09:10
<sideshowbarker>

About the Referer with location.replace or location.assign thing: As detailed in my comment at https://github.com/WebKit/WebKit/pull/19093#discussion_r1378090398, testing (from a new test I wrote) indicates that Gecko and Blink set the referrer from the parent frame.

I still don’t clearly know what the particular relevant spec requirement is, so I updated the commit message to remove any claim that the change makes the implementation spec-conformant — and instead just kept the part of the commit message which states it’s for interop/compat with Gecko and Blink.

I need to ping Alex to re-review it again, but I otherwise have no remaining changes of my own to make. If the goal is to have interop on this behavior across all engines, then I think the state of the implementation in the PR branch is what’s necessary for doing that.

11:05
<Seirdy>

idk if this is the best place to share this, but I wrote a thing:

Proposal: an HTML element for spoilers

opened an issue in whatwg/html, hope this gains interest
11:46
<annevk>
krosylight: in https://bugzilla.mozilla.org/show_bug.cgi?id=1864395 is type=number a typo?
11:47
<krosylight>
krosylight: in https://bugzilla.mozilla.org/show_bug.cgi?id=1864395 is type=number a typo?
No, but I guess I shouldn't do that
11:48
<krosylight>
0009 != 9 😛
11:48
<annevk>
type=number is so attractive for identifiers, I also saw that mistake with device-eid
11:49
<annevk>
Perhaps inputmode=numeric needs a more attractive shorthand, but pretty hard to beat type
11:51
<krosylight>
inputmode=numeric also doesn't do anything on desktop, hmm