Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign uptests: Speed up CSS tests by using smaller Japanese font file #11630
Conversation
|
@bors-servo try |
tests: Speed up CSS tests by using smaller Japanese font file <!-- Please describe your changes on the following line: --> Replace the Japanese font with a much smaller font file to speed up relevant CSS tests. --- <!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: --> - [X] `./mach build -d` does not report any errors - [X] `./mach test-tidy` does not report any errors - [X] These changes fix #11452 (github issue number if applicable). <!-- Either: --> - [ ] There are tests for these changes OR - [X] These changes do not require tests because modifying test to swap font <!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. --> <!-- Reviewable:start --> --- This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/11630) <!-- Reviewable:end -->
|
|
highfive
commented
Jun 7, 2016
|
ccb917f
to
66cb220
highfive
commented
Jun 8, 2016
|
New code was committed to pull request. |
|
r? @SimonSapin |
| @@ -20,10 +18,10 @@ | |||
| </head> | |||
| <body> | |||
| <ol> | |||
| <li>い. Gryffindor</li> | |||
This comment has been minimized.
This comment has been minimized.
SimonSapin
Jun 8, 2016
Member
Why this change? I think that’s what’s making ol_japanese_iroha_a.html fail.
This comment has been minimized.
This comment has been minimized.
talklittle
Jun 8, 2016
Author
Contributor
I changed it because the new font was causing the reftest to fail. After this change it no longer fails. I'm guessing the presence of the . is font- and browser-specific. In Firefox it shows a "Japanese comma" instead of a period.
|
I guess the new font is triggering |
|
|
|
@SimonSapin This has been waiting in your queue for a long time now. |
66cb220
to
2c147de
|
Sorry, I got this PR confused with #8374 which was blocked because of test failures.
There is no reason for that to happen. And I’ve verified in a debugger that the code with I’m somewhat opposed to blindly changing a reference file to match a rendering bug in a test case. This is what the test with this PR looks for me, with I don’t know why the latin letters are light grey instead of black. There is no declaration for the During the page load, for a about one second, the rendering looks like this (if I’m quick enough with the screenshot button): Maybe it while the font isn’t loaded yet? The dot is present during that time. @pcwalton any idea what’s going on? |
|
@SimonSapin Probably the font hasn't loaded yet. This would be an excellent use case for the thread profiling tool I'm working on :) |
|
Yes, my second screenshot is when the font hasn’t loaded yet. (This is taken manually, it’s not what the test harness gets.) The thing is why, when the font does eventually load (my first screenshot), do the dot on space disappear? |
|
Oh, that's weird. Does it reproduce with incremental layout off? |
|
With |
|
Tracking the bug this exposed in #15273. Closing because this doesn't appear to be going anywhere. |






talklittle commentedJun 5, 2016
•
edited by larsbergstrom
Replace the Japanese font with a much smaller font file to speed up relevant CSS tests.
./mach build -ddoes not report any errors./mach test-tidydoes not report any errorsThis change is