[scroll-animations-1] currentTime when scroll range is 0?

[flackr]

The spec does not yet include the calculation for the currentTime, however the description of currentTime in the IDL interface suggests that the timeline is only inactive (producing a null time) when the principle box is not a scroll container.

What happens when we have a ScrollTimeline source scroll container with 0 range?

I think the logical conclusion is that this is another case in which the timeline should be considered inactive and return a null time.

[fantasai]

Edited in as inactive. Agenda+ to review with the CSSWG.

CC @bramus @mirisuzanne @rachelnabors

[bramus]

LGTM

[flackr]

There's a couple existing locations that need to be edited:

• ScrollTimeline constructor states when the timeline is active / inactive
• ViewTimeline constructor states when the timeline is active / inactive

[mirisuzanne]

Makes sense to me as well.

[BorisChiou]

Inactive timeline sounds reasonable to me as well.

[css-meeting-bot]

The CSS Working Group just discussed [scroll-animations-1] currentTime when scroll range is 0?, and agreed to the following:

• RESOLVED: In the absence of scroll range, timeline becomes inactive

The full IRC log of that discussion

<emilio> topic: [scroll-animations-1] currentTime when scroll range is 0?
<emilio> github: https://github.com//issues/7401
<emilio> flackr: when you have a scroll timeline the currentTime is effectively the scroll-position / range of scroll
<emilio> ... but if there's no scroll range then you'd have division by zero / infinite range
<emilio> ... we think that the timeline should become inactive when in this state
<emilio> ... rather than being at the beginning or end or something
<emilio> Rossen_: makes sense, question would be: if you have a scrollport that has content which isn't exceeding the scrollable bounds and you resize it, you're saying that only when you have scroll range the timeline is valid
<emilio> ... numerically I get your point
<emilio> ... in the case when the scrollport resizes
<emilio> ... what does that do to the animation?
<emilio> ... you said the timeline becomes invalid?
<emilio> flackr: inactive, which is an existing state, where the currentTime is null
<emilio> ... so basically if you resize so that it's not scrollable the effects will no longer apply
<Rossen_> q?
<emilio> emilio: when you stop applying the effects the animation doesn't pause right? You get discontinuity
<emilio> flackr: that's right, is not weirder than having a very small scroll range
<dholbert> scribenick: dholbert
<emilio> dbaron: can you have overscroll in these elements?
<dholbert> flackr: I don't think so
<emilio> flackr: depends on browser, I think on ~all browsers you don't get overscroll effect
<dholbert> scribenick: emilio
<emilio> ... but on safari desktop it does
<emilio> ... there's also the question of whether the overscroll effect factors in the scroll timeline
<emilio> dbaron: I was just wondering if you could end up with weird effects in the overscroll effect
<emilio> flackr: I think we'd still not have a reasonable range
<Rossen_> ack dbaron
<emilio> ... I think we shouldn't factor the negative range into the effect
<smfr> i agree with not applying overscroll here
<emilio> ... it's more stuff to worry about for developers, plus browsers don't technically limit the amount of stuff you can overscroll
<emilio> dbaron: that's a good enough answer I think
<emilio> RESOLVED: In the absence of scroll range, timeline becomes inactive

[fantasai]

Thanks @flackr, fixed in fdd21a5