revert: manual management of detached WebContents #19011
Conversation
| @@ -460,7 +455,7 @@ WebContents::~WebContents() { | |||
| } else { | |||
| // Destroy WebContents asynchronously unless app is shutting down, | |||
| // because destroy() might be called inside WebContents's event handler. | |||
| DestroyWebContents(true /* async */); | |||
| DestroyWebContents(!IsGuest() /* async */); | |||
There was a problem hiding this comment.
I'm not sure !IsGuest() is what we expect here. @zcbenz opinion?
|
@deepak1556 you were suggesting to add tests: what kind of tests would you suggest? |
| @@ -460,7 +455,7 @@ WebContents::~WebContents() { | |||
| } else { | |||
| // Destroy WebContents asynchronously unless app is shutting down, | |||
| // because destroy() might be called inside WebContents's event handler. | |||
| DestroyWebContents(true /* async */); | |||
| DestroyWebContents(!IsGuest() /* async */); | |||
I couldn't find a way to programmably test the original issue, it seemed to be only visually noticeable. |
|
@alexstrat can you rebase this PR? |
|
@jkleinsc rebase done |
|
This PR needs an approval from @electron/wg-upgrades. |
|
@alexstrat sorry about this but can you rebase again on master to pull in #19127 that should fix the macOS build error. Thanks! |
|
@deepak1556 done |
codebytere
changed the title
|
macOS failure is expected, as code-signing tests fail on PRs from forks. |
|
Release Notes Persisted
|
Description of Change
Following discussions in #18976, revert #14487
This CL that landed with 76.0.3783.1 made #14487 not useful anymore.
I did follow the reproduction steps explained in #14211 and could not reproduce the issue.
cc @zcbenz @deepak1556
Checklist
npm testpassesRelease Notes
Notes: Revert #14487