Whether this means firing DOMContentLoaded on error pages, or inventing some new event that is error-page-specific, it would be extremely helpful to be able to listen for error page loads. It would help with bugs like bug 425001, but also with things like bug 428948
Created attachment 618640 [details] [diff] [review] proposal v1 I'm not yet sure this is the right fix for bug 623155, but, in general, it would be useful. >+ * |aWebProgress| does NOT set this flag when it did not try to load a new >+ * contents. In this case, it should set LOCATION_CHANGE_SAME_DOCUMENT. If this behavior is not allowed, probably docshell needs another member variable to store this flag.
Comment on attachment 618640 [details] [diff] [review] proposal v1 This needs tests.
Created attachment 619197 [details] [diff] [review] Test Test, appending to bug 311007's. BTW, during writing the test, I found that my comment for the idl > Both |aURI| and |aRequest| may be null. is a lie. Always aURI is not null. However it can be converted from null to <about:blank> in a very edge case, though I don't know how to reproduce.
Comment on attachment 618640 [details] [diff] [review] proposal v1 sr=me
Created attachment 619290 [details] [diff] [review] proposal v1.1 Fixed IDL comments.
https://hg.mozilla.org/mozilla-central/rev/4f05a7ffcfb3 https://hg.mozilla.org/mozilla-central/rev/d8c356534c6a \o/