[PrivacyBadger] Using the behaviour engine for other purposes

Alexei alexeiatyahoodotcom at gmail.com
Wed Mar 21 09:52:38 PDT 2018


Refactoring the code to make the engine more reusable sounds cool and
should be doable, but I'm not clear on how it would work. You could take a
look at
https://github.com/EFForg/privacybadger/blob/master/doc/DESIGN-AND-ROADMAP.md
(somewhat out of date but still helpful) and our code (such as
hasCookieTracking, how it's defined, where it's used:
https://github.com/EFForg/privacybadger/blob/42eb9cc7b779b91751de1d6d5d57e418c57621ea/src/js/heuristicblocking.js#L467)
and perhaps post any ideas/suggestions on our GitHub.

I know at least one browser (WARP on Android) implemented parts of Privacy
Badger's functionality, but my guess is that they did it by reimplementing
logic, not by reusing code. Practically speaking, it might be easiest to
use one or more ad blocking/privacy lists with your proxy.

We mean to use Privacy Badger with a crawler as part of
https://github.com/EFForg/privacybadger/issues/1019 (lower priority) and
https://github.com/EFForg/privacybadger/issues/1891 (higher priority).

Here is a quick-and-dirty script that drives Privacy Badger in Chrome with
Selenium: https://github.com/EFForg/privacybadger/pull/1582.

On Sat, Mar 10, 2018 at 8:26 AM, Maxim Nazarenko via PrivacyBadger <
privacybadger at lists.eff.org> wrote:

> Hello!
>
> One issue with forward proxy will be TLS: Privacy Badger needs to
> modify the data, and that is basically MITM attack.
>
> Best regards,
> Maxim Nazarenko
>
> On 9 March 2018 at 17:03, Kjetil Kjernsmo via PrivacyBadger
> <privacybadger at lists.eff.org> wrote:
> > Hi all!
> >
> > Many thanks for the Privacy Badger, it alone makes my EFF membership well
> > worth, and I hope to increase my donation in near future. I would like to
> > share an idea:
> >
> > Would it be possible to refactor the code so that the engine checking
> > whether you are tracked can be reused in completely different contexts?
> >
> > I have a couple of use cases for it: I'd like to run it on a forward
> proxy
> > at home. We have simply too many devices in the house, and to install the
> > browser extension on all is too much work. And who knows what other apps
> > than the browsers are doing...? So, I figured, I'd rather set it on a
> Squid
> > proxy and let that block the trackers...
> >
> > My second use case is to use it an a Web crawler. I'm in Norway, and the
> > public sector is being automated and streamlined. I strongly support this
> > effort, but sometimes they mess things up. For example, they make heavy
> use
> > of Google Analytics, which I think is a bad idea in itself, but it gets
> > worse when they also mess up the settings and enable the DoubleClick
> > cookies, which is a blatant violation of the law. I have used Privacy
> > Badger to alert me to this fact and send complaints, and the response has
> > been OK: "Thank you for the notification, we are truly sorry, and have
> > disabled the cookie promptly", but this is apparently *really* hard,
> > because every agency seems to get it wrong every now and then... :-/
> > So, if citizen vigilance is required, I should at least automate it. It
> is
> > straightforward to hack a crawler that monitors these sites, but it would
> > be nice to re-use the code that Privacy Badger has to determine the
> > likelyhood of trackers.
> >
> > Cheers,
> >
> > Kjetil
> > _______________________________________________
> > PrivacyBadger mailing list
> > PrivacyBadger at lists.eff.org
> > https://lists.eff.org/mailman/listinfo/privacybadger
> _______________________________________________
> PrivacyBadger mailing list
> PrivacyBadger at lists.eff.org
> https://lists.eff.org/mailman/listinfo/privacybadger
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.eff.org/pipermail/privacybadger/attachments/20180321/a3697969/attachment.html>


More information about the PrivacyBadger mailing list