919899 - [Lockscreen] There's a tiny black line during unlock animation

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::System::Lockscreen, defect)

Description

[Al Tsai [:atsai]]

While user drag the blue ball to unlock the screen, there's a tiny black line show in the blue unlock screen bar.

Comment 2

[Ivan Tsay (:ITsay)]

Hi Greg,

Is this on the duplication bug? Or are you working on this one?

Comment 3

[Tim Guan-tin Chien [:timdream] (please needinfo)]

Set to block until we find the duplicate bug.

Comment 4

[Greg Weng [:snowmantw][:gweng][:λ]]

I'm currently working on Bug 919410 and Bug 921360, while the later would be a workaround patch.

Comment 5

[Lucas Adamski [:ladamski]]

These vertical lines are present even when you aren't dragging to unlock, they are just harder to see.  Looks like maybe we just have some partially-opaque layers that are overlapping.

Comment 6

[Ivan Tsay (:ITsay)]

Greg, do we need this one to be fixed?

Comment 7

[Greg Weng [:snowmantw][:gweng][:λ]]

(In reply to Ivan Tsay (:ITsay) from comment #6)
> Greg, do we need this one to be fixed?

The Bug 921360 had almost fixed the issue, but it's just a workaround patch. Because the symptom even vary from different devices (FirefoxNighly: no such vertical lines; Unagi: black lines due to the glitches; Nexsus 4 & one another LG device: white lines due to the pixels are too crowded).

Jerry from the Gecko graph team have given me some opinions and helped me to do the test. According to the information, the current implementation may be replaced with other possible methods, and that would be a ultimate fix but brings a lot of changes.

Comment 8

[Lucas Adamski [:ladamski]]

This is still quite visible on the Peak at least.  We should fix this or backout the UI.

Comment 9

[Greg Weng [:snowmantw][:gweng][:λ]]

It's difficult because that the issue even has different behavior on different devices. On Nexsus 4 or LG device, its white due to the overlay of DOM elements, and on Unagi and simulator, it almost invisible. Thus the density of pixels obviously affects the issue.
 
Now I'm working on Bug 919410 to fix this, as I mentioned in Comment 4. IMHO, this is actually a duplicated bug. The following discussions can follow on Bug 919410 (would completely resolve the issue) and Bug 921360 (the workaround, and the current situation).

However I must admit that the Bug 919410 is to use canvas to re-implement the whole lockscreen, so the progress may be slow. I may finish a prototype within these days and push a WIP patch, but I still don't know how many difficulties I would encounter before it become perfect. Issues especially like performance tuning may become endless requests.

(use NI to answer Lucas)(I don't know how many people would still read CC info without NI)

Comment 10

[Tim Guan-tin Chien [:timdream] (please needinfo)]

I've filed bug 937023 to reflect our on-going discussion and the workaround approach.

Comment 11

[Greg Weng [:snowmantw][:gweng][:λ]]

The Bug 919410 has been landed, so far so good.
Tim: If this bug is not focus on v1.2, then we may close it. The canvas version no longer comes with glitches like the previous version.

Comment 12

[Tim Guan-tin Chien [:timdream] (please needinfo)]

(In reply to Greg Weng [:snowmantw] from comment #11)
> Tim: If this bug is not focus on v1.2, then we may close it. The canvas
> version no longer comes with glitches like the previous version.

It's not, although I will still mark v1.2 as unfixed if people changed their mind and want to revisit it. Thanks!