627215 - Post-FF4: Enable browser.sessionhistory.optimize_eviction

Status: RESOLVED FIXED

(Core :: DOM: Navigation, defect)

Description

[Alon Zakai (:azakai)]

Followup to bug 610417. We enabled the pref for Fennec, but left it off for Firefox to not risk regressions.

After FF4, we should enable this pref for Firefox, since it is beneficial (see details in previous bug, and TODO comments in the patch for that bug).

Comment 1

[Martijn Wargers (dead)]

From the patch in bug 610417:
// NB: After 4.0, we should remove this option and the corresponding
//     pref - optimization should always be used

So I guess now that Firefox 4 has been released, this optimization should be used and the browser.sessionhistory.optimize_eviction pref should be removed?

Comment 2

[Alon Zakai (:azakai)]

Attachment: patch

Thanks for the reminder mw22!

Comment 3

[Olli Pettay [:smaug]]

Comment on attachment 525119 [details] [diff] [review]
patch

Could we first flip the pref for desktop FF and after some time
remove this code.

Comment 4

[Alon Zakai (:azakai)]

Attachment: change pref to true

That does sound safer.

Btw, I don't know how these prefs update existing installations - will users that upgrade to this code get the pref changed? Or will it remain as it was set before (so they won't get the optimization)?

Comment 5

[Alon Zakai (:azakai)]

http://hg.mozilla.org/mozilla-central/rev/17cbe366cd51

Comment 6

[Jim Jeffery not reading bug-mail 1/2/11]

(In reply to comment #4)
> Created attachment 526775 [details] [diff] [review]
> change pref to true
> 
> That does sound safer.
> 
> Btw, I don't know how these prefs update existing installations - will users
> that upgrade to this code get the pref changed? Or will it remain as it was set
> before (so they won't get the optimization)?

I just installed the latest hourly build and checked before installing that the pref was 'false' - after the install to latest hourly/Nightly - the pref is now 'true' , so users will get the optimization without issue.

Comment 7

[Alon Zakai (:azakai)]

Thanks Jim, good to know.