Firefox 20 Will Finally Fix Private Browsing Mode

darthcamaro writes "Unlike every other major browser vendor, Mozilla today does not allow users to have their private mode browser window open at the same time as a regular browser window. That's now set to change. This is a flaw that has been in Bugzilla since 2008 and has been the subject of heated discussion for years."

Comments

How about tabs in the same window?

[addie]

For a Chrome-addict like me, what I've been waiting for is the option to open an incognito tab within the same window as regular tabs. Apparently this lack of functionality is a feature intended to ease confusion among users. For me, I just find it irritating.

Re:How about tabs in the same window?

[Anonymous Coward]

I would totally mess that up. Open porn in the wrong tab and forget to clear cache because I'm used to not having to do so.

It doesn't help that I have 50 tabs open at one time, usually in the same window.

Re:How about tabs in the same window?

[crow]

Private mode tabs should have a different theme or color for the browser portion above the web page so that it's obvious that it's different. There's no need to force them into a different window.

Re:How about tabs in the same window?

[ccguy]

I'm waiting for incognito mode not only not to leave track on the computer itself but also on the remote sites I visit. How is it incognito if I connect somewhere I've been before to and you send the cookies that were already saved for that site, for example?

So basically, fix the thing :-)

Re:How about tabs in the same window?

[Rhaban]

That's how it works in opera: anonymous tabs not windows (but you can put the tab in a separate window if you want to).

As a web developper, I often use this to have several sessions with different users on the same website.

Re:How about tabs in the same window?

[SJHillman]

This is how Opera has done it from the beginning... I'm really surprised the others haven't at least made it a non-default option yet

Re:How about tabs in the same window?

[Anonymous Coward]

Erm, if the browser is sending a cookie, that means the cookie was saved in your incognito session, so the thing is broken in the first place. It shouldn't save cookies at all.

Re:How about tabs in the same window?

[calzones]

Reminds me how I've wished for a new http "undo" feature.

Basically, if I make a request of a page from a server and decide it was a mistake, I want to invoke "undo" and have my browser history go back, wipe any cookies or history or cache trace, plus delete anything downloaded... AND THEN ALSO send an "undo" header to Apache to request wiping my visit from the logs.

Of course that would be open to abuse. So servers should only honor such "undo" requests if they happen within X seconds (say, 120) after the last non-ajax bit was sent to the browser, and as long as no further requests are made by the browser after the first one. For example, click a link on the page, interact with a form widget, or invoke a new ajax request... and you'd totally kill the ability to "undo".

Re:How about tabs in the same window?

[chris.alex.thomas]

ok, at least you're honest about it, but holy shit, really, 50 tabs? you really need that much porn open at once.....you must be like a rabid sex monkey 23 hours a day....

Re:How about tabs in the same window?

[Anonymous Coward]

are you really THAT paranoid about ... well, i dunno, something?

oh wait, this is slashdot, so, yes.

Re:How about tabs in the same window?

[calzones]

Actually, as a web guy, I also care about logs that are free from false positives due to accidental clicks and redirects. A feature like this would help me verify that traffic to a page on the site is purposeful and desired by the end user.

Re:How about tabs in the same window?

[icebike]

For a Chrome-addict like me, what I've been waiting for is the option to open an incognito tab within the same window as regular tabs. Apparently this lack of functionality is a feature intended to ease confusion among users. For me, I just find it irritating.

No, its to prevent leakage of data via the container. Each tab is supposedly running in a sandbox, but if they are in the same container window there is a risk there.

Re:How about tabs in the same window?

[icebike]

I'm waiting for incognito mode not only not to leave track on the computer itself but also on the remote sites I visit. How is it incognito if I connect somewhere I've been before to and you send the cookies that were already saved for that site, for example?
 

According to the help page about Incognito mode,
it explicitly states it deletes all cookies when you exit incognito mode. Use it for single purpose at a time, and close it out after the fact, there will be no cookies left for them to find.

Never log into any account while in incognito mode, unless you ONLY log in there while in incognito mode.

Re:How about tabs in the same window?

[Anonymous Coward]

Actually, as a web guy, I also care about logs that are free from false positives due to accidental clicks and redirects. A feature like this would help me verify that traffic to a page on the site is purposeful and desired by the end user.

And he really did opt in.

Re:How about tabs in the same window?

[Splab]

Not a very bright web guy? Just look at how long they are on your page, less than X minutes and you just wipe their session again (or reverse it and just mark sessions that are actually busy enough to count as "interested").

(and don't wipe the data, amount of people who aren't interested in your site is probably as important as how many are)

Re:How about tabs in the same window?

[ls671]

Just watch where you click. Touchpads with mouse clicks enabled anyone? Touch screens?

Am I the only one finding it is easier to click somewhere you don't intend to of make other input mistakes when using one of these compared to a conventional 3 button mouse with a scroll wheel?

Am I just too old school? Seems to me like doing flawless input with these is an ability challenge in itself.

Re:How about tabs in the same window?

[Anonymous Coward]

They already have the different firefox button colors. So just do orange tabs for regular and purple for private tabs. As a plus, you can only show the different tab colors when they actually have an private session going.

Re:How about tabs in the same window?

[X0563511]

Easy enough to change - modify the tab colors so it's VERY obvious the tab is in incognito mode.

Re:How about tabs in the same window?

[X0563511]

I once had 4 instances of mplayer running at one time.

It was.... interesting...

Re:How about tabs in the same window?

[X0563511]

Why does that matter? It's all under the same process.

Re:How about tabs in the same window?

[jones_supa]

Reminds me how I've wished for a new http "undo" feature.

Basically, if I make a request of a page from a server and decide it was a mistake, I want to invoke "undo" and have my browser history go back, wipe any cookies or history or cache trace, plus delete anything downloaded... AND THEN ALSO send an "undo" header to Apache to request wiping my visit from the logs.

Of course that would be open to abuse. So servers should only honor such "undo" requests if they happen within X seconds (say, 120) after the last non-ajax bit was sent to the browser, and as long as no further requests are made by the browser after the first one. For example, click a link on the page, interact with a form widget, or invoke a new ajax request... and you'd totally kill the ability to "undo".

That would just make things unnecessarily complicated.

Re:How about tabs in the same window?

[icebike]

Checked your task list. Chrome spawns a sandboxed task per tab.

Re:How about tabs in the same window?

[dolmen.fr]

ok, at least you're honest about it...

Honest, but Anonymous Coward... Does that really count?

Re:How about tabs in the same window?

[dolmen.fr]

Actually, as a web guy, I also care about logs that are free from false positives due to accidental clicks and redirects.

As a web guy that is security aware I would not want that a hacker uses this "feature" to hide his tracks after his forfeit.

Re:How about tabs in the same window?

[Anonymous Coward]

about 410 open at the moment in chrome here... (yes, I like and need them, they aren't the problem, the problem is when software limits them!)

Re:How about tabs in the same window?

[Johann Lau]

Give private tabs a fat, colorful border. Or make them negative even (white text on black bg). Surely there are a million ways to make this obvious.

Re:How about tabs in the same window?

[Johann Lau]

Am I the only one finding it is easier to click somewhere you don't intend to of make other input mistakes when using one of these compared to a conventional 3 button mouse with a scroll wheel?

Nope. Can't do without a mouse. I guess a tablet (WACOM etc.) might be sweet too, but none of that other crap for me. It's just crap... I'm not even a gamer, but I have a somewhat gamer mouse; and that precision and speed is just something I would not want to miss. Flawless is a good word to describe it :)

Re:How about tabs in the same window?

[allo]

who uses the firefox button? Everyone i know just re-enables the menubar.

Re:How about tabs in the same window?

[Anonymous Coward]

Wouldn't that make it a little less incognito?

Re:How about tabs in the same window?

[swillden]

I know this will shock many slashdot users... but people do use the web for things other than porn (and slashdot).

Re:How about tabs in the same window?

[Archenoth]

That's how it works in opera: anonymous tabs not windows (but you can put the tab in a separate window if you want to).

As a web developper, I often use this to have several sessions with different users on the same website.

Yep..!
You can also close all private tabs without effecting your non-private ones with [Ctrl] + [Shfit] + [Q].

You can even assign any of the actions related to private browsing to a mouse gesture or keyboard shortcut in Opera -> Settings -> Preferences -> Advanced -> Shortcuts.

Re:How about tabs in the same window?

[Anonymous Coward]

This is how Opera has done it from the beginning... I'm really surprised the others haven't at least made it a non-default option yet

I know, and I hate it.

Opera lets you open a private tab in a window, browse incognito that way, and then if you hit ctrl-t it creates a *normal*/logged tab. Definitely not Least Surprise as the context switch is easy to miss (it's caught me several times).

It's unwieldy that Opera's Private Window option isn't at the top level context menu when launching from the quickstart icon in Windows 7, so I find myself having to launch Opera with a "sacrificial" window and then spawn the Private Window from there and (finally) close the extraneous sacrificial window.

I'm all for choice, so I propose you be allowed your non-default option in other browsers, and I should be allowed to configure Opera to operate like all the rest of the browsers (ie. a less painful privacy mode). While they're at it, they can perhaps also make popups that aren't bound to their parent window (painful composing in GMail, anyone?) and tabs you can tear off their window to promote them to full blown windows on their own and tabs you can tear off and attach to different window.

Opera was a real innovator when it came to early use of tabs, but they've been resting on their laurels for the past decade or so now.

Re:How about tabs in the same window?

[Anonymous Coward]

No, what I meant was that the color changes depending on the mode, so they already have the ability to change the UI depending on the privacy mode. And since the orange and purple has sort of been established, that is how they should color the tabs when private mode is enabled.

Finally!

[c0l0]

That's one of the very few features that I'd always wanted Firefox to adopt from Chromium, and now it's actually happening - yay for Firefox 20. Can't be longer than a few weeks any more anyway; now can it? ;p

Re:Finally!

[Lennie]

2013-04-02 to precise.

But you can probably download the daily build right now, the pre-Beta/Aurora in the week of 2013-01-06 or the beta on 2013-02-19.

https://wiki.mozilla.org/RapidRelease/Calendar

Re:Finally!

[Anonymous Coward]

That's one of the very few features that I'd always wanted Firefox to adopt from Chromium

You mean "from Opera". Opera is always 5+ years before the other browsers featurewise, at least with those features you actually want. In this case, I think they already had (primitive) private browsing in c:a 2000 (primitive barriers between sites in different tabs (opera had "tabs" in the 1990's) and windows, and don't save history, cache and cookies on exit), and more advanced private browsing c:a 2005. To lazy to check if I remember correctly.

Re:Finally!

[Zaiff Urgulbunger]

There's scope for improvement over Chrome; with Chrome, if I open one incognito window, log in to something or other and then open a completely fresh incognito window, it'll still have the same cookies as the other one.... I'd like to have new windows be completely separate incognito sessions. So I hope Firefox does this right!!

Re:Finally!

[Compaqt]

Just right click on the Chromium launcher in the Unity dash. It has these options:
-Open a new window
-Open a new incognito window (shares cookies among all incog sessions)
-Open a new window with a temporary profile (this is what you want)

Re:Finally!

[Zaiff Urgulbunger]

I''m running 12.04LTS and I do not have these options... but you intrigue me sir! Is this a 12.10 thing, or have you got "special" unity launcher options?

Re:Finally!

[Anonymous Coward]

A new porn mode for my thirtieth birthday. How fitting!

Re:Finally!

[Compaqt]

No, I'm running Precise Pangolin (12.04). I just installed Chromium from the repositories. I did not customize my dash.

These options are highly useful for web testing (test with a fresh start).

Re:Finally!

[Zaiff Urgulbunger]

Cheers!

I'm using the full Google Chrome rather than Chromium, which it turns out, doesn't have a --temp-profile option so I think I'll have to install Chromium too!

Um, Safari?

[Anonymous Coward]

I just tried this in Safari, and the private browsing mode seems to be all or nothing.

Re:Um, Safari?

[elashish14]

Yes, but the summary mentioned major browsers

*ducks*

version 20

[Anonymous Coward]

So this feature will show up by Wednesday?

Re:version 20

[Drinking Bleach]

Tuesday

I Don't Care Anymore

[Anonymous Coward]

The latest firefoxes are all so horrible that I just don't care anymore. They insist on putting a google search bar somewhere on the page, when what I want is a URL and only a URL field. I don't want it adding .com and I sure as hell don't want it sending something I paste into the URL field anywhere if it's not a valid hostname. In fact, I don't want it making any outbound connections that I didn't specifically ask for. It's been progressively downhill from about version 3.0.

Re:I Don't Care Anymore

[e065c8515d206cb0e190]

You realize you can configure it to do just that, right?

Re:I Don't Care Anymore

[gparent]

Paranoid people should take the time to look at the options. They're there for a reason. Options that nobody care about may not be in the options menu but they'll be in about:config.

finally

[physlord]

Finally, a decent porn mode. Why it took that long.

Re:finally

[Anonymous Coward]

Us Opera users have been wanking it for years.

Firefox?

[Threni]

Oh yeah, the desktop browser - the one I used to use before switching pretty much exclusively to mobile browser. Do they still bother to compete with Chrome, Dolphin, and the stock Android browser? Last time I tried it was huge, clunky and setting up syncing of bookmarks was a joke (as opposed to, say, Chrome which just syncs up for you). I used to love FireFox but it just got bigger and slower; there's just no point any more.

Re:Firefox?

[epyT-R]

Well some of us do more with a computer than check up on the latest kardashian kraze and failbook. While we have browser preferences, we tend to choose the one that has the plugins we need, even if it is technically inferior software.

Re:Firefox?

[micheas]

Firefox on android syncs with my desktop well enough. (android 4.0 on the phone and debian on the desktop.)

Re:Firefox?

[Anonymous Coward]

Well I for one do not want my porn history uploaded to Chrome at work. Even if I do not click on porn at work I have routers that sniff for things over the network. I work for a school district so that tells you how strict it is.

Pr0nzSec 2.0

[bambewn]

Never worried much about it really... Didn't think it was an issue. After private lemonbrowzing, I would just Ctrl-Shift-P back to my, worksafeish shenanigans.

When will they add HSTS security?

[WillAffleckUW]

I'm kind of wondering when they'll add HSTS security, which ensures all traffic that starts on HTTPS does not get redirected to insecure HTTP instead.

Re:When will they add HSTS security?

[gparent]

Maybe if you upgrade from Firefox 3.6 you'll find out about the new features they added?

Re:When will they add HSTS security?

[WillAffleckUW]

Mine says Firefox 17.0 and it just update to that.

Re:When will they add HSTS security?

[Anonymous Coward]

Then you've had HSTS for the past 13 versions or so.

Re:When will they add HSTS security?

[BZ]

HSTS support was added in Firefox 4, a bit over year and a half ago. So wonder no longer!

does it fix downloads cutting out though?

[Anonymous Coward]

the worst thing in firefox when you're downloading some crap off a slow ass website and then hop into porn mode for a fap and it stops your downloads because they were in normal mode! this has made me rage hard so many times. fix it, please.

But But

[fermion]

Chrome is version 23, so Firefix is still lame, and lord help IE stuck at version 10.

Re:But But

[mister_playboy]

Thanks for mentioning Chrome's current version number.

I've got real tired of seeing people here bitch about FF's version number inflation causing them to move over to Chrome... ignoring the fact Chrome has an even higher version number after fewer years of existence compared to FF.

This basic math failure is quite egregious on an site supposedly for nerds.

Re:But But

[jez9999]

But Slashdot doesn't talk about "Chrome 23"...

Re:But But

[Waccoon]

IE is several times older than Chrome. We only use mature software around here, bro.

Re:But But

[Anonymous Coward]

You mean IE6?

Re:But But

[Anonymous Coward]

I think the big difference is that on Chrome the update just works, whereas on Firefox it is still broken (!). I really hope that the next update will fix it, but I thought that so many times before, I am not sure I should put my hopes up.

Re:But But

[WillHirsch]

Guess how many of Chrome's major version updates have come with a you-shall-not-pass dialog while it checks for add-on compatibility? Hint: it's less than 22.

Private mode as default

[sunderland56]

Why would anyone, ever, browse the web in "public" mode??? Isn't that like saying "Yes, please track me"?

First thing I do on Firefox is set it to dump all cookies/cache/history/etc/etc every time it closes. 100% on all the time private mode would be just fine, thanks. Unless you *like* big brother watching you.

Re:Private mode as default

[SJHillman]

Not everyone is that paranoid. Some people like the convenience of saving cookies and cache across sessions.

Re:Private mode as default

[e065c8515d206cb0e190]

How do you handle logins and passwords to various sites?

Re:Private mode as default

[neminem]

Because I don't care. Things can track me all they like. I seriously don't care whether they do or not. (Except every once in a blue moon, in which case, I toggle on private mode. Or just log out of google, because that's where most of the major tracking is going on... :p)

Re:Private mode as default

[imsabbel]

Or you want usability. Or do you tell me you never use bookmark, history, or anything in your web browsing. Always type every URL per hand?
Cause thats more or less what it would boil down to be always in privacy mode

Let me guess, you thing you are a big shot because you have them on your google account...

Re:Private mode as default

[Drinking Bleach]

Definately not what most people want -- I certainly would hate that, too.

If you want it though, it's easy enough to enable: https://support.mozilla.org/en-US/kb/private-browsing-browse-web-without-saving-info#w_how-do-i-always-start-firefox-in-private-browsing

Re:Private mode as default

[epyT-R]

alternatively, you could say "most people aren't aware of the serious negative side effects that come with convenient defaults."

Re:Private mode as default

[epyT-R]

our brains..

Re:Private mode as default

[godel_56]

How do you handle logins and passwords to various sites?

KeePass, plus the KeeFox addon to do the log-ins. Or any of the other password managers if you don't like that combination.

KeeFox detects the login page about 80% of the time in my case, and Autotype in KeePass works when KeeFox doesn't.

Re:Private mode as default

[wmac1]

Your brain can remember hundreds of different passwords?

Or perhaps you use repeated passwords. In that case, good for you and your security.

Re:Private mode as default

[tbird81]

That's because there are none. Seriously, start taking your schizophrenia meds, because no-one is out to get you. No-one cares what websites you visit. It will not affect your life in any meaningful way.

Re:Private mode as default

[Anonymous Coward]

Or LastPass.

Re:Private mode as default

[PlusFiveTroll]

No-one cares what websites you visit. It will not affect your life in any meaningful way.

If no-one cares where is there a billion dollar tracking industry, and Google spends a lot of effort tracking everything.

It will not affect your life in any meaningful way, unless you don't get along with your government, get divorced, are a public figure, may become involved in politics, etc, etc, etc.

Just because no one gives a shit about your masturbation habits in your moms basement, doesn't mean there aren't plenty of people that other groups would love to track.

Does this mean I can

[Spy Handler]

post politically correct things (Linux rules and Bush sucks!) to Slashdot while logged in, and at the same time open a Private Browsing window and troll as AC?

Re:Does this mean I can

[Idbar]

As long as you don't forget and start replying to yourself non-AC, seems like fair game.

Re:Does this mean I can

[Anonymous Coward]

DISREGARD THAT, I SUCK COCKS

Re:Does this mean I can

[teh31337one]

Mod parent funny please.

Re:Does this mean I can

[Anonymous Coward]

And mod grand parent insightful.

Since there is no mod that both covers sexiest man alive and big pile of wisdom

Journal

Re:Does this mean I can

[Anonymous Coward]

Yeah, yeah., now you post this as an AC with a link to your journal... We all know you said what you really wanted to say in that post about sucking cocks.

Multiple Profiles are More Functional

[Jah-Wren Ryel]

Firefox has supported multiple simultaneous sessions since at least the 3.x days.

Use these command-line options:

    -ProfileManager -new-instance

Then create as many different profiles as you want. They will all have their own history, bookmarks, add-ons, cookies, etc. The only place you have to worry about cross-profile pollution is with plugins like flash that keep state (like flash-cookies) in their own directory rather than under the firefox profile directory.

I have about 8 different profiles - one for gmail, one for my bank, one for slashdot, one for IMDB, etc and I keep a special "anonymous" profile that is basically a private-mode session, it wipes everything on exit, cookies, disk cache, history, etc. I even use the "User Agent Switcher" add-on so that each profile pretends to be a slightly different version of Firefox to make browser fingerprinting a little bit harder.

Re:Multiple Profiles are More Functional

[Celarent Darii]

Javascript profiling will identify your browser even with user agent switcher. You can find out what browser they are using even with a simple DOM tree check. Heck even CSS can be used to find out what Browser you are really using. The agent string is only for convenience.
Javascript example: http://www.corephp.com/blog/hardcore-javascript-browser-and-computer-fingerprinting/
Paper on different method: http://w2spconf.com/2011/papers/jspriv.pdf
Old CSS history method, now mitigated : http://ha.ckers.org/weird/CSS-history.cgi
tl;dr version: the internet is a public network, you are never really private in a public space.

Re:Multiple Profiles are More Functional

[Jah-Wren Ryel]

Yep, I really did mean "a little bit harder" when I wrote it.

Security is never 100% - on both the attacking and the defending sides. It is always a trade-off between effort and results. I figure the majority of trackers will not go to those extremes to track people because not enough people even go so far as to diddle the user-agent string. It just isn''t worth their time to do it and do it reliably when pay-out is such a small fraction of the total.

Re:Multiple Profiles are More Functional

[Anonymous Coward]

A couple points:

1. Anyone doing this likely has NoScript and/or Ghostery installed, so that crap won't happen anyway.
2. The user-agent string is one factor of many which is (in fact) used to try to identify a browser as uniquely as possible.

Re:Multiple Profiles are More Functional

[Jah-Wren Ryel]

I forgot to mention that using different themes for each profile really helps me to keep from accidentally using the wrong window. I also use the customize_titlebar_v2 add-on to change the titlebar to something unique to each profile, that helps when looking at iconified windows where you can't see the theme.

Re:Multiple Profiles are More Functional

[Anonymous Coward]

On linux it is --no-remote -ProfileManager, same usefulness.

Re:Multiple Profiles are More Functional

[Anonymous Coward]

No -new-instance is correct on all platforms. It used to be that -no-remote did the same thing, but now it turns off remote-control (like via script) which while esoteric enough that few would notice, isn't the intended use here.

Re:Multiple Profiles are More Functional

[girlinatrainingbra]

Look at EFF's Panopticlick website to see the breadcrumbs you're leaving behind. And don't forget that if you're coming in from the same IP address, even with all of those different purported browser-agent strings, it's easy enough to collect those data together and make a profile for that IP address and for the various sites hit at the various times of day. If you've got certain niche websites which you visit, the combination of websites visited could also be seen as a fingerprint also.
;>)
https://panopticlick.eff.org/

Re:Multiple Profiles are More Functional

[grumbel]

Firefox has supported multiple simultaneous sessions since at least the 3.x days.

They don't work properly in Ubuntu. Do a "firefox -P myprofile" while you have another profile running and Firefox will open a new Window with profile that was already running, not the one you gave on the command line. It's pretty badly broken and nobody seems to care.

Re:Multiple Profiles are More Functional

[vsync64]

You need to do -a blargle to get it to avoid just sending a request to the running session. I'd say the behavior isn't ideal but this is at least a workaround that lets me have multiple profiles running for dev/test/browsing.

Re:Multiple Profiles are More Functional

[grumbel]

The -a option right now is:

            -a or --debugger-args Specify arguments for debugger

And doesn't seem to have any effect for me. If I remember correctly, -a used to be to select the running process instance in the past, but even back then it never worked for me either. The relevant bug report from 2006 about the profile mess.

Re:Multiple Profiles are More Functional

[Jah-Wren Ryel]

Why aren't you using the -new-instance argument? It worked for me under ubuntu before I switched to Mint where it continues to work.

Re:Multiple Profiles are More Functional

[grumbel]

That's essentially the same as "-no-remote" and just gives a error message if an instance is already running:

"Firefox is already running, but is not responding. To open a new window, you must first close the existing Firefox process, or restart your system."

If you are wondering why I am not simply opening a new window via the GUI if an instance is already running: Sometimes the last window left of the instance will be a download window and the download window doesn't give you an option to open a new window.

Either way, it's just a damn old and annoying bug in Ubuntu's Firefox that turns using multiple profiles into quite a pain.

Re:Multiple Profiles are More Functional

[tbird81]

Oh no! Not the websites I visit! People will realise that I post on Slashdot!

Re:Multiple Profiles are More Functional

[Jah-Wren Ryel]

The best I can understand you have now defined two different scenarios:

1) Open a new firefox instance when there is no running instance with that specific profile
2) Open a new firefox window when there is already a running instance with that specific profile

These works for me on mint, I don't see why it wouldn't work on Ubuntu or any other platform:

1) firefox -P foo -new-instance
2) firefox -P foo -remote 'OpenUrl(about:blank,new-window)'

Re:Multiple Profiles are More Functional

[dolmen.fr]

2. The user-agent string is one factor of many which is (in fact) used to try to identify a browser as uniquely as possible.

Is it still the case now that browser versions change every month, and the User-Agent string with it?

Re:Multiple Profiles are More Functional

[doom]

I do this on ubuntu, and it seems to work:
firefox -no-remote --ProfileManager

It's useful for me because I like to use non-standard color settings, but there are sites that are unusable if you don't let them set their own colors. (It'd be nice if there were a better solution, but switching profiles as necessary is what I've been living with.)

Cool! I'll be able to use Firefox now.

[Sowelu]

I probably have to post some comment text or I bet this won't go through.

Do it now with Multifox

[Anonymous Coward]

http://br.mozdev.org/multifox/

It only took 4 years?

[Arker]

For firefox this is actually pretty good. My personal favorite bug is still plaguing users of FF over 11 years after being reported.

Re:It only took 4 years?

[Anonymous Coward]

TBF this is also a problem in chrome.

Re:It only took 4 years?

[Arker]

Yes, chrome is awful too. Browsers in general are pretty awful. Traditionally they have bent over backwards to enable plugins to do whatever they want, which is why this bug occured and more importantly why it has never been fixed and cannot be fixed. It's not a technical problem, it's a political problem, at this point a patch that fixed the original bug would never be accepted, it would cause too many problems with existing plugins.

The only browser I know of that at least partially avoids this bug is Opera. And I havent thoroughly tested it yet, so it may still be affected. But at least it passes one simple test that all the others fail. Go to youtube, start a video playing, then hit ctrl-t. In Opera that works just as it should.

Re:It only took 4 years?

[X0563511]

Not sure exactly how one would fix this. Either the plugin handles keys, or the browser does.

Re:It only took 4 years?

[Anonymous Coward]

Absolutely inexcusable that a critical bug like this hasn't been fixed in over a decade. Are they honestly going to sit there and say that this is far more complex than say, designing, building, manufacturing and releasing a popular mobile phone during that time? Or releasing Windows Vista, Windows 7 and Windows 8? Or any other number of technical achievements in the past decade?

Tells me a lot about Mozilla - and how broken their internal process must be. As a developer, I would be ashamed and appalled to allow a bug like this to remain open for so long without re prioritisation or resolution. At least be honest and mark it as 'won't fix'.

What a joke.

Re:It only took 4 years?

[Anonymous Coward]

From that Bugzilla thread, it appears that it never happened Safari or Internet Explorer either. There were a number of workable suggestions (such as Ctrl-T) made over the years and, yet, none of them were ever implemented. 11 years - no excuses.

Re:It only took 4 years?

[Shikaku]

Except I just tried it in the latest Chrome, Opera, and Firefox, and they all have the same bug. The fault is Adobe's, it should pass the keypress to the browser.

Re:It only took 4 years?

[Anonymous Coward]

This affects all plug-ins, not just Adobe's (see the Bugzilla thread) - so it's not their fault. Also, Safari and Internet Explorer don't experience this problem.

Point being is that you have a bug marked as Critical open for over a decade and it has never been resolved. It's also very poor usability coming from folks who are supposedly concerned about such things.

Re:It only took 4 years?

[Arker]

Hmm I could have sworn I have experienced it on both IE and Safari at some point, some sort of plugin blocking ctrl(cmd)+w at the very least. It seemed to me that all the browsers, not just firefox, were keen to 'enable' plugins. There was all that talk about the browser being the OS, you might recall - trying hard to encourage people to develop their programs with the browser as the target platform. I can understand programmers targetting that 'platform' wouldnt be happy about being told they wont even receive any meta-key input.

I havent read through the entire thread for several years and I might be confused.

Re:It only took 4 years?

[Arker]

Opera at the very least isnt susceptible to the same bug in the most common test case. I dont pretend to have tested it thoroughly. I just now tested IE and it passes the same. Chrome and Firefox failed. The fault is the affected browsers. They should be getting that keypress first, and not passing it through to the plugin.

Re:It only took 4 years?

[Arker]

The browser should always get the input first, and only pass it to the plugin after examining it and if appropriate.

Re:It only took 4 years?

[Anonymous Coward]

I don't know what version of Chrome you tried, but this does not happen to me at all when I tried it in Windows. Same thing for Internet Explorer.

You're either lying or using a really old version on some other platform.

Re:It only took 4 years?

[Anonymous Coward]

For firefox this is actually pretty good. My personal favorite bug is still plaguing users of FF over 11 years after being reported.

Perhaps there's a reason for that...

Roman 2012-09-07 02:57:01 PDT

WHERE THIS BUG STANDS AS OF TODAY

There is a technical side to this issue: Firefox is technically UNABLE TO FILTER THE KEYS [...]

There is a bounty standing at $164 as of today on this bug [...]

Yeah, but did they fix the leaks?

[Anonymous Coward]

I have a test virtual machine that has a clean Firefox 17 instance consuming 2.4GB of memory.

Seriously, it's fairly trivial to do this. Just open a bunch of GMail tabs and leave them open for a few days.

Re:Yeah, but did they fix the leaks?

[UltraZelda64]

Memory leaks have been a problem with Firefox since back around version 2.0. Already at version 17, clearly Mozilla Corporation is more interested in inflating the version number than actually fixing decade-old performance problems. Ironically, such high version numbers with the same old problems make them look even worse than they did with a sane versioning system... apparently that never crossed anyone's mind at Mozilla.

Re:Yeah, but did they fix the leaks?

[epyT-R]

haha hilarious..

Re:Yeah, but did they fix the leaks?

[Anonymous Coward]

Cool story bro. Why don't you use another browser then, or do you just like to whine?

Re:Yeah, but did they fix the leaks?

[X0563511]

Why are you doing that? Don't leave the browser open when you're not using it. Don't leave a page open when you're not using it. Problem solved.

Window Drag Handle

[darkain]

I'm still waiting for the window drag handle to be the entire "glass" area at the top, and not just the top X% of it. Ever since Windows Vista/7, Microsoft has made it a defacto standard that any part of a window that is "glass" is a drag handle, and Chrome does this nicely. It is very annoying to not have a visual indicator of where the drag handle starts/stops, and more annoying to have that empty glass space become more or less "useless" if the browser isn't full of a million tabs.

Re:Window Drag Handle

[AxeMurder]

Ummm I can click anywhere in the glass on Firefox, including the area that fills with tabs and it drags just fine...

Re:Window Drag Handle

[darkain]

Well, if it works on some systems and not others, it is still a bug.

Re:Window Drag Handle

[rjim]

I'm noticing behavior in Firefox 17 where all the glass can be used as a drag handle when the menu bar is disabled. However, when the menu bar is enabled, only the window title glass can be used as a drag handle. The glass surrounding the menu bar and the tabs don't respond to mouse dragging.

There is a bug report 667234 regarding this issue for Thunderbird which has been addressed (IIRC), but it also applies to Firefox.

Re:Window Drag Handle

[X0563511]

Seamonkey seems to behave normally.

Re:Window Drag Handle

[Anonymous Coward]

I still miss 'Linux' (don't know whether it was something special by the window manager or not) where you could alt+click-drag a window in any position of the window. Similarly you can easily resize a window. That way you don't have to go to the titlebar or borders of a window to move/shape it. After using Linux for quite some years and running Windows again, I could not live without a tool that did that for me on Windows;

- Windows: http://code.google.com/p/altdrag/
- OS X: http://superuser.com/questions/53051/altclick-drag-window-resizing-on-a-mac-similar-to-x-windows

That was quick.

[UltraZelda64]

So, Firefox is already set to get out of its teens. What's next? Firefox 30 late 2013? Version 100 by 2016?
Pretty soon they're going to have to do something, because people are going to get fucking sick of counting up an entire version pointlessly every god damn month.

Re:That was quick.

[Anonymous Coward]

https://addons.mozilla.org/en-US/firefox/addon/version-sanitiser/

Re:That was quick.

[sateh]

I totally agree! These numbers are getting out of hand!

Don't get me wrong, I love the rapid release cycle and all the new features and improvements they put in those release, but those version numbers are driving me insane.

I honestly don't know how many more version number increments I can deal with. I've totally had it. I'm done with it.

Maybe they can switch to fuzzy cat/fox names like Apple does?

This bizarre version number incrementing has to stop.

Re:That was quick.

[SpectreBlofeld]

Just use month/years in your release numbers. Seriously. A version of Firefox released in december 2012 is firefox 12.12. Simple. Gives you a clear idea of its modernity. No confusion or consternation.

If anyone else needs simple solutions, I'll be over here sitting in the Idea Booth.

Re:That was quick.

[mister_playboy]

This should become an obligatory link in all future Slashdot FF stories.

Thanks.

Re:That was quick.

[X0563511]

So long as you don't tie stupid names like Crackwhore Canary that's perfectly acceptable.

Re:That was quick.

[Ash-Fox]

Don't get me wrong, I love the rapid release cycle and all the new features and improvements they put in those release, but those version numbers are driving me insane.

It's just a version number, take a chill pill. At least it's clear if you're on the latest version of Firefox this way, compared to what I have to deal with in my own software (which is date based), where people just keep reciting the major version number (telling me the year) whenever there is a problem (fairly unhelpful).

Version 20?

[Anonymous Coward]

Software at version numbers above 10 bear two sets of image problems, especially when there is a new major release very other month.

a) they are perceived as old and out dated. Old software means heavy. Heavy means slow. Nobody wants slow software.
b) they indicate a broken lifecycle model. Major versions are meant to signify substantial, sometimes incompatible changes. Nobody in their right mind wants incompatible changes. And nobody believes an organization that claims to have reinvented the wheel every couple of months.

So *please* change this non-sense and start counting minor versions again. There was a time when Mozilla felt that FF could be seen a a product in its infancy at versions 3.x, when all other major browsers boasted versions >6,7,8. Well guess what, they've managed to turn this around and FF is now perceived as the grandfather as opposed to the young hip kid. Nice job.

Re:Version 20?

[snl2587]

To be fair, Chrome is at version 23 now. I'd say no one thinks of that browser as the "grandfather".

Really, people put way too much stock in version numbers, especially for projects with rapid release cycles.

Re:Version 20?

[Anonymous Coward]

So *please* change this non-sense and start counting minor versions again.

Oh, they do, kind of. You see, in a perfect world according to Mozilla, it would go like this:
Firefox 4, 5, 6, 7, 8, 9, 10, 11, 12 ... a new "release" a month.

But here in reality, they've been bitten in the ass by the fact that bugs do exist, so they can't completely do away with point releases. So what it seems like they're doing (and no doubt they don't like it) is:
- Beginning of the month: A major release.
- 2-7 days later: A minor, .1 release do get rid of some major bugs their pointless rapid-release schedule creates.
- A week to a week and a half after that (ie. midway through or toward the end of the month): A minor .2 release to fix additional bugs.
- A week or two after that: A new month starts, a new pointless fucking "release" of Firefox...

Damn, I bet Mozilla Corporation really hates bugs, don't you think? Fuck, if they had their way, they would abolish the decimal from the decimal system because, hey, if it's not a whole number, then it must be inferior!

20???

[Anonymous Coward]

what happened to their numbering system? i was using firefox 4.x waiting for 5. then i took a nap. now firefox 20 is coming out? i rmember similar happened to winamp. i was using a version that worked, was small and didn't try to take over your computer, 3 i think it was. i took a short nap, and when i awoke, version 9 or something was out. maybe its jsut me, but i am gettin the feeling that these large software companies have forgotten how to count. are they in a hurry to reach a certain number? will that version number magically make t he product actually work? example, IE never worked well, always backwards and hard as possible. no matter how 'easy' they say it is. there is no way to stop google's Crome from updating ALL THE TIME. i use it for only 1 site. i cant stop t he updating. i tried to stop the service, and it finds a way around that. crap! then theres opera, same deal with FF, they both get this thing that they have to upload massive amounts of data for no reason to nowhere in cyberspace. just keeps me from using the web. i dont like web browsers.

back to the point

how did they suddenly get to 20? and why is there no truth in advertising? secure browsing? so the server doesn't log what ip is connecting to it?
20, seriously...

Re:20???

[UltraZelda64]

The difference is, Winamp's version numbering actually made sense. There was Winamp 2, and then the flop that brought with it freeform skinning known as Winamp 3. Because Winamp 3 was a failure (I don't know why, I liked it...), Nullsoft decided to go back to the drawing board. They decided to extend the current Winamp at the time (version 2) with plug-ins to handle freeform skinning and various other features they planned in the rewrite (version 3). The result of combining (version) two and (version) three was... drumroll... 5. 2+3=5.

But that happened many years ago; Winamp has been at version 5.xxx ever since. The main problem I see with Winamp was not exactly its version numbering system (although it can be confusing at times with so many digits after a single decimal, though that is not a problem with the current 5.63), but the fact that AOL has for years been adding a whole shitload of garbage to the installer, and it's installed by default, with no sane "full, but without the useless options" setting. And that's not including the endless unchecks to get rid of unwanted sponsored garbage. There's literally so much useless junk installed in a "full" installation, it's actually very difficult to get a nice, light, clean install without loads of crap that you don't need.

I miss the days of being able to select a "full" installation, uncheck maybe 6-12 things, install and be done with it. Now... half the new garbage, I don't even know what it is... and after running Linux for so long (ie. no Winamp, unfortunately), it's even worse trying to figure out what's what. What it needs is more options than just light/minimal and full... all or nothing is just not a good compromise.

Re:20???

[Arker]

Right about the time firefox 4 came out they went full retard. Every time someone fixes a typo they have a new major revision. The ridiculous versioning wasnt the only sign of collapse, although it's hard to miss. I relied on first Netscape then Firefox as my primary browser for well over 10 years but they finally made it suck so bad I even prefer IE. I absolutely loathe IE.

I switched to Opera as my primary browser and although I am far less than 100% happy with that (I would definitely prefer a Free solution) I must admit it works pretty well, and if you set aside the blobware nature of the release as well as notscript appearing to be slightly inferior to noscript, Opera wins every other comparison in my opinion. Personally I put pretty heavy weight on those two things, much more than most people would, yet new Firefox is so bad I just uninstalled it.

Re:20???

[mister_playboy]

The result of combining (version) two and (version) three was... drumroll... 5. 2+3=5.

"nobody wants to see a Winamp 4 skin" is more amusing. :)

Re:20???

[X0563511]

VLC long ago replaced Winamp for me. Does everything I want.

Re:20???

[UltraZelda64]

VLC has been to unstable for me and in general just doesn't tend to work as it's supposed to. In fact, my experience with it has been so consistently bad, I honestly don't get why it's so damn popular.

Re:20???

[X0563511]

Strange, as I have had zero issues and the number of crashes I could count with my fingers. I've been using it for years.

Re:20???

[hobarrera]

I belive winamp 3 was a failure because it used up too many resources in comparison to winamp 2. At the time, winamp 2 was extremely lightweight - and that's what made it so popular.

Re:20???

[UltraZelda64]

Possibly, I'm not really sure, but I actually thought it had something to do with the fact that WA3 was such a major departure (IIRC it was a brand new program, it didn't even support classic Winamp skins).

Please fix plugins.exe, thank you

[Anonymous Coward]

No really, we've all been wanting this to be fixed or at least have a fresh face and get promises every year but it never happens.

Re:Please fix plugins.exe, thank you

[Hsien-Ko]

I wish some day, the very word any browser uses to market themselves, that word being "fast"... actually meant something.

Tor for the highly privacy seeking or the paranoid

[girlinatrainingbra]

Dude, at that level of paranoia, go ahead and install Tor ( http://en.wikipedia.org/wiki/Tor_(anonymity_network)) and use the version of Firefox it comes with to route your requests through the onion router.

Speaking of long standing bugs...

[Anonymous Coward]

Maybe if they do implement this fix, they'll finally get around to fixing this bug after nearly 11 years and counting!

Re:Speaking of long standing bugs...

[Shikaku]

So... which browser has this "bug" fixed? Because none of them I've found work as intended.

Re:Speaking of long standing bugs...

[Anonymous Coward]

Chrome and Internet Explorer 9 definitely have this bug fixed now in Windows 7.

Wonder what old bugs they'll fix next year...

[Anonymous Coward]

in Firefox 100

Re:Tor for the highly privacy seeking or the paran

[Anonymous Coward]

TERRORIST PEDOPHILE!

The major tablet

[tepples]

Yes, but the summary mentioned major browsers

What's the only browser that comes on the major tablet? The iPad is not just a plurality but a majority.

Re:The major tablet

[Anonymous Coward]

The iPad is not just a plurality but a majority

And PC sales still easily out-strip tablet sales by a factor of three or four every quarter. So iPad Safari is still a minor player.

Anything like KeePass for Linux?

[tepples]

KeePass

Requires .NET Framework 2.0. Or does it work in Mono? What password managers have Linux users found useful?

Re:Anything like KeePass for Linux?

[X0563511]

It does work in mono. It's actually mentioned right on the damn download page.

Grab the .zip version, extract it, and do 'mono KeePass.exe'

Carrier-grade NAT

[tepples]

it's easy enough to collect those data together and make a profile for that IP address and for the various sites hit at the various times of day.

And with the depletion of IPv4 addresses, you eventually end up trying to distinguish among 200 unique visitors behind one carrier-grade NAT.

Re:Carrier-grade NAT

[girlinatrainingbra]

re you eventually end up trying to distinguish among 200 unique visitors behind one carrier-grade NAT. But that's where the browser fingerprinting as described at https://panopticlick.eff.org/ comes in: even if you have javascript disabled, your browser sends along information about your:
-- media types accepted
-- cookies enabled
-- HTTP-accept headers
-- and of course, your user-agent
.
Even behind noscript, my browsing leaks 17.96 bits of information, according to the EFF panopticlick survey for me. If we allow javascript, then this other information can also be gathered:
-- fonts available
-- addons available
-- browser plug-in details
-- pixel dimensions of display
-- color depth of display
-- time zone
.
Allowing javascript leaks 21.18 bits of identifying information. In fact, just the browser-plug in details alone seem to be enough to allow my visit to eff to be recognized as unique from all other browsers that have gone there before. Of course, if you use a combiation like "IE version X" running on "MSWindows $ident", then you're more lkely to be a bit more anonymous than someone running a debian system with a lot of non-common browser plug-ins on Firefox-cutting-edge-version.

Re:Carrier-grade NAT

[Jah-Wren Ryel]

You don't seem to realize that all those bits of identifying information are a lot less useful when some of them are deliberately polluted.

So, carrier-grade NAT to mix traffic between multiple different users plus minor variations in fingerprinting information makes it exceptionally difficult to correlate different website accesses with a unique user. The problem has now gone from one that can be reasonably automated to one that is going to require human judgment, so the risk has gone from being caught up in mass-profiling systems to being specificaly targetted. At that point, you are probably a person of interest to an organization with enough resources to just send in someone to install a network sniffer on your local lan.

Version numbering insanity.

[Anonymous Coward]

I'm going to start my own operating system. The initial release will be Version 10.0, followed after a few days by Version 11.0, then after a few weeks, I'll have hit 20.0 But I think I may make the version numbers logarithmic and base-10 instead, so version 10 is followed by version 100, then by version 1000, etc. After a year following the initial release, you'll here somebody say, "hey, I just upgraded to version 10000000000000000.100000.1000.10, when it will really be version 3.62.010

Can we stop the madness yet? Firefox fucking 20. More like 5.23.10.1

A good feature

[jones_supa]

There is a feature that I very much like in Firefox: if you have ordinary horizontal tab bar and you open a lot of tabs, it makes the tab strip scrollable, instead of squeezing the tabs smaller and smaller. With all other browsers you just end up with tiny tabs which text you can't read, which is horrible.

64

[Vlijmen Fileer]

Maybe that will give them time to finally build and support a 64 bit version.
32 Bit should be the exception nowadays, but apparently they decided to drop (!) any further 64 bit effort...
Yeah, i also still play 8 bit Commodore 64 games on my computer, perfectly normal :)
If somebody understands that 64 bit killoff please explain to me...

Re:64

[GeorgeWright]

The issue is not so much that 64-bit is dropped or deemed unimportant; the issue is that Mozilla as a corporation has limited resources to devote to 64-bit Windows builds.

Basically, the main blockers are:

- Plugins. 64-bit plugins on Windows are still not 100% and there currently isn't a way of loading 32-bit plugins in a 64-bit Firefox. Yes, ideally Firefox would have this, but again - resources.
- Testing. It'd add another column onto the test matrix which is a non-negligible cost overhead to the release engineering guys (who are already massively overworked as it is). For a feature that Mozilla as a corporation isn't prioritising, this burden on releng is unacceptable.
- Benefits. The benefits from switching to 64-bit code aren't actually as plentiful as you might think. Basically, the major one is that Firefox would be able to address more memory instead of limiting at 4GB. However, project memshrink (https://areweslimyet.com/) has been working pretty hard on reducing the memory footprint of Firefox, which is the correct fix in this case *except* in the case of those people who are using hundreds of tabs amongst several browser windows. Unfortunately, they're in a small enough niche that, again, Mozilla can't dedicate resources towards a black hole to accommodate them (yet).

The myths that building the browser in 64-bit results in somehow faster code, or more efficient execution are just that - myths. In fact, in a lot of cases, code that was written for x86 that's been recompiled in x86_64 results in slower code because your pointer sizes are twice as big and you start smashing through your CPU cache more quickly.

For a more detailed description on the cost supporting these has incurred, Ben Hearsum wrote a very good post on dev-platform: https://groups.google.com/forum/#!msg/mozilla.dev.platform/UOn1iQetA5w/-DNzeCOMdrcJ

TL;DR - it's not a question of whether Mozilla wants to do this (they do); it's a question of whether the resource/benefit tradeoff makes sense at this time.

I would also like to remind people that software engineers aren't just assets that can be moved arbitrarily from project to project, so all those people saying "stop working on X feature and concentrate on 64-bit instead" - stop thinking like that. That's how bad managers are made.

Re:64

[Ash-Fox]

- Plugins. 64-bit plugins on Windows are still not 100% and there currently isn't a way of loading 32-bit plugins in a 64-bit Firefox.

Yes there is, a 32bit plugin-container.exe with 64bit firefox.exe.

Nonsense owl

[Dwonis]

What has been seen cannot be unseen.

You mean non-private tabs should be labeled?

[dutchwhizzman]

I really like MicroSofts decision to make do-not-track default in IE10. The same should be set for other web sites and people should get a large popup for each site that requires them to allow tracking (yes, per site base exceptions).

Also, It should be easy to configure Tor or other proxies for do-not-track sessions, or even per domain/site that's being visited. Storing IP addresses will often make tracking still feasible and often rather simple. FaceBook keeps "ghost profiles" for people based solely on cookies and IP addresses, I'm fairly certain Google does the same.

If i was in the USA I'd have the constitutional right to be left alone and a lot of companies are not honouring that right. At least give me the tools to make it hard for them to violate my privacy without making it a technical nightmare to do so if I still want to use the Internet.

Re:You mean non-private tabs should be labeled?

[dolmen.fr]

people should get a large popup for each site that requires them to allow tracking

You do not seem to understand what is "do-not-track". This is just a declaration of the user to "Do not track me". Nothing technically forces the site to not track you. In fact, most of the advertising agencies that say they implement it say that they will just not use the tracking data to show you personnalized ads. You will get only generic ads not based on your profile, but will still be profiled.
Worse, if "Do-not-Tack" is enabled in Firefox, this adds an additional bit of valuable information to your tracking profile: the fact that you are privacy-aware.

So don't count on Do-not-Track only if you want that your privacy be respected. Use additional tools such as AdBlock Plus and Ghostery and allow cookies only siste by site with a whitelist.

Lets hope they fix the other bug too

[clickclickdrone]

When browsing in private mode, if you then switch back to normal, when you close Firefox and as the tabs shut down, you usually get one of the supposedly private pages flash up briefly. Clearly not all buffers are cleared when exiting privacy mode.

no remote

[Anonymous Coward]

ever heard of firefox.exe -no-remote -p

Oooh! Crappy Title

[blade8086]

Saying this is 'broken' (by saying it is 'fixed' now) because it doesn't let you mix windows is a total crap article title.

Slashdot is getting as bad as information week with the sensationalist crap titles

Weird bug introduced with FF17?

[Angstroem]

After upgrading to FF17, the browser started crashing whenever the file dialog pops up for up- or downloads. It doesn't matter whether the file is double-clicked or selected via "Ok" -- the browser instantly crashes afterwards, sending home a crash report.

That very behavior also takes place with FF18. When running the with strace, the crash does not happen. FF16, in term, does not show this bug.

Any idea how to hunt this down or even fix it? (I'm running Ubuntu 10.04LTS w/ Trinity/KDE3.5)

Re:Weird bug introduced with FF17?

[Ash-Fox]

I'm running Ubuntu 10.04LTS w/ Trinity/KDE3.5

That version of Ubuntu is no longer supported by Canonical. I'm not really sure it's worth time debugging an issue on an unsupported operating system.

Re:Weird bug introduced with FF17?

[Angstroem]

Anything helpful to say?

It still is supported, btw, Desktop until 04/2013, Server even until 05/2015. So much for that. (You do know what LTS stands for, right?)

Besides, behavior like this

• FF will not crash, when strace'd.
• FF will also not crash, when started via "ssh localhost".

does not sound like some "get your .mozilla directory fixed" kind of problem (which doesn't help btw.), but some race or whatever mess going on.

Re:Weird bug introduced with FF17?

[Ash-Fox]

It still is supported

Why'd they remove it from the download page? Seems kind of weird to remove supported versions from the download page.

Re:Weird bug introduced with FF17?

[Angstroem]

Why do I still get updates? Seems kind of weird to me to update unsupported versions ...

In any case: It's not even related to Ubuntu versioning (discussing the Ubuntu version was a moot point anyway, but some people like to rather jump on these rather than dealing with the problem).

As I found out meanwhile, it's a problem of the Trinity desktop environment regarding mapping GTK widgets to QT widgets, potentially affecting every user of that desktop environment, no matter what recent or not-so-recent Linux distribution they are using ... Removing the kgtk-kde3 package solves the issue.

Workaround

[advantis]

I created a new profile in the profile manager of Firefox and wrote a tiny script that I called '~/bin/privatefox' with this command in it:

#!/bin/bash
exec /usr/bin/firefox -P new -no-remote -private

Bug fixed :)

I've switched away from Chrome since Chrome started adding my incognito cookie and javascript exceptions to the persistent list. Everything else Chrome did to tick me off was tolerable, but the leaking of incognito exceptions... GTFO

Bully for you, Mozilla

[Rogerborg]

But since my #1 required feature is "not randomly crashing out, sometimes ten seconds after startup", I'm now writing this via Chrome. Too bad you couldn't stick to your original vision: small, fast, stable.

Re:Bully for you, Mozilla

[Ash-Fox]

How do I reproduce your issue?