1406364 - Google Slides scrollbar broken in Nightly (drags slides instead)

Status: RESOLVED DUPLICATE of bug 1405397

(Core :: DOM: Core & HTML, defect, P1)

Description

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Broken behaviour in Google Slides in 58.0 (20171003100226 nightly, Windows 10).. STR:

1. Create new presentation
2. Hit the + sign in the top left a bunch to insert enough slides so that a scrollbar appears in the slide list on the left
3. Move mouse to scrollbar thumb, click and hold to drag thumb

Expected:
Drag thumb to scroll slides around

Actual:
You end up grabbing a slide that's sorta near where you clicked and start moving the slide around instead.

Comment 1

[Andrew Overholt [:overholt]]

I can reproduce this in nightly, also on Windows 10. I cannot reproduce in release (56) or beta (57).

QA, can you please run mozregression to see when this regressed?

Comment 2

[Andrew Overholt [:overholt]]

[Tracking Requested - why for this release]: we shouldn't let Google Slides regress like this.

Steven, have you had any reports of this or do you know offhand if recent changes on your end could have caused this?

Comment 3

[Steven Saviano]

Thanks for reaching out!

We have not heard of reports nor know of a recent change that would have caused this. Seems suspicious that it specifies repro in 58 but not 57/56

Comment 4

[Andrew Overholt [:overholt]]

Thanks for confirming, Steven.

FWIW, I tried disabling mouse move event coalescing (dom.event.coalesce_mouse_move) and pointer events (dom.w3c_pointer_events.enabled) and, after restarting, could still reproduce.

Comment 5

[Steven Saviano]

Really appreciate the attention to detail here as well, and concern to not let Slides regress. Thank you!!

Comment 6

[Marcia Knous [:marcia]]

ni on Andrei to help with the regression range.

Comment 7

[Ciprian Georgiu, Desktop QA]

(In reply to Marcia Knous [:marcia - use ni] from comment #6)
> ni on Andrei to help with the regression range.

Hi, please see below the mozregression output.

Last good revision: ce3e3c2f159845603124c4e1ee0b78a284713136
First bad revision: 4c8b85e80aebb27cf3a5334c2c183e49636b9ece

Pushlog:
https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=ce3e3c2f159845603124c4e1ee0b78a284713136&tochange=4c8b85e80aebb27cf3a5334c2c183e49636b9ece

Comment 8

[Timothy Nikkel (:tnikkel)]

The patches in bug 1405397 should fix this.

Comment 9

[Timothy Nikkel (:tnikkel)]

Should be fixed now. Please re-open if not.

Comment 10

[Andrew Overholt [:overholt]]

This is fixed in today's nightly AFAICT. I'll dupe to bug 1405397. Thanks!

Comment 11

[Hsin-Yi Tsai (she/her) | away [:hsinyi]]

Mirroring the firefox status58 of bug 1405397.

Comment 12

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Confirmed fixed for me in Google Slides; however, I suspect the fix caused bug 1409003 which is similar breakage on Telegram's Web UI.