| 17:14 | <Noam Rosenthal> | Alexander Kalenik: good catch, I think the conclusion you and Dominic Farolino reached is correct. One of my commits may have caused this (when I refactored navigation redirects to use the manual redirect steps) |
| 17:30 | <bkardell: holidays 4/10> | It seems there are quite a lot of cases where getComputedStyle(.slottedThing).direction will give one answer but will match the opposite on the slottedThing:dir selector in both Chrome and WebKit... I mean, that should never be the case, right? Or am I misunderstanding why we'd want that to be the case? |
| 17:34 | <bkardell: holidays 4/10> | I'm trying to improve a lot of things about my own understanding about this as I am trying to come up with the spec text but then I write a bunch of tests and find "mixed" answers like that - I'm not sure why... feels like a bug, but I'd like to be certain. Also, assuming there should be only 1 answer - very curious in those cases which they meant it to be |
| 17:35 | <bkardell: holidays 4/10> | https://codepen.io/bkardell/pen/rNqaVQe. contains a few examples... |
| 17:36 | <bkardell: holidays 4/10> | #5 and #7 have different answers like that in the same browser for chrome/webkit |
| 18:06 | <bkardell: holidays 4/10> | interesting they both have the same discrepancy on those two - they computed style direction is ltr, but the style matches :dir(rtl) |