1031226 - [vibration] navigator.vibrate([]) fails to cancel a pre-existing instance of the processing vibration patterns algorithm

Status: VERIFIED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Zhiqiang Zhang]

User Agent: Mozilla/5.0 (Windows NT 6.2; WOW64; rv:30.0) Gecko/20100101 Firefox/30.0 (Beta/Release)
Build ID: 20140605174243

Steps to reproduce:

Navigator to http://www.w3c-test.org/vibration/cancel-with-empty-array-manual.html

Click the 'Vibrate!' button.


Actual results:

The device vibrates for a short period of time (roughly one second)


Expected results:

It vibrates for a longer time (roughly five seconds, it should feel somewhat long)

Comment 1

[Boris Zbarsky [:bzbarsky]]

We explicitly bail out early for navigator.vibrate on empty pattern, and there's even a comment there saying this is what the spec says.

In particular, the early return on empty pattern was added in bug 884935.

It does look like the current spec draft has "Cancel the pre-existing instance of the processing vibration patterns algorithm, if any. " before the "If pattern is an empty list, or if the device is unable to vibrate, then return true and terminate these steps." bit.

I did check, and the spec draft that was current back when bug 884935 was fixed also has the "cancel" verbiage...

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

That was indeed a mistake and bug 1014581 didn't fix that case.
	

Andrew, want to fix this, or should I?

Comment 3

[Olli Pettay [:smaug][[e-mail]]]

(Andrew, feel free to take this if you have patch before me.)

Comment 4

[Zhiqiang Zhang]

Seems I mistake the 'Actual results' and 'Expected results'; can someone swap them?

BTW, what does 'Flags: needinfo?' mean? Do you need more info from me, or something else?

Comment 5

[Olli Pettay [:smaug][[e-mail]]]

I was asking needinfo from Andrew.

And bugzilla comments can't, in general, be modified afterwards.

Comment 6

[Olli Pettay [:smaug][[e-mail]]]

Attachment: empty_pattern.diff

As far as I see, Hal should be able to take care of the right behavior.

Comment 7

[Mounir Lamouri (:mounir)]

Comment on attachment 8450402 [details] [diff] [review]
empty_pattern.diff

Review of attachment 8450402 [details] [diff] [review]:
-----------------------------------------------------------------

r=me, sorry for not spotting that.

Did you check that Hal is actually doing the right thing here when the pattern is empty?

Comment 8

[Olli Pettay [:smaug][[e-mail]]]

Based on code inspection it should do the right thing.

Comment 9

[Olli Pettay [:smaug][[e-mail]]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d6a41b0f9794

Comment 10

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/d6a41b0f9794

Comment 11

[Zhiqiang Zhang]

Thanks for the fixing. The test passes on Nightly (2014-07-13).