[INVALID] NS_ERROR_FAILURE → no more Flash :(

[dhouwn]

After viewing some videos of Youtube on Firefox 11 Firefox 12 (Aurora) with NoScript 2.3.1rc1, I suddenly can't run any Flash content any-more (no matter if as a document or embedded, other plugins seem to work). I click on the placeholder and it disappears but the plugin seems to not run (plugin-instance.exe is not even started).

Probably related error messages:

Code: Select all

Error: [Exception... "Failure"  nsresult: "0x80004005 (NS_ERROR_FAILURE)"  location: "JS frame :: jar:file:///C:/Program%20Files%20(x86)/firefox/extensions/%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js :: <TOP_LEVEL> :: line 4885"  data: no]
Source File: jar:file:///C:/Program%20Files%20(x86)/firefox/extensions/%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js
Line: 4885

Code: Select all

Error: [Exception... "Failure"  nsresult: "0x80004005 (NS_ERROR_FAILURE)"  location: "JS frame :: jar:file:///C:/Program%20Files%20(x86)/firefox/extensions/%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js :: <TOP_LEVEL> :: line 3716"  data: no]
Source File: jar:file:///C:/Program%20Files%20(x86)/firefox/extensions/%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js
Line: 3716

Other information: Latest Flash beta, also might have been that I killed the plugin process before, which I sometimes do because sometimes it hogs an absurd amount of memory for no apparent reason.

If you want to test me something, I am still running this Firefox instance for now.

P.S.: Interestingly, the icon also seems not fit to deal with such a situation: It still displays a puzzle tile after I click the placeholder but the puzzle tile disappears after I switch to another tab and then back again (you determine the icon state newly whenever one switches to a tab?). This it not the case for standalone plugin content where it disappears right away (What makes the difference? After all it's still embedded in a document, is it the "embed" element?).

[therube]

What is your build date on 11?
Isn't it going to be old, cause Aurora would be up to 12 ever since 10 was released?
(this stuff has always & always will confuse me)

What beta of Flash are you running?

[dhouwn]

What is your build date on 11?

Oooops, got myself confused there, it's actually Firefox 12 (aurora) I am running there, the build from yesterday (http://hg.mozilla.org/releases/mozilla- ... 788bc83248).
I am using Firefox 11 (beta) too but with another profile.

(this stuff has always & always will confuse me)

Since Firefox 10 released (actually some days later), Firefox 11 moved from the aurora to the beta channel, where it is even unlikelier (compared to aurora) that patches get applied to, much less feature-adding or string-changing patches. It's not outdated in terms that this is more or less what the general user population will get 6 weeks later.

What beta of Flash are you running?

11.2.202.197 (Adobe Flash Player 11.2 beta 5), the latest from http://labs.adobe.com/technologies/flas ... layer11-2/

And just in case, my other active add-ons (all at their newest version):

• NoRedirect
• NoSquint
• TabSubmit
• Vertical Tabs

[Giorgio Maone]

This surely seems a bug either in Gecko's plugin host or in the Flash plugin.
Could you try finding regression windows, if any, for both?

[dhouwn]

Should have clarified that this bug is not permanent across Firefox instances, after a restart Flash works again, no console message. For finding a regression window I guess I would first need to reliably reproduce it. I believe to remember that I had this bug at least once before but that could have been a month or two ago.

BTW, I would have really love to know whether I was right with my assumption that

you determine the icon state newly whenever one switches to a tab?

[Giorgio Maone]

BTW, I would have really love to know whether I was right with my assumption that

you determine the icon state newly whenever one switches to a tab?

[/quote]
I do, but the DOM analysis is performed only on page completion and whenever an object instance attempts to load: the results are cached for performance reasons.

[GµårÐïåñ]

This is an actual bug in Gecko and is known on Bugzilla. They were supposed to have patched it on the nightly but I guess it took a crap again. I will try and find you the bug report so you can read through it and if you have anything further to add, do it there where they can try and fix it. Its not related to NS at all.

[GµårÐïåñ]

Here you go, Bug# 723154 and there are dependent issues and cross related issues too that you can get onto from there.

[dhouwn]

I believe the bug you linked to is one of the afterpains of fixing bug 90268 (as you can see it blocks it) which did not land in Firefox 12.

[GµårÐïåñ]

That bug although claiming to have fixed the issue has shown up to still be an issue and shows there is a regression to earlier versions. Including the 10.x that I am running. I am just pointing to issues they can review for a fix. The issue is not NoScript related, so that's the point.

[dhouwn]

And again I am in this error state:

Code: Select all

Error: [Exception... "Failure"  nsresult: "0x80004005 (NS_ERROR_FAILURE)"  location: "JS frame :: jar:file:///C:/Program%20Files%20(x86)/firefox/extensions/%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js :: <TOP_LEVEL> :: line 4885"  data: no]
Source File: jar:%7B73a6fe31-595d-460b-a920-fcc0f8843232%7D.xpi!/components/noscriptService.js
Line: 4885

[therube]

(Can't seem to duplicate with today's Nightly & Flash, release version.)

[dhouwn]

OK, there is now good evidence that this is indeed non caused by NoScript, as the error message pointing to the NoScript source just comes after I killed the plugin-container process but Flash stops working before that but no error message appears.

[GµårÐïåñ]

I refer you to when I told you that a while back. Never hurts to beat a dead horse I guess.