TPAC 2021 Breakout minutes

Minutes for the "Gamepad API Breakout" - 21 October, 2021

Present: Matt Reynolds (Google), Vincent Scheib (Google), Ben Mathwig (Microsoft), Reilly Grant (Google), Marcos Caceres (W3C) Scribe: Marcos Caceres

gamepad input events #152

issue 152

MR: Gamepad API is polling API currently, with some limited events
...some events were proposed for axis events early
...But nothing landed in the spec gamepad
...We still see a use cases and interest from Cloud gaming clients
...There is interest of seeing a prototype
...It is possible to poll as really fast to get similar effect, but it's problematic.
...It may take some time to implement in Chrome, with potential partnership with Microsoft/Edge
...It would be great to have folks at Microsoft to have a look, particularly around button changes.
...Now is the best time to get feedback, before we start seriously implemented
...It's very likely the proposal will change
RG: We really want hear about how various versions of the proposal will want to hear about that.
BM: even coalescing will likely be very important to consumers
MR: we plan to flush onbutton button up and requestAnimationFrame
...one event per animation frame, so you never get more than to match the display's refresh
...Other design decision was which events do want.
...We had buttons and axis as separate events
...But now we might want to combine them into a single event or maybe two
...We'd really like to get some feedback on how many events we need.
...BM: button up ,button up, and the axis because they wanted to move the cursor
...< cross talk > ... they need to do things like diagonal, to activate menus up and to the right.
MR: that's all I have from this for now
RG: Any signal from Mozilla
RG: any other interest from the other implementers? Wonder if Mozilla is interested?
BM: this might be useful to the iOS version of Safari
MC: We can got and chat to the webkit directly if need be, over on their Slack channels.

Preview | Diff

Proposed Haptics API

BM: we are ready to start sharing it
...It's currently in an origin trial
...Plan is to move into the WICG initially to get feedback. 
...I'm trying to see if I can retrofit it into the Gamepad API
...I'd like look at how to integrate it with the eventing model
...Declarative vs event driven is still a challenge
...Right now, experimenting with a particular pen hardware device
MR: if it's all HID, I wonder if we could do a poliyfill 
BM: we might also look at a "language" to talk about haptics, frequencies and ample
...so how do we describe a buffered web form to describe HID
..Or do we want to treat it like an audio file, and maybe go to the music/audio space for inspiration
MR: this is very interesting
BM: I've also looked at how one transition a wave across a controller (across multiple motors)
MR: you want to control them independently...
Be: so, that's why I see it as different audio tracks... the haptic actuator on have a single array, but would could send a bunch of tracks the get sent to the different actuators.
MR: I'm against sending them as array, because mapping them to a "standard" layout hasn't worked well
BM: ack. You could get a list of actuator.
BM: So you play waveform function into actuators.
MR: how close are you to having a draft?
BM: a couple of weeks
MC: if you need any early review, please reach out.

Issue triage

MR: Chromium might need to make a breaking change to switch to live object.
MR: It would be good change Firefox side
MR: On the Firefox side, the objects are kept alive so long as you don't call getGamepads() again
MR: Ben, you had mentioned about HID integration... the current Gamepad API
< cross-talk >
BM: We want to make sure we can do HID cross OS, so talking to various teams about that.
MR: That's all the topics I wanted to discuss. Any one else have anything else?

Meeting adjourned!