W3C home > Mailing lists > Public > public-device-apis-log@w3.org > May 2017

Re: [sensors] Use ''sampling frequency" instead of "polling frequency"

From: Tobie Langel via GitHub <sysbot+gh@w3.org>
Date: Fri, 26 May 2017 10:00:02 +0000
To: public-device-apis-log@w3.org
Message-ID: <issue_comment.created-304242470-1495792800-sysbot+gh@w3.org>
> Even platform APIs do not allow this. I don't think we should set such goals for JS API.

I think we're running around calling different things with different names and that makes the whole conversation very difficult.

For the sake of this argument, let's say that setting a `samplingFrequency` to `1000` will allow the Web developer to get 1000 readings per second.

Now let's agree that setting `reportingFrequency` to `60` and `samplingFrequency` to `1000` will allow the Web developer to get 60 readings per second, all of which with a latency smaller than `1ms + PIPELINE_LATENCY`. Once some form of bundling mode is added, this would enable the developer to grab around 17 samples 60 times per second. 

In contrast,  setting `reportingFrequency` to `60` and `samplingFrequency` to `60` would only ever allow developer to get 60 readings per second, all of which with a latency that could be up to  `~17ms + PIPELINE_LATENCY`.

Now, how this  `samplingFrequency` actually translates at the platform API level or the HW level is an implementation detail, as long as the contract described above is respected.

-- 
GitHub Notification of comment by tobie
Please view or discuss this issue at https://github.com/w3c/sensors/issues/209#issuecomment-304242470 using your GitHub account
Received on Friday, 26 May 2017 10:00:12 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 12:18:53 UTC