Slashdot Mirror
Microsoft Accuses Google of Violating Internet Explorer's Privacy Settings
New submitter Dupple writes with a followup to Friday's news that Google was bypassing Safari's privacy settings. Now, Microsoft's Internet Explorer blog has a post accusing Google of doing the same thing (in a different way) to Internet Explorer. Quoting:
"By default, IE blocks third-party cookies unless the site presents a P3P Compact Policy Statement indicating how the site will use the cookie and that the site’s use does not include tracking the user. Google’s P3P policy causes Internet Explorer to accept Google’s cookies even though the policy does not state Google’s intent. P3P, an official recommendation of the W3C Web standards body, is a Web technology that all browsers and sites can support. Sites use P3P to describe how they intend to use cookies and user information. By supporting P3P, browsers can block or allow cookies to honor user privacy preferences with respect to the site’s stated intentions. ... Technically, Google utilizes a nuance in the P3P specification that has the effect of bypassing user preferences about cookies. The P3P specification (in an attempt to leave room for future advances in privacy policies) states that browsers should ignore any undefined policies they encounter. Google sends a P3P policy that fails to inform the browser about Google’s use of cookies and user information. Google’s P3P policy is actually a statement that it is not a P3P policy."
In other words, if your server delivers a garbage or blank P3P header, the browser assumes there are no privacy implications? Sounds like a hole in the standard to me, such headers should be ignored IMO. Though Google really should have tested this properly with all browsers before deploying it in production it sounds to me like an oopsie, not at all like the Safari thing.
Google has been claiming "oopsies" almost weekly over the last couple months. In this case they put this in their policy: 'P3P: CP="This is not a P3P policy! See http://www.google.com/support/accounts/bin/answer.py?hl=en&answer=151657 for more info."' in what is meant to be a machine-readable field. Following the spec would have been easy-- omit the field altogether. Instead Google violates the spec in a way that benefits them. It's possible Google is just really incompetent over all these "oopsies", but they sure try to represent themselves as a company with above-average engineers. It has to be one or the other.
E pluribus unum
funny: I'll have to remember this to rub their noses in it, next time I run into a googler.
or, if they interview me, I'll ask THEM: "so, what is the proper response to a machine parsable field? TLV things or human-intended english? please support your answer."
sigh. I cannot give google a pass. they act like god's gift to networking yet they make 'mistakes' like this? sorry, but I don't buy it.
--
"It is now safe to switch off your computer."
When I was configuring P3P for Mozilla/Firefox, it distinguished between what exactly the P3P policy was stating. If the site didn't say in the P3P policy what it was doing with cookies, Firefox assumed the worst. It seems to me that if the IE devs were dumb enough to stop after seeing a P3P policy presented and didn't bother checking what it said, or if they assumed a lack of a statement indicated respect for privacy, that's a failure in IE. The code needs to start out assuming personal information is collected and used without consent, and then upgrade only if the P3P header specifically says something better. It's not like that's hard to implement.
Then again, we've seen similar problems in Microsoft software time and time again: they assume the best (input's valid, doesn't contain special characters, etc.) until they detect otherwise, even though best practices say to do the opposite (assume input's invalid until analyzed and proven correct, list the known non-special characters and filter out or escape everything not in that list).
P3P sounds like a stupid idea anyway. How does it protect user privacy if something as trivial as the attack described above totally defeats it?
If the IE or Safari teams really cared about user privacy, they would be more strict about allowing sites to set or read cookies. This is just an excuse for Microsoft and Apple to publicly bash one of their competitors while continuing to not give two hoots about their users.
Palm trees and 8
Google has been claiming "oopsies" almost weekly over the last couple months. In this case they put this in their policy: 'P3P: CP="This is not a P3P policy! See http://www.google.com/support/accounts/bin/answer.py?hl=en&answer=151657 for more info."' in what is meant to be a machine-readable field. Following the spec would have been easy-- omit the field altogether. Instead Google violates the spec in a way that benefits them. It's possible Google is just really incompetent over all these "oopsies", but they sure try to represent themselves as a company with above-average engineers. It has to be one or the other.
Can't say I really can fault Google for this. Explaining why would require an understanding of how P3P 1.0 objects are configured and how limited those types really are.
P3P 1.1 work has stalled (albeit in provisionally final state) and is likely to not restart; in its absence is P3P 1.0 which exists firmly in the world-as-it-was of 2000/2001. It covers cookies and certain types of form transmission, but doesn't cover privacy aspects of other types of persistent data, new transmission protocols (like SPDY), advanced caching techniques, or HTML5 storage. Technology has advanced past the point that P3P 1.0 is useful -- and quite simply, it's doubtful it ever really was. If you visit the link Google supplies it explains some of their reasoning, and it's pretty dang valid for a post-2007 view of the Web.
Those chucking bombs over this would be better served to focus their efforts on either modernizing or replacing P3P 1.0 -- or, better yet, trying something radically different like PRIME or Policy-Aware-Web tried to do.
In IE iframes will block cookies if you don't have the right P3P policy. There where other bugs that would prevent your site's cookies from being read.
I've "faked" a P3P header just so users of certain IE browser versions could use my site.
At the end of the day the standard is a proposal and only MS thinks it's worth a hill of beans.