in godot, also will be using my gesture library and arbitrary sensors lmao) that is going to have to tackle very similar issues of abstracting inputs from all sorts of sensors, gamepads, and mouse/KB/touch. Im also interested in expanding the input capabilities of tooll3 after our next release and have begun to sneak silk windows into the codebase already. so this seems like a really good opportunity to get the whole damn .net ecosystem covered

if we wanna go that deep I'm absolutely game and have started to put some thought into that. I work with weird sensor inputs in C# for a living and have had my fair share of pain points so I can probably brainstorm something that more or less covers all the bases. I'm down to propose an everything-input-proposal but would need to seriously ramp up on Silk to keep it relevant

agree