| 09:20 | <sideshowbarker> | annevk: About the FrozenArray<Element>? WebIDL design for the various *Elements attributes in ARIAMixin https://w3c.github.io/aria/#ARIAMixin: I’m wondering if you happen to know the history behind that |
| 09:47 | <sideshowbarker> | hmm OK looking at https://github.com/alice/aom/blob/e1d084046dafb9867bab5514c87d8e56f796e239/explainer.md#use-case-2-setting-relationship-properties-without-needing-to-use-idrefs I find this:
|
| 14:18 | <annevk> | sideshowbarker: I think some mixture of premature shipping and the better array type not being quite ready yet. There should certainly be some history captured in whatwg/html issues. |
| 19:59 | <Yagiz Nizipli> | Is there any other browser/runtime that implements URLPattern other than Chrome? |
| 20:03 | <Yagiz Nizipli> | I think there is a WPT that is wrong. If I'm correct, new URLPattern({ "protocol": "http", "port": "80 " }) should not fail, because port setter defined in URL spec trims leading and trailing whitespace characters from the input on port setter. Chrome throws because it doesn't comply with spec... |
| 20:05 | <Yagiz Nizipli> | Any idea why new URLPattern({ "protocol": "http", "port": "80 " }) fails? |
| 20:39 | <Yagiz Nizipli> | It seems the spec doesn't mention port validation, but Chrome is implementing it. https://chromium.googlesource.com/chromium/src/+/main/extensions/common/url_pattern.cc#101 |
| 20:43 | <Noam Rosenthal> | https://urlpattern.spec.whatwg.org/#high-level-operations step 6 seems relevant? |
| 20:44 | <Yagiz Nizipli> | This just sets the processed_init port to empty string. It doesn't validate or throws an error for it. |
| 20:47 | <Yagiz Nizipli> | I've opened an issue for this: https://github.com/whatwg/urlpattern/issues/239 |
| 20:53 | <Yagiz Nizipli> | From my understanding, this is only throwing because Chrome doesn't follow URLPattern spec and call the URL setter. |
| 20:53 | <Yagiz Nizipli> | But I don't understand how this landed into WPT |