343352 - Session restore clobbers my window position on restore

Status: VERIFIED FIXED

(Firefox :: Session Restore, defect)

Description

[Adam Guthrie]

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a1) Gecko/20060626 Minefield/3.0a1

Steps to reproduce:
1. Set browser.sessionstore.resume_session to true.
2. Make sure the window /isn't/ maximized, or in the top left of the screen.
3. Restart Firefox.

Results:
The browser window is positioned in the top left corner of the screen.

Expected results:
The browser window's position is restored to where it was previously.

This works when I don't use session restore, so it's gotta be something session restore does.

Comment 1

[Adam Guthrie]

This also happens on branch and when just resuming the session if it crashed.

Comment 2

[Adam Guthrie]

So, does session store explicitly store the window position (doesn't look like it after a glance at the code) or just expect localstore.rdf to take care of this? Or maybe session store is overriding what localstore.rdf says...

Comment 3

[Simon Bünzli]

(In reply to comment #2)
SessionStore explicitly saves the windows' positions (in _updateWindowFeatures) and relies on localstore.rdf only for the very first window to be restored (otherwise restoreWindowFeatures should move the window back where it was).

Comment 4

[Simon Bünzli]

*** Bug 364389 has been marked as a duplicate of this bug. ***

Comment 5

[Simon Bünzli]

Attachment: never use the window dimensions from localstore.rdf

Comment 6

[Asaf Romano (gone)]

Comment on attachment 331808 [details] [diff] [review]
never use the window dimensions from localstore.rdf

r=mano

Comment 7

[Dão Gottwald [:dao]]

http://hg.mozilla.org/index.cgi/mozilla-central/rev/6c8bb2e4fb41

Comment 8

[Michael Kohler [:mkohler]]

Verified with Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.2a1pre) Gecko/20090617 Minefield/3.6a1pre ID:20090617031528

Comment 9

[u279076]

in-litmus-
There are multiple tests in litmus that include window size/position in their Expected Results.  A specific test case is not needed IMO.

Comment 10

[u279076]

(In reply to Dão Gottwald [:dao] from comment #7)
> http://hg.mozilla.org/index.cgi/mozilla-central/rev/6c8bb2e4fb41

This appears to have caused a regression (see bug 680001)

Comment 11

[Paul O'Shannessy [:zpao] (not bugmail, email directly)]

(In reply to Anthony Hughes, Mozilla QA (irc: ashughes) from comment #10)
> (In reply to Dão Gottwald [:dao] from comment #7)
> > http://hg.mozilla.org/index.cgi/mozilla-central/rev/6c8bb2e4fb41
> 
> This appears to have caused a regression (see bug 680001)

Technically yes, but meh. That should be solved differently.