Re: Vibration

[+public-device-api]

On 09 Oct 2014, at 15:46, Anne van Kesteren <annevk@annevk.nl> wrote:

> Is there anything in http://dev.w3.org/2009/dap/vibration/ you would
> have done differently today in terms of the API? We want to add
> vibration support to the notifications API as part of a general set of
> behaviors, see https://github.com/whatwg/notifications/issues/22 Was
> just wondering if the API should be the same or slightly different. In
> particular overloading long with a sequence of long does not seem
> great.

We tried to keep the scope as tight as possible and did not add any extras, thus not much to do differently.

The overloading aspect of the API was there from the start. This is something we inherited from the early Mozilla experimental implementations and managed to stood the test of time. Some pretty cool projects were built atop this API, so the design seems to work fine from the web author’s perspective at least.

Some people questioned whether there should be means to detect the existence or the actual vibration motor. We settled on a design that the API behaves identically regardless of whether the hosting device has an actual vibration motor or not. This is to allow implementers to use alternative means to emulate vibration (sounds, shake screen, electric shock, whatnot the future brings) in devices that do not have the supporting hardware. Secondly, we’re not leaking any information that could be used for fingerprinting.

Someone also proposed strength/intensity feature, but since it was demonstrated one can emulate intensity with the existing API using pulse modulation methods pretty easily, we did not see great rush to spec that. Also hardware support for strength was/is limited.

The current API ships in both desktop and mobile browsers [1], and we have a test suite [2].

Thanks,

-Anssi

[1] http://caniuse.com/#feat=vibration
[2] http://w3c.github.io/test-results/vibration/all.html

Received on Thursday, 9 October 2014 13:33:17 UTC