WHATWG on 2022-11-01

00:50	<Domenic>	Patrick Meenan: we broke "ancestor browsing contexts" intentionally in https://github.com/whatwg/html/commit/0a97a81da77bc4cb0ab5b16420605ca001ff5b17 . What spec is this? We will work to update it.
07:13	<annevk>	Domenic: Fullscreen is broken then too
07:13	<Domenic>	I should be able to squeeze out a fullscreen fix in the next hour or so
07:45	<Domenic>	Wasn't so bad: https://github.com/whatwg/fullscreen/pull/208 but needs some exports from HTML.
10:48	<Jerry Murray>	👙❤️🚝🧛🧛‍♂️🧛‍♀️🧛🦸🥷👩‍🚀🧑‍🚒👍️
13:41	<foolip>	How do we use the "needs implementer interest" label? Do we remove it when there's one implementer interested, or two?
14:01	<annevk>	foolip: two
14:53	<Patrick Meenan>	Patrick Meenan: we broke "ancestor browsing contexts" intentionally in https://github.com/whatwg/html/commit/0a97a81da77bc4cb0ab5b16420605ca001ff5b17 . What spec is this? We will work to update it. This was from the fetch spec. PR showing the same issue here: https://github.com/whatwg/fetch/pull/1523
14:55	<Patrick Meenan>	foolip: two Should the implementers comment on the issues in the relevant specs? Chrome has shipped Priority Hints and Mozilla is working on implementing it and there are issues on both the HTML and Fetch trackers (PR's inbound). Just need to know who needs to comment where to properly reflect the implementors interest.
14:57	<foolip>	Patrick Meenan: if implementation is in progress in Gecko, then cc'ing the person working on that on the two issues is a good first step. Ask if they'd be supportive of the changes.
14:57	<jgraham>	I might be imagining things, but did someone at some point make a graphviz-type visualisation of the fetch spec?
15:02	<jgraham>	Hmm, there's https://docs.google.com/presentation/d/1eUtpVsuoPPPLzl6ccNoaUH9Rwmd63hgCboVh-gDLkiI/edit#slide=id.g2c389e7de3_0_17 but that's now 5 years old
15:16	<annevk>	If there's an easy way to generate such a diagram and we can be sure it can be updated 10 years later I'm all ears
15:20	<jgraham>	Hmm, well in theory bikeshed has enough information to generate it. Maybe one could even hack something together with the HTML output, because at some level it's just a graph of links, but limited to the ones that only point to internal algorithms.
15:33	<annevk>	I mean I'm happy to duplicate work, just need some kind of tool that helps with it, makes it accessible, and makes updates somewhat easy.
23:18	<Domenic>	I feel like the hard part of these things is placing the boxes. Every tool I've tried (including implementing one algorithm I found in a paper by hand) has done poorly.