You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
What problem does this solve or what need does it fill?
We need to be able to verify that Taffy's implementations of layout algorithms (Flexbox, CSS Grid, etc) are correct. Our current gentest system is excellent and allows us to compare to Chrome (and potentially other browsers). However, it only allows us to test features which our text fixtures cover, and while our test suite is reasonably extensive it pales in comparison to the WPT test suite, which is not only larger but is maintained by "tier 1 browser" (Chrome, Firefox, etc) developers and is what they use as the basis for ensuring interoperability
What solution would you like?
Create a WPT runner for Taffy that can run the (a subset of) the WPT tests against Taffy.
https://github.com/jkelleyrtp/stylo-dioxus could likely form a basis for a runner. It has rendering capabilities so it could be used to do "visual similarity" for "ref tests". And while it doesn't currently have scripting capabilities, those could potentially be bolted for running WPT tests.
What problem does this solve or what need does it fill?
We need to be able to verify that Taffy's implementations of layout algorithms (Flexbox, CSS Grid, etc) are correct. Our current gentest system is excellent and allows us to compare to Chrome (and potentially other browsers). However, it only allows us to test features which our text fixtures cover, and while our test suite is reasonably extensive it pales in comparison to the WPT test suite, which is not only larger but is maintained by "tier 1 browser" (Chrome, Firefox, etc) developers and is what they use as the basis for ensuring interoperability
What solution would you like?
Create a WPT runner for Taffy that can run the (a subset of) the WPT tests against Taffy.
https://github.com/jkelleyrtp/stylo-dioxus could likely form a basis for a runner. It has rendering capabilities so it could be used to do "visual similarity" for "ref tests". And while it doesn't currently have scripting capabilities, those could potentially be bolted for running WPT tests.
A good initial target would be to run most of:
What alternative(s) have you considered?
Additional context
The text was updated successfully, but these errors were encountered: