Mason Freed

> Seems extremely chilish for adults that work for a company as large as Google to keep locking threads where there are comments criticizing this push to drop XSLT, such...

> I do apologise if it was not actually you or any other user from Google who locked those threads It wasn't me, but again, I appreciate that they did...

> We should add the interface to the Historical section. That would more clearly indicate its status. Done. > Looks good otherwise. Thanks! > Yeah, or else — based on...

At least in Chrome, the behavior as of the latest Canary is "correct": you can drag the dialog and the tooltip stays anchored, even with a transformed dialog. In 142...

> What happens if appearance changes while the picker's open? That’s something that came up in discussion last week. To avoid possible oscillation or hysteresis, if the `appearance` property changes...

> This would also require `showPicker()` to perform some kind of style resolution, which seems suboptimal. That'll be required anyway, for ``, to decide which picker to show.

> No, that doesn't require it to be done synchronously as far as I can tell. Can you elaborate? It seems like the picker should be opened synchronously or nearly-sync,...

> But taking a step back I don't really like the idea that CSS can be used to toggle API behavior in such a fundamental way. That seems like a...

It'll be really unfortunate if the whole point of the new customizable-`` work was to improve the status quo, and then the pushback to improving things is always "well, we...

> You cannot use testdriver with reftests? That seems like something that ought to be fixed if true. Not at the moment, see https://github.com/web-platform-tests/wpt/issues/13183. I agree we should get it...