Live data from Hacker News

I too know the websites you visited

oxplot.github.com

81–90 of 98 posts

Re: I too know the websites you visited

#81

I think it got 100% for me, on Safari/Mac. Note that it doesn't need to be 100% accurate to be effective. If it guesses better than 50% (i.e. coin flip), then it could be used to give guesses with at least some confidence. No different than analyzing any other noisy dataset. Because this all works client-side, it can also be done quite invisibly.

Right, but those applications will be for things like online advertising, which is already tracking your visits and/or just assuming you use the popular sites anyway. Can this be used to violate privacy in a meaningful manner if it misses 50% or more of the time? You'd have complete plausible deniability if you were pinned to have accessed some site you don't want people to know you accessed. It can't be used as evid…

I oppose user tracking not necessarily because there's something to hide, but because a sense privacy is a fundamental component of a sense of independence, and for me, independence is one of the critical components of happiness.

Re: I too know the websites you visited

#83
post #39

Not even close, while all the sites it said I visited, I had, it missed tons of other sites I'd visited.

Same here. There were 3 correct positives listed, and I had about 7 or 8 false negatives.

Same here, although I visited multiple sites listed in the very same session on the browser I have open, the tool only reported one as visited, and Google as a Whoops. It should have reported at least 6 more!

Re: I too know the websites you visited

#84
I just tried it twice, once on a public wi-fi network. And then again when I got home. It worked very well on the public wifi, and had many false positives at home.

It seems to work better on slower internet connections. The script returns calls a site "visited" if the response time of the potentially cached image is less than 1/20 the time of the certainly uncached image.

On slow connections the cache is much faster than the uncached. On fast connections it's only slightly faster. However, the known uncached images sometimes have "10x increase in latency" so it seems that based on my (and other's experience) that this is a major problem.

One could attempt to normalize this for the sites where appending random query string causes higher latency. Simply precalculate the added latency from images with the random query string on a per site basis. Then subtract it from "uncachedTime."

Re: I too know the websites you visited

#85
post #47

It said "not visited" for EVERYTHING, except google which says "whoops". I have visited nearly all of them in the last 2 months. But don't despair, I have one of the most hostile browser settings. I have RequestPolicy, NoScript,and Flashblock.

Safari with adblock and ghostery here, got 8 correct and 18 false positives.

Re: I too know the websites you visited

#86

Throwaway account. My company created an analytics product around the ability to track which sites your visitors have visited. It used a different and (at the time) more reliable technique. About a year after the product launch we were contacted by a powerful washington based lobby group and they wanted to chat. They felt it violated a site visitor's "reasonable expectation of privacy". I agreed. So we pulled the fea…

Related, an advertising company who was accused of tracking history: http://cyberlaw.stanford.edu/node/6695

Re: I too know the websites you visited

#88
When the script is run the second time, it will show that every site was visited. After the first visit, guess its cached and cannot figure out if its a hit or a fail :)

Running in Chrome's incognitive mode is a bit different though. only 7 show up cached the first time its run.

Re: I too know the websites you visited

#89
post #29

Apparently the key to people not knowing where you visited is to use IE. It missed sites like Twitter and Facebook that are open for me all the time. It did get one site correct, HN ;<).

It guessed HN correctly for me the first time, but on the second run it said I hadn't visited any of the sites, including HN. Perhaps a bug..

Re: I too know the websites you visited

#90

Extremely wrong on Chrome. The first time I ran it, it said I had not visited any of them. Google and HN were definitely browsed today. Ran it again, ALL of them appeared visited. Even sites like abebooks, which I have not visited at all.

That's because the second time you ran it, all the images were in your cache from the first time you ran it. That's the expected result.

First of all, I was explaining my experience. Second,I know it is a proof of concept. Still, I don't think it really does prove any concept, since the results are not reliable at all. A proof of concept, requires to give you a reliable results, hence a proof.
Post reply on HN