Re: [html-tests] [html] Account for bug that causes instability (#4828)

> Is it possible to write a test that reliably fails if the bug is present?

That's the intent of this patch. The initial version caused a timeout in Chrome
because it didn't account for variations in the `textDocoration` computed
property (which differs between Chromium and Chrome). I've updated the load
detection mechanism to use the `zIndex` property instead.

As you can see in the stability test results above, Chrome now consistently
fails while Firefox continues to pass.

View on GitHub: https://github.com/w3c/web-platform-tests/pull/4828#issuecomment-280060212

Received on Wednesday, 15 February 2017 16:28:35 UTC