Mozilla Delays Default Third-Party Cookie Blocking In Firefox 106
hypnosec writes "Mozilla is not going ahead with its plans to block third-party cookies by default in the Beta version of its upcoming Firefox 22. Mozilla needs more time to analyze the outcome of blocking these cookies. The non-profit organization released Firefox Aurora on April 5 with a patch by Jonathan Mayer built into it which would only allow cookies from those websites which the user has visited. The patch would block the ones from sites which hadn't been visited yet. The reason for Mozilla's change in plans is that they're currently looking into 'false positives.' If a user visits one part of a group of site, cookies from that part will be allowed, but cookies from related sites in the group may be blocked, and they're worried it will create a poor user experience. On the other side of the coin, there are 'false negatives.' Just because a user may have visited a particular site doesn't mean she is comfortable with the idea of being tracked."
Ummmm.. (Score:2, Redundant)
I found that usually the rest of the 'group' of sites usually host static images and other media so theres usually no reason these sites even need cookies.
Re:Ummmm.. (Score:5, Insightful)
The only thing 3rd party cookies are useful for is tracking you. Anyone who says otherwise makes their living out of stripping you of your privacy.
Re: (Score:3)
not for sites that use 3rd party commenting systems, et al. discus
i'm not saying i like this implementation, but surely this firefox feature will break this. i see it all the time using cookie monster plugin
Re: (Score:2)
that should say "disqus"
Re: (Score:1)
Re: (Score:2)
What about a Safari hack?
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Those sites just aren't doing it properly. By comparison, OpenID works fine without 3rd-party cookies. So do other commercial Single Sign On (SSO) solutions.
Re:Ummmm.. (Score:4, Interesting)
I block third party cookies. What happens when I land on a page that uses Disqus? I have to coax the browser to log me in to Disqus. And - that is just the way I want things to be. Disqus doesn't need to know where I browse, or what I'm reading, unless and until I CHOOSE to summon Disqus.
Children, if you're going to dabble in the arcane arts, you must learn to control those demons - or you will find that the demons control YOU!
Re: (Score:2)
Re: (Score:2)
Subdomains are not affected. They are first party cookies not third party.
Re: (Score:2)
Reading fail! The summary itself says the policy is being delayed because of false positives, ie, things that they are blocking that is causing users to complain.
This is exactly what happened with Safari. Somebody decides that "privacy" can be viewed exclusively through the lens of particular technologies, that advertising is bad and they will "save the users" fr
Re: (Score:2)
What about services like Google, where you can sign into YouTube?
that would be pretty great, if it isolated the cookies so that you would have different login for youtube..
and different login for google.com from gmail. it would be sweeeeeeeeeeeeeet. because google is shit poor guessing which account I want to use, if I want finnish localization or not and so forth.
Re: (Score:2)
Not always, some sites put user content on another domains, For example if those cookies are blocked you will not be able to download file attachments from Gmail
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Re: (Score:1)
I am a free slashdotter. I will not be modded, blogged, DRM'd, patented, podcasted or RFID'd. My life is my own.
ahh, but you will be modded... modded DOWN, that is!
They can do more (Score:1)
No issue. (Score:5, Insightful)
I have third-party cookies (indeed, all cookies, except those from domains specifically whitelisted) blocked. I've never noticed a problem with blocking third-party cookies. I have a heck of a lot more issues with third-party JavaScript (people using Google-hosted or similar JQuery for example).
So, Firefox, take note, there are not going to be any problems for the vast majority of people.
(I use CookieMonster, it works real nice like.)
Re:No issue. (Score:5, Insightful)
I have third-party cookies (indeed, all cookies, except those from domains specifically whitelisted) blocked. I've never noticed a problem with blocking third-party cookies. I have a heck of a lot more issues with third-party JavaScript (people using Google-hosted or similar JQuery for example).
So, Firefox, take note, there are not going to be any problems for the vast majority of people.
I find it laughable that one of Mozilla's excuses for not doing this is "they're worried it will create a poor user experience". Over the last few years Mozilla has made a number of changes to Firefox that were met with user complaints, and continue to be a source of user complaints and the developer's response is always a resounding "fuck you".
As far as cookies go, don't forget that Mozilla currently gets $300 Million a year from Google, whose entire gazillion-dollar-a-year business model is based on tracking people.
Re: (Score:2)
No, it hasn't been blocking third party cookies for years. This is the core of why such policies are a bad idea. It says it blocks third party cookies, but there are actually lots of exceptions [webkit.org] to that rule in order to avoid as the summary says "false positives". You can read about what really happened with Google on Lauren Weinstein's blog [vortex.com], it's very different to how you paint it (there was no "trying to circumvent" involved).
Re: (Score:2)
Oh come on. You think they are actively hostile to user experience?
It's one thing to think they are wrong about everything to do with user experience. Another to think that it's not a motivation for them.
Re: (Score:2)
so you have exactly the same usage as "the vast majority" of people?
Re: (Score:2)
I block 3rd party cookies by default (Score:3, Interesting)
The only thing I notice is I can't comment on Disqus (a 3rd party site that handles comments on some blogs). I don't care about it, block them.
Firefox should focus on privacy, its their usp. Google for example, doesn't let you accept cookies for the 'session only', you accept them or not on their Android browser. At some point you have to accept cookies, so this is a fake choice, you'll end up with that feature always on because its too much fuss to turn it on when its needed.
Firefox 'accept cookies for session only' option is my default, it lets me work on sites that use cookies, but throws them away when I close the browser.
Things like this are why I use Firefox.
Re: (Score:2)
a "session cookie" is literally a different type of cookie as sent from the server.
aka "transient cookies"
they are typically to follow you around that "session" on the website. not the session in the browser.
but yes, the browsers happen to trash them when you exit. but if you run out of cookie memory, they'll page them out or scrap the old ones entirely.
Re: (Score:2)
Firefox has the ability to treat all cookies as session cookies. It's a useful feature.
Re: (Score:3)
The problem Mozilla finds itself in now is that since a large number of people use it, it's harder to make such changes. You might think this is a no brainer, but people who use Disqus or other services which are built around third party cookies, of which there are many, might disagree with their page or sites they visit breaking and either not knowing the cause, or not being knowledgeable enough to fix it.
This wasn't such a problem when using Firefox was more of a techie thing. Now they need to tread light
Re: (Score:2)
The easy way out of that is to include a default white list. Which of course should be open and configurable (add/remove sites from whitelist; disable whitelist completely).
I have been blocking third-party cookies since I found out Mozilla (yes, back then) allowed me to do so. And when switching to Firefox when it became useful I did the same. Never had any problems with it.
Enable it already! (Score:1)
Mozilla should enable it, poorly designed websites need to get in line and clean up their messy cross linking stuff.
Then again I'm an old fart who misses the 90s world wide web with its simple and direct approach without java and flash.
Re: (Score:2)
Then again I'm an old fart who misses the 90s world wide web with its simple and direct approach without java and flash.
You're talking about 1992-1995 then?
The web was so small then that it'd probably fit on a modern desktop hard drive!
Re: (Score:2)
Re: (Score:2)
As someone who experienced the web in the early 90s and has continued to use it since, I prefer today's version.
If someone wants to see the 98% of crap, they can easily watch their cat videos. I personally find it easy enough to drown out the noise.
I've been blocking 3P cookies for years (Score:5, Insightful)
and have never noticed a problem. This has always struck me as a no-brainer and it's annoyed the hell out of me that I have to modify the setting on every platform for each of my five family members.
I can't wait for them to change the default behaviour and I'll be very interested to see if they uncover any side effects that could conceivably be considered undesirable by the user.
My biggest worry is what the websites might do to circumvent the change.
Re: (Score:2)
Re: (Score:1)
Don't worry, we still track people who don't clear history through the use of 1x1 pixel images, display:none anchors, and javascript checking the visited property, and server side analysis of whether or not your browser has cached dynamically generated names.
A bit of jquery doing some XHR & a binary search on :visited goes a long way to getting and setting unique id's.
(block those third party javascripts...)
Re: (Score:2)
My biggest worry is what the websites might do to circumvent the change.
Flash cookies have the potential for great evil.
Bullshit (Score:3, Interesting)
They caved to pressure from advertisers
Re: (Score:1)
Actually they pay for product placement, though I'll grant that the developers can skim a little off the top.
Re: (Score:1)
Re: (Score:1)
Why not?
'Cause I for one am not a she, it should've been 'they'.
Re: (Score:2)
Why not?
'Cause I for one am not a she, it should've been 'they'.
Because there are no women on the Internet. Only men pretending to be women. Everyone knows that.
Re: (Score:2)
Because there are no women on the Internet. Only men pretending to be women.
That may have been true, up until the advent of Pintrist.
Re: (Score:3)
I'd give you mod points if I had them: +1 for singular they. Using a gendered word for a person of unknown gender is dumb, and singular they is a perfectly reasonable workaround.
Hasn't been a problem for me (Score:4, Insightful)
Disqus is the problem (Score:5, Insightful)
There is one very large product that relies on 3rd-party cookies: Disqus. It is used by a lot of popular sites such as Thingiverse and StackOverflow. Disqus simply needs to fix the problem. There is actually a discussion on StackOverflow about this: http://meta.stackoverflow.com/questions/126764/why-does-registration-require-third-party-cookies-to-be-enabled [stackoverflow.com]
The last time I looked at it it claimed the problem was fixed, but I just now tried to register and it says this:
Third Party Cookies Appear To Be Disabled
This site depends on third-party cookies, please add an exception for https://openid.stackexchange.com/ [stackexchange.com].
Re: (Score:2)
Re: (Score:3)
I'm sure that I'm naive, but can't they just run a little script that detects the cookie, and if not found asks the user to click a link to enable comments? Then the user would have visited the site (Disqus) and the Firefox block would be removed forever forward.
Re: (Score:3)
...by providing a login button that does the redirect dance back and forth.
that's how such system would have been meant to be used in the first place. of course they wouldn't get random visitor tracking as their business model that way but meh, those are the breaks.
Re: (Score:2)
Re: (Score:3)
You're going to have go into more detail. At the very least:
1. Explain how having to reload the page (Jump to Disqus and then bounce back) going to be positive for the user's experience. I certainly don't see how it would be remotely positive.
2. How is this going to work without the host installing something on their server? As I said, a selling point of Disqus is that it doesn't need anything on the hosts' server at all, just some boiler plate HTML that inserts the Disqus Javascript script.
I don't see your solution as being "How they should have done it all along". It's inefficient, kludgy, and fails the ease-of-installation test.
the solution can entirely be javascript included in the page source, mostly as it is. the only thing that would break with breaking of cross site cookies/storage would be that you wouldn't be already logged in when you go to another disqus enabled site.
though, admittedly, I viewed it as a bonus that the login is intrusive and the user has to visit the site of the service he's authenticating to. think of it as one-click-sign-on instead of already logged in when you go to a new site single-sign-on.. the page
Re: (Score:2)
As far as I know, this problem is not intrinsic to the design of OpenID itself. It is designed to use redirections to basically pass data back and forth between the OpenID provider and the web site. I don't think other OpenID implementations have this problem. I don't know enough details about OpenID to describe exactly how, but I think the answer comes down to "follow the specification" and "do what other sites do."
Re:Disqus is the problem (Score:5, Interesting)
Really, who cares about Disqus? I immediately added a filter for them to adblock when I noticed a suggested thread 'Soandsos baby mama' on the AngularJS API docs.
Anyone the least bit privacy conscious should be blocking Disqus along with G+, Facebook, Twitter, etc. on their party sites.
Re: (Score:3)
Disqus is only an example but the point is that there are "third party web components" that will be effected by a platform wide block. For cases like this it is good to give legitimate component software a "transitional grace period" to move away from the deprecated behavior before locking it out from modern versions onward.
I view control over "third party sources" in web content as a serious security issue but I also admit that I don't know the full ramifications of an outright ban either where taking the
TLS vee1.1 and 1.2 (Score:2)
Now that your delaying third party cookies hows about using the extra time to add support for new versions of TLS? Why is IE the only browser supporting TLS v1.1 and 1.2? Even chrome supports 1.1 and it uses NSS too.
We are still dealing with a few lazy nessus wielding compliance jackasses invoking BEAST to get EVERYONE to use broken RC4 ciphers because a few users still have not updated their browsers to fix a known problem solved over a decade ago.
It would be nice to one day be in a position to start to
Are they really that bad? (Score:2)
All this Mozilla change means to me is that a l
Re: (Score:2)
From the cookie monster himself (Score:1)
IE used to do it right (Score:1)
Long before Firefox existed, IE6 allowed blocking 3rd party cookies.
However, it would display an icon on the status-bar and when I clicked on it, it would show me a list of blocked items and allow me to white-list them.
Why can't FF do the same? Or is there an extension to do it?
Analyze this, Mozilla. (Score:3)
I've been blocking third-party cookies for years with absolutely no hint of any site failing to load correctly. If there is ever a problem, it is scripting, and choosing to disable NoScript on one or more sites typically sorts that out. Get the advertising industry's dick out of your ass and just fucking block third-party cookies already, Mozilla. It should have been done a hell of a long time ago. This new versioning system can be so amazingly retarded; we're at Firefox 21 already, already talking about Firefox 22, and Mozilla is still dragging their feet around on something as simple as the default fucking setting of a checkbox regarding third-party cookies. Talk about illusion of progress! You know that by this point, Mozilla no longer gives a shit about their actual users and seems to have their priorities in the advertisers; otherwise there would be no question, no delay. Why hasn't there been a fork of Firefox yet? IMO, it's been needing one free of Mozilla's bullshit since the 2.x.x days at the very least, or possible 3.x. This is getting ridiculous.
Once, long ago (Score:3)
Cookies used to be really easy to deal with using mozilla, it wrote them all to cookies.txt. You just went in, deleted cookies.txt once, then mkdir cookies.txt. Then set it to allow cookies across the board. All websites worked fine, but anytime you restarted the browser they were all gone. Not 100% ideal but still a quick and relatively foolproof way to assert some sanity. So of course they changed that.
Now... let me get this straight, they are thinking about maybe, eventually, blocking third party cookies by default. Better late than never I guess, but it seems pathetic both in timing and scope as well, since they appear to be worried only about cookies(!) rather than scripting. Third party scripts are a much bigger problem. Both cases should have been blocked by default 10 years or more ago. At this point, yes, I would imagine some problems.
Re: (Score:1)
Domain weirdness? (Score:2)
I see in Eich's comment where he talks about a site "foo.com" including content from a separate domain "foocdn.com" belonging to the same company. My question is why they're using a separate domain? Why not "cdn.foo.com" which would automatically indicate that this domain's part of "foo.com". Or is this a case of "Doc, I don't want to stop hitting myself in the head with a hammer. I just want you to make it stop hurting."?
Re: (Score:2)
Because the cdn is probably (hosted at) a different company; not hosted on your own server (that has the database and provides all dynamic content and user logins and whatnot - keeping all that important information in house), but that seriously bandwidth-eating stuff comes from say Amazon or some cloud server.
Now why such foocdn would need cookies, that'd be the real question. That's supposed to be static content, downloaded via direct links in the main html.
Re: (Score:2)
And? I delegate cdn.foo.com to the CDN company and let them assign names as they see fit. It's not magic, how do you think the root nameservers delegated foo.com to Foo's nameservers in the first place? C'mon, this is DNS 101, the stuff you were supposed to know before you got your first domain.