detect a non-firefox console and respectfully leave it alone, adding an INFO or WARN message to the console about this. let the user know that: "we have detected a 3rd party window.console API, and any calls to console.info/warn/log/error will not be displayed in the webconsole" or something to that effect.

This bug is split off of bug 583476, which is specifically covering an issue in which the console does not activate on some sites. As a separate issue, this bug addresses the need to tell the user when Firefox's built-in console object has been overridden by JavaScript provided on the page.

For beta 6 I am going to write the method that does the warning part and create the string, as the detection part will have to wait until after string freeze

I would recommend producing that patch as soon as possible...

the warning/error message I have so far for this string is: "The Web Console logging API (console.log, console.info, console.warn, console.error) has been disabled by a script on this page." Sounds good? makes sense? is there a better message?

Part 1 of 2 patches, the second of which will detect the page-created browser. This patch needs to land for b6, the second can land before final.

Does this obsolete the first patch ("patch v1.2") in bug 583476?

(In reply to comment #8) > Does this obsolete the first patch ("patch v1.2") in bug 583476? Not sure. I do think that the problem is not that there is a new console, rather, there are some exceptions that are thrown preventing us from opeing the UI. It may or may no thave to do with the paved over console. Still working on that aspect of it.

Comment on attachment 471985 [details] [diff] [review] v 0.2 Patch >+ logWarningAboutReplacedAPI: >+ function HS_logWarningAboutReplacedAPI(aHUDId) >+ { >+ let domId = "hud-log-node-" + this.sequenceId(); >+ let displayNode = this.getHeadsUpDisplay(aHUDId); >+ let outputNode = displayNode.querySelectorAll(".hud-output-node")[0]; see this a lot. instead of hard-coding the selector everywhere, maybe should make a convenience function this.getHUDOutputNode(aHUDId); r=me otherwise.

Reprioritizing bugs. You can filter the mail on the word TEABAGS.

(In reply to comment #11) > Reprioritizing bugs. You can filter the mail on the word TEABAGS. huh? I do not understand what you are doing here. the first patch is a b6 blocker with strings.

It appears that severity=blocker means something different to other people (see the bug writing guidelines doc) and does not mean the same thing as blocker2.0. At the end of my machinations here, I think I'll have something clearer for the team anyhow.

Mihai has included this change with slightly different wording in bug 583476.

Moving [strings] patches between bugs of different triage status isn't making triage easier, FWIW.

(In reply to comment #14) > Mihai has included this change with slightly different wording in bug 583476. The reason I have been asking mihai not to worry so much about this problem is virtually all of this code will be ripped out next week. The problems he should look into are the exceptions that are raised when trying to display the UI. This is unfair to localizers and is not cool to me as I spent some time on this patch last week - and - it is isolated and easy to land. Enough with the refactoring already. This is not going to affect that many people and is better dealt with when we are not under string freeze.

Passes all tests

David and Axel: I agree that bouncing that string between the bugs is problematic. Mihai is changing bug 583476 to depend on this one for the string. (That said, it looks like the progress on bug 583476 has been good so we can have the console popup on problematic sites in beta 6.)

There's a bug in the latest patch: getConsoleOutputNode: function HS_getConsoleOutputNode(aId) { let displayNode = this.getHeadsUpDisplay(aHUDId); return displayNode.querySelectorAll(".hud-output-node")[0]; }, aHUDId is undefined.

Updated patch to fix the issue reported above.

please file a followup bug for the fix. this patch is already reviewed and approved.

Backed out

Actually, there is one more patch coming from bug 583476

Comment on attachment 473166 [details] [diff] [review] v 0.4 updated with a fix Marking this patch as obsolete, because 583476 will include the fix.

the test for this patch leaked on tinderbox.

Correct. I have a fix for this, in the upcoming patch for bug 583476.