- From: Marijn Kruisselbrink <web-platform-tests-notifications@w3.org>
- Date: Thu, 05 Jan 2017 17:10:39 GMT
- To: public-web-platform-tests-notifications@w3.org
In general chrome doesn't actually use console output comparison for testharness.js based tests, but instead just directly checks if the test passed/failed. Which works fine for most cases. But if a particular file has multiple tests, most of which pass, but some of which fail, we have a few choices (with the current system): either mark the entire file as failing, losing any useful test coverage from the file, or falling back to comparing the output. So what chrome's system really needs here of course is a way to mark particular sub-tests as failing, rather than only having test expectations at the file level. View on GitHub: https://github.com/w3c/web-platform-tests/pull/4378#issuecomment-270699566
Received on Thursday, 5 January 2017 17:10:51 UTC