05:30
<Yutaka Hirano>
annevk: Looking forward to your feedback on https://github.com/whatwg/fetch/issues/1438#issuecomment-1150755587.
07:57
<annevk>
Yutaka Hirano: sounds like a good plan; I can look at the PR later today
08:20
<Yutaka Hirano>
annevk: Thanks!
10:42
<Yoav Weiss>
sefeng: The spec's "timing-allow check" link should probably point directly to https://fetch.spec.whatwg.org/#concept-tao-check
10:42
<Yoav Weiss>
which does take redirects into account
10:42
<Yoav Weiss>
if the Chromium implementation doesn't, can you please file a bug? (Or land a WPT that would fail once we roll it in)
10:48
<Yoav Weiss>
sefeng: The comment you're pointing at in the WPT doesn't say that TAO for the redirects is ignored, just that the final resource is not impacting TAO passes or not, because it does
14:54
<wanderview>
annevk: regarding the struct on environment that would contain the various bits of StorageKey... do you have a preference on the name of the struct? what information do you see living in there besides nonce and cross-site-ancestor?
16:11
<annevk>
wanderview: top-level origin and origin, although I'm a little torn on origin given that it's still somewhat inconclusive what networking should do
16:11
<annevk>
But maybe it's okay enough that networking just ignores that part when creating its key
16:12
<annevk>
Given all of that I was thinking "authority" would be fitting and we'd ignore parts of it in specific contexts, but I'm definitely open to other ideas
16:18
<annevk>
Luca Casonato: hey, for full duplex, the tests you pointed to, does the server there support full duplex?
16:36
<annevk>
I added some thoughts to the full duplex issue directly, I think it's worth looking into, but I wonder if the existing setup already got us stuck with half duplex, e.g., due to the blobs
18:36
<Luca Casonato>
Luca Casonato: hey, for full duplex, the tests you pointed to, does the server there support full duplex?
If it is the Deno.dev one, yes