Open jakearchibald opened 4 years ago
This is somewhat contentious and non-interoperable, actually; see https://github.com/whatwg/html/issues/125, https://github.com/whatwg/html/issues/1230, and https://github.com/whatwg/html/pull/4247.
Hah, perfect. I guess we should just do whatever the spec does right now, and change whenever the spec changes.
It will certainly be observable to them that they weren't unloaded. Unlike the load
event, activate
does commit them to going there if it did succeed in loading and so is less useful for passively observing XFO etc than <iframe onload>
.
Only just occurred to me that iframes fire their load event even in the case of navigation 'failures'. If we reject in similar cases we're revealing something new.