WHATWG on 2022-10-14

03:07	<sideshowbarker>	Aren’t a `template` element’s attributes just normal attribute nodes of the `template` element — in the same way as for any other element? Or instead are a `template` element’s attributes part of its template contents?
03:12	<Domenic>	The former
03:14	<sideshowbarker>	OK thanks
03:20	<sideshowbarker>	What made me want to confirm is that https://validator.w3.org/nu/parsetree/ shows them as part of the template contents instead. But I guess that’s just a bug in that tool — in that it’s not correctly showing what the validator.nu parser is actually doing. Because otherwise, if it were somehow making the `template` attributes part of the template contents, then the same behavior would be seen in Firefox (since Firefox uses the same parser).
06:13	<annevk>	sideshowbarker: yeah that's a bug, DocumentFragment cannot have attributes
06:17	<sideshowbarker>	OK — thanks
06:18	<sideshowbarker>	hsivonen: r? https://github.com/validator/htmlparser/pull/77
06:20	<hsivonen>	hsivonen: r? https://github.com/validator/htmlparser/pull/77 r+. Thanks.
09:08	<hsivonen>	I'm curious how the spec and WebKit/Blink got the complication of `about:blank?foo` being `about:blank`-like enough for it to have magic behavior as the first URL to be loaded into a browsing context.
09:27	<Domenic>	I'm curious how the spec and WebKit/Blink got the complication of `about:blank?foo` being `about:blank`-like enough for it to have magic behavior as the first URL to be loaded into a browsing context. https://github.com/whatwg/html/pull/6869#discussion_r675520796 is where we first started down that path, I think....
09:30	<Domenic>	https://domenic.github.io/rewrapper/ fails less and has more test coverage (https://github.com/domenic/rewrapper/tree/master/testcases). Maybe one day we can run all of HTML through it.
09:30	<hsivonen>	https://github.com/whatwg/html/pull/6869#discussion_r675520796 is where we first started down that path, I think.... Thanks. That explains the spec history but not how WebKit got the behavior. (Blink presumably inherited from WebKit.)
11:00	<annevk>	hsivonen: WebKit has had a pretty good URL parser for a long term where `?foo` would not be considered part of the path; that might have something to do with it
11:00	<annevk>	(I think Chrome ended up regressing on that, but perhaps not everywhere due to tests?)
11:02	<hsivonen>	hsivonen: WebKit has had a pretty good URL parser for a long term where `?foo` would not be considered part of the path; that might have something to do with it But surely someone has had to write the code where the URL of the initial about:blank gets swapped so that it becomes about:blank?foo in place. That seems like something that should give one a pause while implementing.
11:16	<annevk>	Oh, I misunderstood. That sounds weird.
11:16	<annevk>	https://github.com/WebKit/WebKit/search?q=about%3Ablank%3Ffoo only returns tests.
12:32	<sideshowbarker>	TabAtkins: does the Bikeshed `Spec` object metadata include the last-modified date?
21:03	<TabAtkins>	sideshowbarker: Yeah, `spec.inputSource.mtime()` (for input types that support it)