Home
Posted By: alternaut Web privacy: cookies vs browser specs - 10/04/11 04:12 PM
Browsing my daily sites I just came across Panopticlick - How Unique, and Trackable, Is Your Browser?, offering a (free) service that may be of interest to FineTunedMac users. As the site explains: 'Is your browser configuration rare or unique? If so, web sites may be able to track you, even if you limit or disable cookies.'

I vaguely recall someone (tacit?) mentioning the unique character of one's browser 'footprint' in the context of cookie scrubbing efforts to minimize one's web profile. With Panopticlick you can now easily get some specifics on your browser(s), that could conceivably help to adjust things back to lesser uniqueness. Of course, there are other ways of doing this, but for those who don't or can't go that route Panopticlick may be useful.
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 05:40 PM
Thanks for the link; I'll have to give it a bit of a read when I've got more time.

In the meantime, though...

Quote:
Your browser fingerprint appears to be unique among the 1,789,368 tested so far.

Currently, we estimate that your browser has a fingerprint that conveys at least 20.77 bits of identifying information.

I took a quick look, but I didn't see anything that put the "20.77 bits of identifying information" into any context.
Posted By: ryck Re: Web privacy: cookies vs browser specs - 10/04/11 07:17 PM
I had exactly the same result...." browser fingerprint appears to be unique among the 1,789,368 tested so far....... a fingerprint that conveys at least 20.77 bits of identifying information".

I also looked quickly through the documentation (although much was over my head) hoping to find either context or some ideas about improving the numbers but was not successful.

ryck
Posted By: jchuzi Re: Web privacy: cookies vs browser specs - 10/04/11 07:35 PM
Ditto
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 07:49 PM
Originally Posted By: ryck
I had exactly the same result...." browser fingerprint appears to be unique among the 1,789,368 tested so far....... a fingerprint that conveys at least 20.77 bits of identifying information".

I also looked quickly through the documentation (although much was over my head) hoping to find either context or some ideas about improving the numbers but was not successful.

ryck

Well, how on Earth could we both be among the same group of 1,789,368? That number should have gone up 1, at the very least (assuming it changes dynamically).

Our both being unique doesn't sound all that unlikely, but both our browsers "convey(ing) at least 20.77 bits of identifying information" sounds a bit too coincidental.

I need to know how many permutations and combinations there are of the criteria considered in making those determinations and where 20.77 stands in the scale of results in order to put our results into context.

As I see it at the moment, we haven't learned anything!

And, most important, I'd like to know what I can do to make myself melt into the mix; being unique makes me a target.
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 07:51 PM
Originally Posted By: jchuzi
Ditto

"Spuriouser and spuriouser," said Alice?
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 08:07 PM
You've been around about as long as I have, Jon; does this story ring a bell in your head? I mean the part about multiple users getting the exact same results in a "test" that, seemingly, should have returned some differences, that is.
Posted By: jchuzi Re: Web privacy: cookies vs browser specs - 10/04/11 08:24 PM
Something does, indeed, ring a bell but I can't recall what it is. The test's list of plugins is accurate, however, as is the info about my display.
Posted By: Pendragon Re: Web privacy: cookies vs browser specs - 10/04/11 09:01 PM
And from here...

Your browser fingerprint appears to be unique among the 1,789,696 tested so far.

Currently, we estimate that your browser has a fingerprint that conveys at least 20.77 bits of identifying information.

Alas, I no knot what conclusions to draw from all this. confused



Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 09:03 PM
Originally Posted By: Pendragon
And from here...

Your browser fingerprint appears to be unique among the 1,789,696 tested so far.

Currently, we estimate that your browser has a fingerprint that conveys at least 20.77 bits of identifying information.

Alas, I no knot what conclusions to draw from all this. confused

Nor I, but at least the number of those tested so far has changed, if not the test results.
Posted By: Bensheim Re: Web privacy: cookies vs browser specs - 10/04/11 09:41 PM
I got.....

Within our dataset of several million visitors, only one in 894,868 browsers have the same fingerprint as yours.

Currently, we estimate that your browser has a fingerprint that conveys 19.77 bits of identifying information.
----------------------------------------------
Don't believe a word of it.
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 09:47 PM
Originally Posted By: Bensheim
I got.....

Within our dataset of several million visitors, only one in 894,868 browsers have the same fingerprint as yours.

Currently, we estimate that your browser has a fingerprint that conveys 19.77 bits of identifying information.
----------------------------------------------
Don't believe a word of it.

Interesting that your results are worded differently than the others; I wonder if you got into a different database...maybe because you're there and we're here?

Edit: xx.77 is a popular number, no?
Posted By: dkmarsh Re: Web privacy: cookies vs browser specs - 10/04/11 09:59 PM

Since 1 in 894,868 is the same as 2 in 1,789,736, it seems as if your browser configuration has a doppelgänger out there somewhere. Perhaps one of your other iMacs?

The fingerprint conveys one less bit of identifying information than those of the other contributors to the thread, namely, which of the two computers that browser is running on! laugh
Posted By: tacit Re: Web privacy: cookies vs browser specs - 10/04/11 10:07 PM
Originally Posted By: Pendragon
And from here...

Your browser fingerprint appears to be unique among the 1,789,696 tested so far.

Currently, we estimate that your browser has a fingerprint that conveys at least 20.77 bits of identifying information.

Alas, I no knot what conclusions to draw from all this. confused


My understanding is that the number 1,789,696 refers not to the total number of people who have visited the site, but to the number of unique browser configurations which have been recorded. Since many visitors may have the same configuration, it's not surprising that number doesn't change often. In fact, it's a little surprising the number is that high; it means they have seen 1,789,696 combinations of browser type and version, operating system type and version, browser plugins, and so on.

The "20.77 bits of information" is referring to binary bits. If you take a snapshot of the information that is being returned, you end up with uniquely identifying information equivalent to a binary number between 20 and 21 bits long. That corresponds to a pattern of information that can identify you down to a precision of 1 in approximately 1,048,575. Or, to put it another way, if a Web site sees that particular combination of variables, the odds that it is someone OTHER THAN you are 1,048,575 to one against.

It's kind of like the way people are identified by their DNA. We don't sequence someone's entire genome when we do forensic DNA analysis; instead, we look at only certain combinations of genetic markers. It is theoretically possible that someone else could have the same combination of markers you have, just by pure random chance; the odds that someone else walking around has the same combinations of markers as you are about 3 billion to one. So when the police say "This is your DNA," what they're saying is "We are certain this is you to within a probability of 3 billion to one." This corresponds to approximately 22 bits of information.

When the Web site says that your browser is leaking 20 bits of information, that's a way of saying that it can positively identify you to a probability of 1,048,575 to one; the odds that someone else's browser signature is exactly the same as yours, just by pure random chance, is 1,048,575 to 1 against.
Posted By: artie505 Re: Web privacy: cookies vs browser specs - 10/04/11 10:18 PM
Thanks for the explanation of what it all means.

As I said earlier... I'd like to know what can be done to make oneself as close to anonymous as one can get? (I believe you've posted that info already... sometime...somewhere...probably in whichever [MFIF?] thread, Jon and I recollect.)
Posted By: alternaut Re: Web privacy: cookies vs browser specs - 10/05/11 05:49 PM
More Panopticlick discussion here.
Posted By: Bensheim Re: Web privacy: cookies vs browser specs - 10/05/11 07:42 PM
Originally Posted By: artie505
Originally Posted By: Bensheim
I got.....

Within our dataset of several million visitors, only one in 894,868 browsers have the same fingerprint as yours.

Currently, we estimate that your browser has a fingerprint that conveys 19.77 bits of identifying information.
----------------------------------------------
Don't believe a word of it.

Interesting that your results are worded differently than the others; I wonder if you got into a different database...maybe because you're there and we're here?

Edit: xx.77 is a popular number, no?


Nah, I just turned off cookies before I did it.

Then I had to turn them back on again to log in here and post that. (log in doesn't work with cookies turned off.)
Posted By: Bensheim Re: Web privacy: cookies vs browser specs - 10/05/11 07:46 PM
I love reading your posts, Tacit. They're so learned and erudite. Sometimes I read your posts on things of no concern to me, just to read your awesome knowledge on, well, just about everything in the world of Macs and computers.

That is an honest post. I don't do sarcasm.

BTW, have been to Portland Oregon and still remember one of the finest hotel breakfasts I've ever had in my life anywhere on earth in any hotel, and I'm very fussy about hotels and their breakfasts. In Portland, Oregon, on the way from San Francisco to Seattle, many years ago, and I still remember it.

Posted By: tacit Re: Web privacy: cookies vs browser specs - 10/06/11 04:14 AM
Now you have me curious...what hotel was it?
© FineTunedMac