Open Bug 639839 Opened 14 years ago Updated 3 years ago

Page zoom not kept between page navigation

Categories

(Firefox :: General, defect)

x86_64
Linux
defect

Tracking

()

People

(Reporter: jstpierre, Unassigned)

Details

(Whiteboard: [bugday-2011-05-27])

User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:2.0b13pre) Gecko/20110308 Firefox/4.0b13pre Build Identifier: 4.0~b13~hg20110308r63349+nobinonly-0ubuntu1~umd1~maverick When reloading a page in the same domain, the zoom size should be kept. Right now, for a number of sites, it resets the zoom size. Reproducible: Always Steps to Reproduce: 1. Go to http://google.com/ 2. Press Ctrl-+ to zoom in 3. Go to http://google.com/search?q=foobar Actual Results: Zoom size went back to 100%. Expected Results: Zoom size should be kept per-domain. This started happening at most a week ago. Doesn't work on all sites: BMO is working properly, but it fails on other HTTPS sites. The bug also has the effect with a hash, like: http://google.com/# -- so AJAX sites that use this to navigate are quite annoying for people with bad eyesight like me.
Still happen in safe mode / with a clean profile? http://support.mozilla.com/kb/Safe+Mode http://support.mozilla.com/kb/Basic+Troubleshooting#w_make-a-new-profile Failing that, can you find the regression range please: http://harthur.github.com/mozregression/
Version: unspecified → Trunk
Safe Mode doesn't work, but a clean profile does.
Does copying your prefs.js from the profile that doesn't work to the one that does make it also not work?
It seems to be content-prefs.sqlite that causes the bug.
OK, definitely not a core text rendering issue...
Component: Layout: Text → Preferences
Product: Core → Toolkit
QA Contact: layout.fonts-and-text → preferences
Status: UNCONFIRMED → NEW
Ever confirmed: true
Component: Preferences → General
Product: Toolkit → Firefox
QA Contact: preferences → general
Do you guys still me to do a bisect?
That would be useful in pinning down the problem, yes... Make a copy of the content-prefs.sqlite first, so the bisecting doesn't mess it up?
OK, it looks to just be a corrupt content-prefs.sqlite, not a regression. $ sqlite3 content-prefs.sqlite sqlite> SELECT * FROM groups 1|bugzilla.mozilla.org ... 673|bugzilla.gnome.org Error: database disk image is malformed Still, you should probably do something about that.
Whiteboard: [bugday-2011-05-27]
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.