Mirko Brodesser

@annevk can you PTAL again.

> It would help to have a pointer where the throwing is defined. I didn't see it in the linked issue either. Indeed not obvious. The throwing stems from: https://github.com/whatwg/dom/pull/1247...

> Thanks, this looks like the correct test. > > I wonder if we should also test the inverse. That when you take an element out of a TT global,...

> It's relevant for ensuring the specification is implemented correctly. True; but is that scenario actually relevant in practice?

> That does not matter for conformance tests. The whole point is that we can't know what websites might do and might rely on. Agreed. Added https://github.com/w3c/trusted-types/issues/425#issuecomment-2056452603 so that another...

@annevk: can you please merge this test, I lack rights.

A fatal error would be better than the current state, since especially in headless mode, one might not notice that release instead of Nightly was used.

> I wonder how intrusive it would be to make the WPT helpers trusted type compliant? Don't know. Presumably the only way to find out is to do it. @otherdaniel...

> Yeah that's the trade off, benefit it's less intrusive to the rest of the suite, it wouldn't need to be the whole harness like just the helpers we use...

> You should be able to use `fetch_tests_from_window`, compare for example https://github.com/web-platform-tests/wpt/tree/master/FileAPI/url/resources/fetch-tests.js Okay, that's https://searchfox.org/mozilla-central/rev/9bb5d5f55da6cc7163dd93825c25609b769822f9/testing/web-platform/tests/resources/testharness.js#3886-3897 and should work.