Hi Joey, Ravjit, could you please take a look at this?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Need one more reviewer for third_party/blink/web_tests/VirtualTestSuites, Anders could you please take a looK?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
HTMLUserMediaElementMediaStream::From(*element_).SetMediaStream(stream);
element_->DispatchEvent(*Event::Create(event_type_names::kStream));
}These lines should have less indentation
It isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
HTMLUserMediaElementMediaStream::From(*element_).SetMediaStream(stream);
element_->DispatchEvent(*Event::Create(event_type_names::kStream));
}These lines should have less indentation
Oh, thanks for catching this. I don't know why `git cl format` did not work and the presubmit did not warn about this.
It isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
This is to add `BypassPepcSecurityForTesting` args to the tests. Basically we are mirroring the logic of <geolocation> element (HTMLGeolocationElement), however, the test in <usermedia> contains a "click test" which is disabled by security measures.
I will add a comment about that.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I need to fix the test again. So, remove reviewers and I will add you back
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
HTMLUserMediaElementMediaStream::From(*element_).SetMediaStream(stream);
element_->DispatchEvent(*Event::Create(event_type_names::kStream));
}Thomas NguyenThese lines should have less indentation
Oh, thanks for catching this. I don't know why `git cl format` did not work and the presubmit did not warn about this.
Done
Thomas NguyenIt isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
This is to add `BypassPepcSecurityForTesting` args to the tests. Basically we are mirroring the logic of <geolocation> element (HTMLGeolocationElement), however, the test in <usermedia> contains a "click test" which is disabled by security measures.
I will add a comment about that.
Added comment and moved to the right place.
Thomas NguyenIt isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
Thomas NguyenThis is to add `BypassPepcSecurityForTesting` args to the tests. Basically we are mirroring the logic of <geolocation> element (HTMLGeolocationElement), however, the test in <usermedia> contains a "click test" which is disabled by security measures.
I will add a comment about that.
Added comment and moved to the right place.
Done
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Thomas NguyenIt isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
Thomas NguyenThis is to add `BypassPepcSecurityForTesting` args to the tests. Basically we are mirroring the logic of <geolocation> element (HTMLGeolocationElement), however, the test in <usermedia> contains a "click test" which is disabled by security measures.
I will add a comment about that.
Thomas NguyenAdded comment and moved to the right place.
Done
So is the virtual test suite supposed to be there forever? Could we instead check whether we are running in a web test to bypass the security check?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Thomas NguyenIt isn't obvious to me why this patch adds a virtual test suite. Want to add a paragraph to the commit message explaining why it is added in this patch as opposed to an earlier one which actually added these flags?
Thomas NguyenThis is to add `BypassPepcSecurityForTesting` args to the tests. Basically we are mirroring the logic of <geolocation> element (HTMLGeolocationElement), however, the test in <usermedia> contains a "click test" which is disabled by security measures.
I will add a comment about that.
Thomas NguyenAdded comment and moved to the right place.
Joey ArharDone
So is the virtual test suite supposed to be there forever? Could we instead check whether we are running in a web test to bypass the security check?
We’ll keep the click tests as they are.
and yes, we can check in the native code by allowing WPT-driven clicks (`IsRunningWebTest`. I’m slightly worried about future tests that might need to block synthetic clicks, but since there’s nothing like that in the WPT right now, let's keep it simple. If we need to, we can always bring back the virtual test suites later.
Updated.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
await new Promise(resolve => t.step_timeout(resolve, 600));What exactly is this waiting for? Does it not work if you put 0 instead of 600 in the timeout? Or maybe 0 with a rAF?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
await new Promise(resolve => t.step_timeout(resolve, 600));What exactly is this waiting for? Does it not work if you put 0 instead of 600 in the timeout? Or maybe 0 with a rAF?
In the first 500ms after being added to DOM and rendered on a page, the element is unclickable. So at least it should be 501ms.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
await new Promise(resolve => t.step_timeout(resolve, 600));Thomas NguyenWhat exactly is this waiting for? Does it not work if you put 0 instead of 600 in the timeout? Or maybe 0 with a rAF?
In the first 500ms after being added to DOM and rendered on a page, the element is unclickable. So at least it should be 501ms.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Exportable changes to web-platform-tests were detected in this CL and a pull request in the upstream repo has been made: https://github.com/web-platform-tests/wpt/pull/59276.
When this CL lands, the bot will automatically merge the PR on GitHub if the required GitHub checks pass; otherwise, ecosystem-infra@ team will triage the failures and may contact you.
WPT Export docs:
https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md#Automatic-export-process
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Commit-Queue | +2 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
15 is the latest approved patch-set.
The change was submitted with unreviewed changes in the following files:
```
The name of the file: third_party/blink/web_tests/external/wpt/html/semantics/permission-element/usermedia/set-constraints-combinations.tentative.html
Insertions: 2, Deletions: 2.
@@ -82,10 +82,10 @@
await new Promise(r => t.step_timeout(r, 600));
if (should_trigger) {
- await test_driver.bless("click", () => usermedia.click());
+ await test_driver.click(usermedia);
await stream_promise;
} else {
- await test_driver.bless("click", () => usermedia.click());
+ await test_driver.click(usermedia);
// Wait a bit to ensure no stream event is fired
await new Promise(r => t.step_timeout(r, 200));
}
```
[CE] Correct logic of DefaultEventHandler in HTMLUserMediaElement
Ensure that HTMLUserMediaElement does not initiate multiple concurrent
media stream requests. Previously, rapid user activations could trigger
overlapping requests before the initial one had completed.
Additionally, the activation logic was updated to use the base
capability element's activation handler, ensuring that permission
checks and media requests are correctly synchronized with user events.
Fixed: 500303904
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
The WPT PR for this CL has been merged upstream! https://github.com/web-platform-tests/wpt/pull/59276
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |