Open Bug 933470 Opened 6 years ago Updated 3 months ago

ctrl + pgdown / pgup not switching tab when on Google+ and Jira


(Firefox :: Tabbed Browser, defect)

25 Branch
Windows 7
Not set




(Reporter: dmladek, Unassigned)


(Blocks 1 open bug)


(Keywords: access, regression, ux-control)


(1 file)

Attached video FFv25-regression.mp4
User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:25.0) Gecko/20100101 Firefox/25.0 (Beta/Release)
Build ID: 20131025151332

Steps to reproduce:

This must be regression between FF v24 and v25 to which version I've upgraded a couple days ago and since my user experience with this browser change so I can't use it conveniently anymore.

1. Start Firefox
2. Open several TABs with different URLs, e.g.,,
   this bug:,, etc.
3. Be sure that one of those pages is
4. Select a TAB which is different from Google+ 
5. Press CTRL + PgDown several times and watch how the tabs are changed

Actual results:

Actual Results:
- When Google+ tab is reach, another pressing of CTR + PgDown or PgUp won't navigate to next/previous tab. Instead of that it scrolls up/down (in jumps) or tries to interact with navigation widgets in the page.

The same behaviour you get if you use open a ticket in Jira  (v4.4.5#665) (another backtracking system from Atlassion). Then you cannot navigate from that TAB using CTRL +PgUp/PgDown.

See attached video with navigation issue on Google+ and Jira pages

Expected results:

Expected Results:
- Using CTRL + PgUp/PgDown should always control navigation between browser's TABs, not navigation around webpage.
I manage to reproduce this bug on the latest Nightly build:

Build identifier: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:28.0) Gecko/20100101 Firefox/28.0
Build Id:20131103030204
OS: Win 7 x64
Component: Untriaged → Tabbed Browser
Ever confirmed: true
Interesting. I can see a benefit from having browser's keyboard shortcuts overwritten by a webapp, however it shouldn't be in control of the webapp itself, but in a control of a user and their browser IMHO.

So ideally, there should be a configuration option which would allow overwrite browser's keyboard shortcuts per site or in global. 
Problem here would be how to let the user know they have this option - kind of popping up notification bar on the site which tries to overwrite it?
There needs to be an option to prevent webpages from overriding keyboard actions. This particularly becomes relevant to users who use keyboard layouts that include non-US letters or extended layouts for non-ASCII letters/symbols (such as m-dash, and others); these layouts typically - on Windows - use CTRL+ALT combinations. If these websites would just stick to CTRL *or* (not "and/or") ALT combos that **did not** conflict with neither OS nor browser hotkeys, things would not be as bad... but that is not what happens.
Websites overriding the keyboard seems to also break Firefox's 'Search as You Type' functionality (for example, SAYT does not work on Facebook).
Do people actually use any of these website created hotkeys, anyway?
(In reply to busi.bugzilla from comment #4)

> There needs to be an option to prevent webpages from overriding keyboard
> actions.

Let's put it the opposite way. There can be an option to allow Web pages to override keyboard actions. But the default situation has to prevent this. Even the option would lead to awkward UI, to a broken Firefox.
> Everyone agress this being a dupe of bug 380637?

Except that the title of file bug 380637 asks a question. The file bug 933470 answers the question with NO. In order to have duplicates, we should change the file bug 380637 to "Don't allow Web pages to override keyboard actions". It would be in accordance with the vast majority of the opinions that have been expressed.
I am seeing the same regression- with Novell GroupWise WebAccess, which now also "eats" Ctrl+PgUp/PgDn
while going between tabs, effectively preventing switching to a different tab.
This bug seriously annoys me as I'm SW Tester and Jira is my daily tool which I can't use effectively any more with FF (Switching with dozens and dozens open TAB by mouse, I'd develop RSI syndrom within a couple of days).

Just another version of FF was released and no improvement nor indication that is going to be fixed. Luckily, there is Chrome or other browsers where pay a bit more attention to their usability.
+1 here,  with duplicate bug 949020.  We should not allow javascript to intercept key combinations that the browser is intending to use as a default behaviour, though I do agree that giving user permission to allow the overriding might be a good idea.

For the record, IRCCloud is affected by this bug.  They are now going to be checking the status of the modifier keys in Javascript, but their interception was more generic (just PageUp/PageDown), and they had no desire to override FF's behaviour.  I suspect the vast majority of web developers who end up intercepting these keystokes do so unintentionally.
Interestingly, this problem does not happen with newer jira. 
Eg. I could not reproduce it on following jira version:  Atlassian JIRA (v6.1.4#6159-sha1:44eaede) 
Visit this link for example:
Ignore my earlier comment, I had disabled keyboard shortcut in my jira profile so that problem did not show up.
I've made an addon to workaround this issue.
Blocks: 657080
Regression window(m-i)
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:25.0) Gecko/20130724 Firefox/25.0 ID:20130724230740
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:25.0) Gecko/20130724 Firefox/25.0 ID:20130724230940

Triggered by:  Bug 501496
Blocks: 501496
Depends on: 1008772
You need to log in before you can comment on or make changes to this bug.