Skip to content
Raymond Hill edited this page Aug 17, 2014 · 58 revisions
Intro

If you really care very much about your privacy (not being tracked, data mined, etc.), µBlock is a crutch (not a bad one though), even with EasyPrivacy enabled (this is true for any "ad blocker"). If you want more than a good crutch, HTTP Switchboard is the way to go.

Settings

Unlike HTTP Switchboard, µBlock can't foil cookie headers. I strongly suggest privacy-minded users to...

  • Enable "Block third-party cookies and site data" in "Content settings" / "Cookies".
  • Enable "Click to play" in "Content settings" / "Plug-ins".
Command line switches

I personally use these command line switches (Chromium on Linux):

  • --disable-component-extensions-with-background-pages
    • "Disable default component extensions with background pages" (ref)
    • I believe this prevent Hangout Services to be launched by the browser as a background process. I wasn't too happy to find out there was such a process launched even though I do not use Google's Hangout.
    • With other Chromium-based browsers, maybe more stuff would be disabled, you decide whether this is good or bad.
  • --disable-background-networking
    • "Disable several subsystems which run network requests in the background" (ref)
  • [add more switch of interests whenever new ones are found]
Regarding EasyPrivacy

In case you were not aware, using EasyPrivacy doesn't protect completely against Google Analytics. So if you were using Adblock Plus with EasyPrivacy (as recommended by the EFF), you might have thought you were protected against Google Analytics. This is not necessarily the case.

If you are using µBlock, it protects you more against Google Analytics out of the box -- via "Peter Lowe's Ad server" list. Yet, given that an exception filter may exist somewhere in one of the many lists, I can't guarantee the protection is 100%.

Twitter widget

I don't know why this one is not blocked by Fanboy Annoyance, as the list already blocks many other twitter widget-related stuff. So if you use above list, you may want to add the following to your filters:

||platform.twitter.com/widgets.js$third-party

Gravatar (et al)

Each time you visit a site which pull cute little avatar images aside (typically) a commenter's name, there is a corresponding request to Gravatar's web site, and the HTTP referer header contains the site you are visiting. The tracking potential is too much for me, so I block all these requests:

||gravatar.com^$third-party

I don't know if, and how much this breaks things. But for now I am happy to not have my browsing habits disclosed to gravatar.com. I can live without these cute thumbnails.

But this applies to any domain which is ubiquitous enough, gravatar.com is just one example among so many.

To deal with this easily, I find HTTP Switchboard to be the best tool, as to blacklist a ubiquitous domain with 100% certainty is simply a matter of point and click.

Clone this wiki locally