Parallelize and wait for endpoint requests before rendering Discovery Stream
Categories
(Firefox :: New Tab Page, defect, P1)
Tracking
()
People
(Reporter: thecount, Assigned: thecount)
References
Details
(Keywords: github-merged)
Attachments
(2 files)
52 bytes,
text/x-github-pull-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
lizzard
:
approval-mozilla-beta+
|
Details | Review |
To deal with loading jank, for now we're going to wait until everything is ready until we render the layout.
Assignee | ||
Updated•6 years ago
|
Comment 1•6 years ago
|
||
Updated•6 years ago
|
Assignee | ||
Comment 2•6 years ago
|
||
For QA I would say just make sure dev-test-all or basic load and nothing looks out of place.
This bug is likely to break in very noticeable and dramatic ways, so if it's displaying content still I think we're good.
Assignee | ||
Updated•6 years ago
|
Updated•6 years ago
|
Updated•6 years ago
|
Comment 3•6 years ago
|
||
Updated•6 years ago
|
Updated•6 years ago
|
Comment 4•6 years ago
|
||
Looks like this depends on Bug 1520258 to be uplifted
Comment 5•6 years ago
|
||
Assignee | ||
Comment 7•6 years ago
|
||
Comment on attachment 9043424 [details]
Bug 1525413 - Parallelize and wait for endpoint requests before rendering Discovery Stream
Beta/Release Uplift Approval Request
Feature/Bug causing the regression
None
User impact if declined
Experiment data less reliable. User performance issues.
Is this code covered by automated tests?
Yes
Has the fix been verified in Nightly?
No
Needs manual test from QE?
Yes
If yes, steps to reproduce
- Set the pref
browser.newtabpage.activity-stream.discoverystream.config
to{"enabled":true,"show_spocs":true,"layout_endpoint":"https://getpocket.com/v3/newtab/layout?version=1&consumer_key=40249-e88c401e1b1f2242d9e441c4&layout_variant=dev-test-all"}
- Load up a new tab.
- Watch for when new tab loads, it loads all in one shot, and not concurrently, ie elements shouldn't load until everything is ready, and load and render once, and not load elements as they come in, causing the page to jump around.
I'm not able to throttle this using responsive design mode tools as responsive design mode does wonky things in about:home, but with the above steps, I have been able to do it fast enough to see that everything gets loaded at once. If you need to try it again, you can change the pref from the above one in step 1 to {"enabled":true,"show_spocs":true,"layout_endpoint":"https://getpocket.com/v3/newtab/layout?version=1&consumer_key=40249-e88c401e1b1f2242d9e441c4&layout_variant=dev-test-all" }
, notice the extra space after dev-test-all" which triggers a fresh load/cache clear.
List of other uplifts needed
Risk to taking this patch
Medium
Why is the change risky/not risky? (and alternatives if risky)
It's preffed off and has tests, but really hard to verify because it relies on network requests..
String changes made/needed
None
Comment 9•6 years ago
|
||
There are 2 different issues found on Mac OS and Windows OS respectively.
Windows :
"dev-test-all" end-point doesn't work on Windows 10 Pro.
Tracking here : bug 1527816
Mac OS :
Layout has some words cutoff specifically for the letter g on some cards (maybe css issue?)
Tracking here : bug 1527819
Comment 10•6 years ago
|
||
Brahmini, it sounds like those issues are unrelated to this patch and don't need to block the uplift.
- For Bug 1527816, we're unable to reproduce this/don't have any reports of nightly not working on Windows 10 – are you sure you have your configuration set up correctly?
- For Bug 1527819, this reproduce-able on nightly but unrelated and non-blocking for this uplift.
Comment 11•6 years ago
|
||
Comment on attachment 9043424 [details]
Bug 1525413 - Parallelize and wait for endpoint requests before rendering Discovery Stream
Let's go ahead with this uplift for beta 8 since the issues found by QA aren't blocking.
Comment 13•6 years ago
|
||
I had no clue about the pref browser.newtabpage.activity-stream.discoverystream.optOut.0
(which was true), so having incorrect value on Windows caused the issue.
After pref reset, it works as expected.
Thanks for the clarification.
![]() |
||
Comment 14•6 years ago
|
||
bugherder uplift |
Comment 15•6 years ago
|
||
Tested bug 1527816.
Closing as verified.
Updated•6 years ago
|
Updated•6 years ago
|
Comment 16•6 years ago
|
||
I have verified that the issue is no longer reproducible on the latest Beta 66.0b8 (Build ID 20190214195736) on Windows 10 x64, Mac 10.14, and Arch Linux 4.14.3.
Updated•6 years ago
|
Description
•