1381383 - facebook_ail_click_photo_viewer_right_arrow is too slow

Status: RESOLVED WORKSFORME

(Core :: CSS Parsing and Computation, defect, P3)

Description

[MarkYen]

STR:
0. Logged in with Whitehat account, go to Linda's photo album
1. Click right-arrow button on photo to next photo

Short Gecko Profile(cover only Input Lantency action):
https://perfht.ml/2tZOHOE
on Ref Device

Reports: https://goo.gl/DjIwNb
Notes: https://docs.google.com/spreadsheets/d/1RgiyTm41Af-mC0lVkqiDT1sp9yKNkZ5cBoj7bApZWtY/edit?ts=596c1cbc#gid=256706414

Comment 1

[Bobby Chien]

Paul, could you help to have the first analysis? Thanks.

Comment 2

[Paul Adenot (:padenot)]

Bobby, what are the details for this account?

Comment 3

[Bobby Chien]

Mark, please help Paul to access Hasal whitehat account. thanks.

Paul, Facebook provides whitehat account in each FB account. You could either use your own whitehat, or use Hasal whitehat. The differences are the contents, like news feed, etc. Please ask Mark anyting you need in measurement and profiling. Thanks.

Comment 4

[MarkYen]

(In reply to Paul Adenot (:padenot) from comment #2)
> Bobby, what are the details for this account?

What details would you like to know more about?

Comment 5

[Paul Adenot (:padenot)]

Well, the login/passwd or whatever I need to have a look at the page and guess what is happening here.

Comment 6

[MarkYen]

Attachment: sample video for this test case

Files contain the video recordings of Facebook photoviewer test case on window10.

Comment 7

[MarkYen]

(In reply to Paul Adenot (:padenot) from comment #5)
> Well, the login/passwd or whatever I need to have a look at the page and
> guess what is happening here.

I have uploaded recordings of this testing case. You can check the steps and account details in the video. However, the videos were not recorded on the exact same reference device, so there might be some differences. But it still provides you with a insight of our test case, since the account and steps are the same.

Comment 8

[Paul Adenot (:padenot)]

What I see is a bunch of js code being executed in the child. I'm not expert, but it looks like we interpret some of it, and it runs off a mousemove event. This seem to spam the parent, the parent its therefore less responsive.

Where do we go from here ?

Comment 9

[MarkYen]

(In reply to Paul Adenot (:padenot) from comment #8)
> What I see is a bunch of js code being executed in the child. I'm not
> expert, but it looks like we interpret some of it, and it runs off a
> mousemove event. This seem to spam the parent, the parent its therefore less
> responsive.
> 
> Where do we go from here ?

What if we check whether there are any unnecessary events and remove them? Will it help solve this situation.

Comment 10

[Paul Adenot (:padenot)]

(In reply to myen from comment #9)
> (In reply to Paul Adenot (:padenot) from comment #8)
> > What I see is a bunch of js code being executed in the child. I'm not
> > expert, but it looks like we interpret some of it, and it runs off a
> > mousemove event. This seem to spam the parent, the parent its therefore less
> > responsive.
> > 
> > Where do we go from here ?
> 
> What if we check whether there are any unnecessary events and remove them?
> Will it help solve this situation.

This is probably not web-compatible.

Comment 11

[Bobby Chien]

Sean, could you help to follow profile analysis from Paul? Thanks.

Comment 12

[Bobby Chien]

Change component to Layout per depends bug 1388179.

Comment 13

[Daniel Holbert [:dholbert]]

(In reply to Bobby Chien from comment #12)
> Change component to Layout per depends bug 1388179.

That bug's looking like it's mostly a style-system issue (per bug 1388179 comment 4), so bumping this to CSS.

Comment 14

[BugBot [:suhaib / :marco/ :calixte]]

The bug assignee didn't login in Bugzilla in the last 7 months.
:emilio, could you have a look please?
For more information, please visit auto_nag documentation.

Comment 15

[Emilio Cobos Álvarez [:emilio]]

This would need new profiles and repro steps.

Comment 16

[Mayank Bansal]

Should this bug be closed?

Comment 17

[Mayank Bansal]

All dependencies are closed, and looks like the website code has changed.
I will close this bug, but Please reopen if needed.