753129 - Flash still broken for some Nightly testers due to bug 751641

Status: RESOLVED WONTFIX

(Core Graveyard :: Plug-ins, defect)

Description

[u279076]

A fallout from bug 751641 resulted in several Nightly testers having broken Flash in Firefox 15.0a1. The side-effect is that Flash is just not loading for some sites (Youtube, Vimeo, etc). 

Reinstalling Flash 11.2 is known to fix the issue.

Comment 1

[Johnny Stenback (:jst)]

My guess here is that a user running a nightly build with the bug that was fixed in 751641 ends up with a busted pluginreg.dat file laying around, and that's causing problems for those users. The only way to fix this is to trigger re-generation of that file by either installing a plugin or manually deleting the file.

Josh, would it make sense to bump the version in the pluginreg.dat file here to force re-generation of the file so the users who run into this get a new file and functioning plugins?

Comment 2

[Josh Aas]

The pluginreg.dat bug was only around for a couple of days, and if you triggered a re-generation during that time in order to see the bug you'll probably trigger it again to get rid of it. You can manually get rid of the problem by just deleting pluginreg.dat. If we bump the version we'll be triggering an unnecessary plugin scan (disk activity, jank) for millions of regular users when we release it.

Comment 4

[Alex Keybl [:akeybl]]

(In reply to Josh Aas (Mozilla Corporation) from comment #2)
> The pluginreg.dat bug was only around for a couple of days, and if you
> triggered a re-generation during that time in order to see the bug you'll
> probably trigger it again to get rid of it.

None of the bugs duped to bug 751641 mention updating plugins as part of the STR to get broken Flash. If this is confirmed behavior, and updates of pluginreg.dat happen very infrequently, we can wontfix.

Comment 5

[Josh Aas]

(In reply to Alex Keybl [:akeybl] from comment #4)

> None of the bugs duped to bug 751641 mention updating plugins as part of the
> STR to get broken Flash. If this is confirmed behavior, and updates of
> pluginreg.dat happen very infrequently, we can wontfix.

Updating plugins isn't the only way to get pluginreg.dat to regenerate. I forget the full list of ways but it's possible (iirc) via the DOM's plugin refresh method as well, which a large number of popular sites use.