Safari 9 scroll is either 1 or -1

[drewbaker]

On Safari 9.0 (the latest), event.deltaY is either 1, or -1. It never goes any higher than that. So it's not consistent across browsers. On Chrome it works fine.

[dmethvin]

I don't have a Mac with iOS to try this on so others will need to provide some more detail. Did Safari make some kind of intentional change here? You're saying the event.deltaY value provided by the plugin is only 1 or -1, what are the raw values Safari is giving?

[drewbaker]

It's strange, because it works with the mouse wheel on my logitech mouse, but if I use the trackpad, it is just -1 or 1. Chrome and Firefox is fine.

Here is a screenshot of the entire event, when using trackpad:

And here it is when using mousewheel on external mouse:

[christian-thomas]

I've also got the same issue — All other browsers are fine, it's just the latest version of Safari that's giving 1/-1.

[dmethvin]

Put a breakpoint inside the plugin, on this line. What path does Safari 9 take, and what values does it calculate? How does that differ from Safari 8?

[christian-thomas]

So, from a brief bit of debugging, I amended lines 106-108:

    if ( 'detail'      in orgEvent ) { deltaY = orgEvent.detail * -1;console.log('detail - ' + deltaY);      }
    if ( 'wheelDelta'  in orgEvent ) { deltaY = orgEvent.wheelDelta;console.log('wheelDelta - ' + deltaY);      }
    if ( 'wheelDeltaY' in orgEvent ) { deltaY = orgEvent.wheelDeltaY;console.log('wheelDeltaY - ' + deltaY);      }

It looks like Safari 9 has some sort of debounce/throttling...

In Safari 8, depending how much I scroll, in a single mousewheel motion/scroll I'll get something like this:

    [Log] detail - 0
    [Log] wheelDelta - -60
    [Log] wheelDeltaY - -60
    [Log] detail - 0
    [Log] wheelDelta - -30
    [Log] wheelDeltaY - -30
    [Log] detail - 0
    [Log] wheelDelta - -90
    [Log] wheelDeltaY - -90

Same mousewheel motion/scroll in Safari 9, just:

    [Log] detail - 0
    [Log] wheelDelta - -3
    [Log] wheelDeltaY - -3

[Awea]

Hi there, any idea for a crappy work around for this one ?

[dmethvin]

It seems like it might be due to this Webkit bug reported by @mzgol : https://bugs.webkit.org/show_bug.cgi?id=149526

[kpettinga]

Experiencing the same issue over here as well. Is this something we'll have to wait for Safari or Webkit to fix?

[johndigital]

I'm not sure if this helps, but I've found a workaround for this. In my situation I'm trying to detect the deltaY when scrolling a non-scrollable page and use the value. Like everyone else I get a totally normal value in chrome and FF, but in Safari I just get 1 or -1 when using a magic mouse or trackpad:

It seems like this is being caused in part by the elastic scrolling that safari does. If the default scrolling behavior is disabled before using the deltaY value then the normal value comes through fine. The two easiest ways to achieve this:

1. Put overflow: hidden; on the body. This disables all scrolling
2. Add event.preventDefault(); before using event.deltaY. This will kill the default mousewheel behavior

Either of those fix it for me both using the mousewheel plugin and the underlying 'mousewheel' event. This obviously won't help if the default scrolling behavior needs to stay in tact, but worked great for my situation.

[dmethvin]

Nice detective work there @jrobson153! This leads me to think that there isn't a solution that the plugin can automatically implement to solve this because either of those has tradeoffs. If I knew of anyone at Apple I could escalate this to I would, but Apple isn't known for its responsiveness on browser issues. What would be the best place to document this, perhaps the README.md file?

[johndigital]

Yea, it could take them a while to catch up to this. I think a brief note in the Readme file would be helpful!

[dmethvin]

Did this get fixed in Safari 9.1?

[mgol]

My initial comment was wrong, the bug https://bugs.webkit.org/show_bug.cgi?id=149526 is still present in Safari 9.1.

EDIT: It's also still present in Safari Technology Preview & WebKit Nightly.

[sandrodz]

ah Safari... new IE never fails to deliver bogus issues, its been 2 years now this has not been fixed?