One can't use RemotePage:Init's .browser.currentURI.spec because it's about:blank. Waiting for RemotePage:Load might be too late. RemotePageManager knows what url content is loading, so provide it.

Comment on attachment 8906809 [details] Bug 1398955 - Include url with RemotePage messages. https://reviewboard.mozilla.org/r/178536/#review183554

The one interesting failure on treeherder is on Windows 7: https://treeherder.mozilla.org/#/jobs?repo=try&revision=e425eac76d86&selectedJob=130206406 It's one of the new tests added in this patch. However, it's not obvious to me whether the failure indicates that something is actually wrong. Is it possible that browser.currentURI on windows 7 simply has a different default value in this case?

23:19:33 ERROR - 915 INFO TEST-UNEXPECTED-FAIL | toolkit/modules/tests/browser/browser_RemotePageManager.js | inited port browser still is blank - Got http://www.example.com/browser/toolkit/modules/tests/browser/testremotepagemanager.html, expected about:blank is the actual error message.

Another way of asking this is that it's not obvious to me why that pre-validation step needs to be added to the test at all. Mardak, can you clarify the thinking behind it?

Yeah, I figured some platforms could potentially have different INIT timings, so I pushed to try before autolanding. I was thinking maybe the test should check for either TEST_URL or about:blank as it shouldn't be something /else/…

Pushed by edilee@gmail.com: https://hg.mozilla.org/integration/autoland/rev/dad72fdc1cbf Include url with RemotePage messages. r=mossop