1327615 - Ruleview looses connection to the element if I toggle a rule several times

Status: RESOLVED DUPLICATE of bug 1330607

(DevTools :: Inspector: Rules, defect, P3)

Description

[arni2033]

>>>   My Info:   Win7_64, Nightly 49, 32bit, ID 20160526082509
STR_1:
1. Open https://xcraft.ru/colonization.html?sp=9
2. Open inspector -> ruleview. Inspect the left image that depicts a spaceship (selector "a.terran")
3. Keep clicking on checkbox near the CSS rule defining background for 5 seconds

AR:  Either A or B happens
 A) The actual applied rules stop updating (image is constantly displayed on the page)
    Obviously, ruleview just looses 'connection' to the node, as the rule displayed in style editor is
   `background: url("...") top center no-repeat;`
 B) The actual applied rules stop updating (image is constantly NOT displayed on the page)
    Obviously, ruleview just looses 'connection' to the node, as the rule displayed in style editor is
   `/*! /*! background: url("...") top center no-repeat; */ */`

ER:
 Checkbox near rule in ruleview should always toggle the rule correctly.
 Double commented rules (as shown in (B)) should never appear in "style" attr/style editor.

   
Notes:
1) (B) is worse than (A), because if I switch to another node and back, (A) seems fixed (i.e. ruleview
   is 'connected' to the node again), but (B) is still broken, because it defines an invalid rule, so
   the only rule displayed in "element { }" section is `*/: ;`, and it is invalid
2) If browser lags (becomes too slow due to high RAM/CPU consumption), I don't need to click too many
   times in Step 4 - that's how I originally encountered it. You can also write a simple content
   script which will quickly click on the checkbox.


STR_2:  (testcase)
1. Open url   data:text/html,<div style="background: lightgray;">A<style>div{background:gray}
2. Open devtools -> inspector -> ruleview. Inspect <div> element
3. Keep clicking on the rule "background: lightgray;"


STR_2 is regression from bug 1069829. STR_1 is not.

Comment 1

[Tom Tromey :tromey]

I could not reproduce this with a build from today.

Comment 2

[arni2033]

>>>   My Info:   Win7_64, Nightly 53, 32bit, ID 20161229030206 (2016-12-29)
Reproducible.
To run into this bug, you should either click really fast, or have FF instance that lags (so clicking at normal speed will be equal to very fast clicking for the browser).

Comment 3

[Tom Tromey :tromey]

> To run into this bug, you should either click really fast, or have FF
> instance that lags (so clicking at normal speed will be equal to very fast
> clicking for the browser).

Yeah, that makes sense.
Also, are you using e10s?

Comment 4

[Patrick Brosset <:pbro>]

Not able to reproduce this but it sounds very possible that there would be a race condition here that, on slower machines, would cause this.

Comment 5

[Tom Tromey :tromey]

Dup.