Closed Bug 1788115 Opened 2 years ago Closed 2 years ago

WPT test toascii.window.js (aka toascii.window.html ) has some failures in Firefox

Categories

(Core :: DOM: Networking, defect, P2)

defect

Tracking

()

RESOLVED FIXED
106 Branch
Tracking Status
firefox106 --- fixed

People

(Reporter: dholbert, Assigned: valentin)

References

(Blocks 2 open bugs)

Details

(Whiteboard: [necko-triaged])

Attachments

(2 files)

I was just glancing at https://wpt.fyi/interop-2022 , and particularly the "Web Compat" focus area there (where we happen to be leading but still have some failures).

It turns out all of our "Web Compat" category failures are in one test, shown here:
https://wpt.fyi/results/?label=experimental&label=master&product=chrome&product=firefox&product=safari&aligned&view=interop&q=label%3Ainterop-2022-webcompat

wpt.fyi link for test: https://wpt.fyi/results/url/toascii.window.html
"Run in your browser" link: https://wpt.live/url/toascii.window.html

Test results:
Firefox: 262/352
...vs. Chromium: 198/352, and Safari: 352/352 (100%)

Our failures seem to all be in subtests that are related to handling punycode.

The in-tree files for this test are:
https://searchfox.org/mozilla-central/source/testing/web-platform/tests/url/toascii.window.js
https://searchfox.org/mozilla-central/source/testing/web-platform/tests/url/resources/toascii.json
...and known-failure annotations at:
https://searchfox.org/mozilla-central/source/testing/web-platform/meta/url/toascii.window.js.ini

Looks like this test was added to the test set in https://github.com/web-platform-tests/interop/issues/18 FWIW (filed in Nov 2021).

In the initial comment there, AnneVK (representing Mozilla at the time, though he's now at Apple) filed the issue to highlight some Chrome-specific subtest failures as being the real compat issue identified by that test ("Firefox and Safari follow the URL Standard and use Transitional_Processing set to false, but Chrome does not.")

He noted "Firefox has some unrelated failures on this particular test, I'm not sure whether those should be in scope or not." Presumably those are still the failures that we have today that are (slightly) dinging us in the interop2022 tally.

Assignee: nobody → valentin.gosu
Blocks: url
Severity: -- → S3
Priority: -- → P2
Whiteboard: [necko-triaged]

According to https://unicode.org/reports/tr46/#ToASCII we have the option of
setting VerifyDnsLength = false for the algorithm. This seems in accordance
to the URL spec and the rest the browser implementations.

The label size is arbitrarily bumped up to 256.

Pushed by valentin.gosu@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/ef8a29a25db2
Do not enforce 63 character limit for IDNA labels r=necko-reviewers,dragana
https://hg.mozilla.org/integration/autoland/rev/2d8d235bd750
Add length check to EmailInputType::PunycodeEncodeEmailAddress r=smaug

Backed out for causing wpt failures on failure.html

Backout link

Push with failures

Failure log

Failure line: PROCESS-CRASH | /url/failure.html | application crashed [@ mozilla::ipc::FatalError(char const*, bool)]

Flags: needinfo?(valentin.gosu)
Pushed by valentin.gosu@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/6a46724141de
Do not enforce 63 character limit for IDNA labels r=necko-reviewers,dragana
https://hg.mozilla.org/integration/autoland/rev/58bd7ec59592
Add length check to EmailInputType::PunycodeEncodeEmailAddress r=smaug
Status: NEW → RESOLVED
Closed: 2 years ago
Resolution: --- → FIXED
Target Milestone: --- → 106 Branch
Flags: needinfo?(valentin.gosu)

(In reply to Daniel Holbert [:dholbert] from comment #0)

I was just glancing at https://wpt.fyi/interop-2022 , and particularly the "Web Compat" focus area there (where we happen to be leading but still have some failures).

It turns out all of our "Web Compat" category failures are in one test, shown here:

With these fixes in, https://wpt.fyi/interop-2022 now shows us at 100% score for the "Web Compat" category. \o/

Thanks for taking this, Valentin!

Regressions: 1793521
Regressions: 1790287
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: