Fills up .xsession-errors with "GetValue variable 1 (1)"
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
firefox-3.0 (Ubuntu) |
Fix Released
|
Low
|
Unassigned |
Bug Description
Binary package hint: firefox
Since (I think) sometime after I upgraded from Firefox 1.5 / Ubuntu 6.10, Firefox on all my machines, with or without safe mode, keeps on emitting uninformative messages on standard error.
For most users, who start up Firefox by clicking on its icon in the window manager of their choice, the effect is not directly visible, but will (soon enough) result in .xsession-errors becoming huge, and more important messages being lost, because the X server stops logging once the file grows large (and even before that, because these uninformative messages drown out more important and useful information).
As far as I can tell, this is a way to repro:
(This will start a second instance of Firefox, so you don't have to quit if you are reading this in a Firefox session.)
vnix$ firefox -ProfileManager -safe-mode
Create a new empty profile, select all the safe mode checkboxes (disable add-ons, reset toolbars and controls, bookmarks, preferences to defaults, etc) and continue in safe mode.
On the bookmarks toolbar, click the Mozilla "Getting Started" link. At this point, you should see the following output at the prompt where you started Firefox:
** Message: GetValue variable 1 (1)
** Message: GetValue variable 2 (2)
** Message: GetValue variable 1 (1)
** Message: GetValue variable 2 (2)
** Message: GetValue variable 1 (1)
** Message: GetValue variable 2 (2)
** Message: GetValue variable 1 (1)
** Message: GetValue variable 2 (2)
I don't know what it takes to get it to display more of these, but as I surf in my regular session, I see a steady trickle of them. I have not been able to pin down exactly what triggers these messages.
You may want to remove the new dummy profile after you're done. To the best of my understanding, you have to quit Firefox and start it with firefox -ProfileManager again.
To my surprise, I am unable to find an upstream bug about this, but that is probably more a reflection of my lack of competence (and/or the lack of a usable search facility in their bugzilla) than proof that it doesn't exist.
Actually the regression in LP #31746 may prevent the repro steps from working in Gutsy. The machine where I'm writing this is on Feisty still.