Closed Bug 1023329 Opened 12 years ago Closed 12 years ago

[VsD Refresh] Lockscreen Visual Refresh > Color overlay with notifications wrong with wallpapers chosen from gallery and camera

Categories

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

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(blocking-b2g:2.0+, b2g-v2.0 fixed, b2g-v2.1 fixed)

RESOLVED FIXED
2.0 S4 (20june)
blocking-b2g 2.0+
Tracking Status
b2g-v2.0 --- fixed
b2g-v2.1 --- fixed

People

(Reporter: mnjul, Assigned: mnjul)

References

Details

(Whiteboard: [p=3], [2.0-flame-test-run-2])

Attachments

(2 external files)

The color overlay (average color of wallpaper) when there are notifications in lockscreen, as specified in bug 950884, only works correctly when the wallpaper is picked from "wallpaper", and not from "gallery" or a picture taken from camera on-the-spot. With the latter two approaches of choosing wallpapers, the overlay color will not be calculated at all, remaining being the color of the old wallpaper.
blocking-b2g: --- → 2.0?
blocking-b2g: 2.0? → 2.0+
Comment on attachment 8439795 [details] [review] Patch (Github PR) O-kay, here is the big, big patch. The fundamental issue of this bug is that I only implemented average color calculation for the wallpaper picker in the wallpaper app, which would be entirely useless if the wallpaper was chosen from the inline activity of gallery or camera. So the calculation mechanism must be moved from wallpaper app to lockscreen (or other apps/objects if more suitable) --and this patch moves/reverts some changes in bug 950884 related to this average color calculation functionality. With this patch, when the lockscreen observes change of wallpaper, it is responsible to calculate the average wallpaper color. Note that the wallpaper app has a canvas for the wallpaper when picking, but gallery/camera don’t, and we need to redo the canvas drawing in lockscreen when calculating average color. To reduce computation/performance overhead for color calculation, in this patch I have: 1. Downsampled (by image dimension) the drawn wallpaper image; the longest edge of the drawn-to-canvas image would be 100*DEVICE_PX. Visually the resulting average color has nearly no difference from that of full-sized wallpaper [1]. 2. Deferred the calculation of wallpaper average color from [when wallpaper image is selected] to [when device is locked]. This can reduce the device lag-ness when choosing wallpaper. [1] FYI, on Flame, calculating the average color from a 480x800 full-size wallpaper would take 600~800ms, while with the reduced-size wallpaper (150*250) it takes less than 100ms.
Attachment #8439795 - Attachment description: WIP Patch (Github PR) → Patch (Github PR)
Attachment #8439795 - Flags: review?(gweng)
Well it's embarrassing: conflicts occur...
Flags: needinfo?(jlu)
Conflicts are resolved, waiting for Travis & TPBL's pass...
Flags: needinfo?(jlu)
Things are green again now, so Greg, please review the patch for me, thanks.
Flags: needinfo?(gweng)
Comment on attachment 8439795 [details] [review] Patch (Github PR) Please fix the nit (function designation) I mentioned on the GitHub page, thanks.
Attachment #8439795 - Flags: review?(gweng) → review+
Flags: needinfo?(gweng)
Target Milestone: --- → 2.0 S4 (20june)
I'm waiting for tests to pass for final landing. For future uplift reference: in this bug, I made changes to modifications by bug 1024395; but bug 1024395 itself will not be uplifted to 2.0 branch because bug 973443 will not be uplifted. So when we're uplifting patch for this bug (as this bug is blocking-2.0), we might have to resolve conflicts manually.
Blocks: 1020115
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(ktucker)
Whiteboard: [p=2] → [p=2], permafail
Whiteboard: [p=2], permafail → [p=2], [2.0-flame-test-run-2]
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(ktucker)
QA Whiteboard: [QAnalyst-Triage+] → [QAnalyst-Triage+][lead-review+]
Status: NEW → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
As this is a 2.0 blocker, I will uplift the patch to 2.0 by myself (partly because need for potential conflict resolution as said in comment 7)
Whiteboard: [p=2], [2.0-flame-test-run-2] → [p=3], [2.0-flame-test-run-2]
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: