Closed Bug 849441 Opened 12 years ago Closed 12 years ago

[B2g] [Webapps] The wrapper loads slowly causing the user to see the background

Categories

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

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(b2g18+ fixed)

RESOLVED FIXED
Tracking Status
b2g18 + fixed

People

(Reporter: dsubramanian, Assigned: crdlc)

References

Details

Attachments

(3 files)

Description: Go to everything.me page. Click on one of the categories listed and tap on any app. While the app is loading the background screen image at the bottom is seen first and then the wrapper loads up. Clicking on home button, the user can see the same background change as the wrapper unloads. This happens very quickly. Repro Steps: 1) Updated to Unagi Build ID: 20130308070202 2) Go to everything.me 3) Click on one of the categories 4) Open an app (while app loading user sees at screen background at the bottom) 6) Tap on the home button to leave the app 7) User sees the background screen changing at the bottom again Expected: The user is able to see the wrapper load and unload properly. Actual: The user is unable to see the loading and unloading of the wrapper appropriately. Repro frequency: 100% Environmental Variables: Verified on Unagi Build ID: 20130308070202 Kernel Date: Dec 5 Gecko: http://hg.mozilla.org/releases/mozilla-b2g18/rev/94927529a55f Gaia: 87801b12cd3aac153fae84844da8f1e6455ea679 Also, Repros in Master Build ID: 20130308031058 Kernel Date: Dec 5 Gecko: http://hg.mozilla.org/mozilla-central/rev/cb432984d5ce Gaia: c7ac21c48e3718a178340a56b17a0508ae903aba Other Notes: Three screenshots attached
Component: Gaia → Gaia::Everything.me
Assignee: nobody → crdlc
Status: NEW → ASSIGNED
Attached file Patch v1
Attachment #723453 - Flags: review?(timdream)
Component: Gaia::Everything.me → Gaia::System
Comment on attachment 723453 [details] Patch v1 It's sad that window manager doesn't comes with unit tests. Please manually test your patch thoroughly before merging the patch.
Attachment #723453 - Flags: review?(timdream) → review+
Comment on attachment 723453 [details] Patch v1 Please, could you test it on master? Thanks a lot
Attachment #723453 - Flags: feedback?(ran)
Attachment #723453 - Flags: feedback?(evyatar)
If feedback is +, I will land on master, thanks Tim
The user is not able to see the wrapper load and unload properly while opening webapps. This issue was found and tested on both in Leo and in Master (also found in the description). Master Build ID:20130312030649 Kernel Date: Dec 5 Gecko: http://hg.mozilla.org/mozilla-central/rev/519c7cf5605b Gaia: ee236c942f2b9abe9065c2fb967b715770a43d9c And also on Unagi Build ID: 20130312070202 Kernel Date: Dec 5 Gecko: http://hg.mozilla.org/releases/mozilla-b2g18/rev/99680a32f297 Gaia: de3e5b9205e6cb1a6bd0858a98d159272ad96d11
tracking-b2g18: --- → ?
Attachment #723453 - Flags: feedback?(ran) → feedback+
Attachment #723453 - Flags: feedback?(evyatar)
https://github.com/mozilla-b2g/gaia/commit/e707086b6ccd9922ba47aada06fbc4e098bcdb4d No risk, very simple and it improves the user experience a lot
Status: ASSIGNED → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED
there will likely be conflicts because of Bug 844990 that removed some of the touched code. Crdlc could you please try to uplift it to v1-train yourself ?
Flags: needinfo?(crdlc)
(In reply to Julien Wajsberg [:julienw] from comment #9) > there will likely be conflicts because of Bug 844990 that removed some of > the touched code. Crdlc could you please try to uplift it to v1-train > yourself ? Yeah, good catch. IMO This is also a regression from 842627 because: The adding class action to wrapper element is also put in windowOpened function, which is delayed until mozbrowserloadend(on ev.me this means wait until all network transmit......) is fired.
Since 844990 already remove the something from 842627, I suggest that we don't need to do this patch :|
It would indeed make sense to revert this and see if Bug 844990 is enough to fix this.
This doesn't happen in current v1-train ;) (with patch of bug 844990, without this bug's patch).
It doesn't happen neither loading wrapper nor closing it? If it is so, it is not needed anymore
Flags: needinfo?(crdlc)
(In reply to crdlc from comment #14) > It doesn't happen neither loading wrapper nor closing it? If it is so, it is > not needed anymore Ya, the wrapper-footer appears/disappears immediately..
(In reply to Alive Kuo [:alive] (PTO during 3/27~4/7) from comment #15) > (In reply to crdlc from comment #14) > > It doesn't happen neither loading wrapper nor closing it? If it is so, it is > > not needed anymore > > Ya, the wrapper-footer appears/disappears immediately.. Wait, the wrapper type app disappears too slow, investigating.
OK, when an app is closing(to homescreen), we will have about ~1sec to wait next-paint event to be fired. This is a general issue in WM, and timdream tells that we have a at most 300ms timer to cancel the event callback if the event is not fired in 300ms "before". So I think this is another regression in recently performance work in Window Manager. I still recommend to back out the change in this bug and set dup to 844990 and open another bug for app closing too slow one. Thanks!
The slow app closing transition is caused by bug 831391's patch.
So what we should do in this bug should be: delay the |removeClass('visible')| like what https://bugzilla.mozilla.org/attachment.cgi?id=723453 does "only".
Sorry, what do you mean? Open other bug only to implement if ('wrapper' in closeFrame.dataset) { + wrapperHeader.classList.remove('visible'); + wrapperFooter.classList.remove('visible'); + } and - -moz-transition-delay: .5s; /* Delay opening a wrapper */ - -moz-transition-property: visibility; true? thanks
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Status: REOPENED → RESOLVED
Closed: 12 years ago12 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: