Home > data science, open source tools > Do not track vs. don’t track

Do not track vs. don’t track

March 4, 2012

There’s been some buzz about a new “do not track” button that will be installed in coming versions of browsers like google chrome. The idea is to allow people their privacy online, if they want it.

The only problem is, it doesn’t give people privacy. It only blocks some cookies (called third-party cookies) but allows others to stick.

Don’t get me wrong- without third-party cookies, the job I do and every other data scientist working in the internet space will get harder. But please don’t think you’re not being tracked simply by clicking on that.

And as I understand it, it isn’t even clear that third-party cookies won’t be added: I think it’s just an honor system thing, so third-party cookie pasters will be politely asked not to add their cookies.

But don’t believe me, visualize your own cookies as you travel the web. The guy (Atul Varma) who wrote this also open-sourced the code, which is cool. See also the interesting conversation in comments on his blog Toolness.

Let me suggest another option, which we can call “don’t track”. It’s when nothing about what you do is saved. There’s a good explanation of it here, and I suggest you take a look if you aren’t an expert on tracking. They make a great argument for this: if you’re googling “Hepatitis C treatments” you probably don’t want that information saved, packaged, and sold to all of your future employers.

They also have a search engine called “DuckDuckGo” which seems to work well and doesn’t track at all, doesn’t send info to other people, and doesn’t save searches.

I’m glad to see pushback on these privacy issues. As of now we have countless data science teams working feverishly in small companies to act as predators against consumers, profiling them, forecasting them, and manipulating their behavior. I’m composing a post about what a data science team working for consumers would have on their priority list. Suggestions welcome.

  1. Richard Séguin
    March 4, 2012 at 12:15 pm

    I’ve been trying out DuckDuckGo and Ixquick the past few days. I don’t yet have an opinion of which is most secure, but they both work well, are very user configurable, and in some respects work better than (non-scholar) Google. In the past two years I’ve noticed Google’s usefulness deteriorate markedly, now often having to wade through pages of commercial fluff before finding the meat of whatever it is I’m interested in. These new search engines feel more like the Google of yore.

    If you’re using OS X you might also consider the Omniweb browser. You can set cookie policy globally like most other browsers, including blocking third party cookies and the ability to automatically throw out cookies at the end of a session. You can also configure each website individually. Right now I’m using this to reject all cookies except for a few web domains which I have configured individually. Safari is superior to printing though and I sometimes use it rather than Omniweb for that purpose.


  2. Tim Cerino
    March 4, 2012 at 3:07 pm

    Fight fire with noise? What if there was an application that would feed random words and phrases into your browser’s search field basically all day long? Your real searches would be lost among the spray of randomness. Perhaps have a white-list and/or black-list of search terms to avoid randomly searching into unsavory territory.

    If it is impossible to disable tracking, perhaps it can be rendered valueless to the tracker?


  3. March 4, 2012 at 6:28 pm
  4. March 9, 2012 at 3:39 pm

    I’m feeling a little 1984-quality claustrophobia now. @mathbabe: What browser/suite of tools are you using?


  1. March 8, 2012 at 6:59 am
Comments are closed.
%d bloggers like this: