Bug 177358 – Speedometer: ensure all TodoMVC tests use the complete latest CSS

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Mathias Bynens]

The following tests get the default unstyled checkboxes in Firefox:

- VanillaJS-TodoMVC
- Vanilla-ES2015-TodoMVC
- Vanilla-ES2015-Babel-Webpack-TodoMVC
- EmberJS-TodoMVC
- EmberJS-Debug-TodoMVC
- AngularJS-TodoMVC
- Angular2-TypeScript-TodoMVC
- VueJS-TodoMVC
- jQuery-TodoMVC
- Inferno-TodoMVC
- Elm-TodoMVC
- Flight-TodoMVC

This can be fixed by ensuring they use the latest available TodoMVC CSS.

[Mathias Bynens]

Created attachment 321583 [details]
Patch

[Addy Osmani]

Ensuring all implementations use the exact same set of styles cross-browser sgtm. Thanks for catching this, Mathias.

[Ryosuke Niwa]

Any comparison of perf across browsers before/after this content change?

[Build Bot]

Comment on attachment 321583 [details]
Patch

Attachment 321583 [details] did not pass ios-sim-ews (ios-simulator-wk2):
Output: http://webkit-queues.webkit.org/results/4632004

New failing tests:
imported/w3c/web-platform-tests/payment-request/payment-request-show-method.https.html

[Build Bot]

Created attachment 321602 [details]
Archive of layout-test-results from ews125 for ios-simulator-wk2

The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews125  Port: ios-simulator-wk2  Platform: Mac OS X 10.12.6

[Mathias Bynens]

(In reply to Ryosuke Niwa from comment #3)
> Any comparison of perf across browsers before/after this content change?

Unfortunately, I haven’t been able to compare the results yet.

Can I assume the failing build is a fluke?

[Ryosuke Niwa]

Yeah, I don't think this can cause a test failure on iOS.

[Mathias Bynens]

Before/after results of the entire Speedometer benchmark:

Chrome Canary v63.0.3222.0:
    63.55 ± 0.62 (0.98%)  →  64.24 ± 0.39 (0.61%)
Safari Technology Preview Release 37 (v11.1, WebKit 12605.1.4.0.1):
    73.23 ± 0.62 (0.85%)  →  73.26 ± 0.66 (0.90%)
Firefox Nightly v58.0a1 (2017-09-24) (64-bit):
    48.60 ± 0.94 (1.90%)  →  48.1 ± 0.62 (1.3%)

Turns out the rendering difference didn’t cause much of a difference. Still, it’d be good to fix it.

[Mathias Bynens]

(In reply to Mathias Bynens from comment #8)
> Safari Technology Preview Release 37 (v11.1, WebKit 12605.1.4.0.1):

That should’ve said Safari Technology Preview Release *39*. The detailed version numbers are correct.

[WebKit Commit Bot]

Comment on attachment 321583 [details]
Patch

Clearing flags on attachment: 321583

Committed r222534: <http://trac.webkit.org/changeset/222534>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/34692999>