Josh Matthews

Ok, I've read some code and here are my thoughts: * there is a clear path for the GC to trace the fields of Rust structs that implement the Callback...

No response from assignee, so this is available again.

Let's use: ``` /// Types that implement the `Callback` trait follow the same rooting requirements as types that use the `#[dom_struct]` attribute. /// Prefer storing `Dom` members inside them instead...

It is safe to switch from DomRoot to Dom for that struct. It will require adding an annotation like https://github.com/servo/servo/blob/99add66f8e463c6b8f42fa000231e16c3727c025/components/script/body.rs#L270 to allow it to compile with `--use-crown`.

Closing in favour of #35989

https://github.com/servo/servo/blob/99ca170ed7eb4bee8100c5bcd8ef7ba59629eb02/components/script/stylesheet_loader.rs#L168-L191

Step 1 of https://html.spec.whatwg.org/multipage/links.html#link-type-stylesheet:process-the-linked-resource is pretty clearly in favour of our current implementation. I wonder if MIME sniffing is supposed to kick in here? Unfortunately, https://mimesniff.spec.whatwg.org/#rules-for-sniffing-in-a-style-context is not enlightening at...

Mmm, we may be missing this logic: https://html.spec.whatwg.org/multipage/semantics.html#processing-the-type-attribute

That still doesn't explain things because the page's HTML is: ``

This specification logic is so convoluted. From the example from the last link: >If one of the two files was returned without a [Content-Type](https://html.spec.whatwg.org/multipage/urls-and-fetching.html#content-type) metadata, or with a syntactically incorrect...