Zap Annoyances on the Web, Part I

Surfing the Web is something most of us do daily.
At home I am on a dial-up connection, and I'm
constantly looking for ways to make my browsing
experience more efficient and to protect my privacy
while on-line. The approach I describe here
allows you to stop many ads, images, Web bugs and
pop-up windows. In case you are not familiar with
Web bugs, they are small graphics included on a Web
site or in an e-mail that are used to track who is
reading. Whether you are on
a dial-up connection or a high-speed connection,
the tools presented in this article are guaranteed
to make your on-line browsing experience more
enjoyable and to protect your personal habits from nosy
marketers.

My environment consists of an IBM A31 ThinkPad with
a 1.9GHz Pentium 4 Mobile CPU and 1GB of RAM
running Fedora Core 1 (fedora.redhat.com).
I am using Mozilla 1.5 (www.mozilla.org) as
my Web browser.

The Approach
The approach we take here consists of the
following three basic parts:
Mozilla's image, JavaScript and pop-up window management capabilities;
a proxy server to cache Web content; and
a program to intercept requests for undesired
content.

If you are not familiar with proxy servers,
here's a quick overview. For the purpose of this
article, we define a proxy server as an
intermediate server used to intercept requests for
Web pages and associated files. The proxy server
either handles the request itself or passes the
request on to another proxy server or to the
original target. A proxy server can be transparent,
intercepting all requests automatically. In our
example we manually configure the browser
to go through our proxy server. Proxy servers
also can be used to proxy other types of requests, but
that falls outside the scope of this article.

We are using Squid (www.squid-cache.org)
as our proxy server. Squid is an extremely popular
proxy server that's in use all over the world.
Squid allows you to cache Web content locally, so
instead of going out to the Internet for content
you have previously downloaded, you can
retrieve it from Squid's local cache. Squid is
smart enough to know if content can be read from
cache or if it must be updated from the originating
Web site. If you are looking at a news
site, for example, Squid knows that it needs to refresh the
text on the site, but it usually loads the static
images on the site from cache. Squid also gives us
the ability to add in something called
redirects, which intercept
requests and possibly modify them in some way. For example,
if you request an image that is known to be an
advertisement, the intercept may return a
placeholder image instead of the image requested,
completely bypassing the request for the original
image.

Implementing Our Solution
The first step is to set up
Mozilla to block pop-up windows. The option to
enable this feature can be found under Edit→Preferences→Privacy
& Security→Popup Windows. You are greeted by the screen shown in
Figure 1. Choosing the option to Block
unrequested pop-up windows does exactly that.
You can play a sound or display an icon when
pop-ups are blocked if you so desire. There also is an
option to allow pop-ups explicitly from certain
sites. If you click on the button labeled Allowed
Sites..., you are shown the screen in
Figure 2. This feature is handy for sites
that pop up external windows for things like
showing progress for a long-running task. You can
leave this list blank to start and add sites
as you go.

Figure 1. Block unrequested pop-up windows from Mozilla.

Figure 2. You can permit pop-ups from sites that use
them for legitimate purposes.

The second step is to configure Mozilla to manage
JavaScript events. These settings can be found
under Edit→Preferences→Advanced→Scripts & Plugins. Check boxes
on this screen allow you to turn off JavaScript completely
for your browser, mail and news readers. You
also can keep JavaScript enabled and configure Mozilla
to stop scripts from moving or resizing windows,
raising or lowering windows, hiding or changing the statusbar, changing images, creating or changing cookies and
reading cookies. The settings I use can be seen in
Figure 3 and should work well for most
situations.

Figure 3. Disable the JavaScript features that Webmasters
use to create annoyances.

The next step is to tackle setting up
our Squid proxy server. If you have never set up a
proxy server before, have no fear; setting up Squid
is much easier than you think.
First, check to see if you have Squid already installed on
your system. This can be done by typing
which squid. If you get
back a path to Squid you're all set. If not, you can
download prebuilt RPMs from
rpmfind.net/linux/rpm2html/search.php?query=squid&submit=Search+...,
or download the source from
www.squid-cache.org. You also need Perl
installed, which is pretty much guaranteed to
be on your system already. Installing from source
is easy; follow the three-step install
instructions in the INSTALL file included with the
source distribution:

% ./configure --prefix=/usr/local/squid
% make all
% make install

The latest version of Squid, at the time of this writing, is
2.5.STABLE4. I am using version 2.5.STABLE3, which
shipped with Fedora Core 1.

Next, we configure Squid. Before updating your
Squid configuration, we issue the stop
command, /etc/init.d/squid stop, to make sure Squid is not running.

Now, edit the /etc/squid/squid.conf
file as root. We are going to
change only one setting in the default squid.conf
file. Locate the line that reads:

# http_port 3128

and change it to:

http_port 8080 

This line changes the port on which Squid is running
from the default port of 3128 to the easy-to-remember port of 8080.
Although this step is not
necessary, you probably want to do it so
you can remember on which port Squid is
running without having to look at the squid.conf
file in the future. Keep in mind, though, that port 8080 may
conflict with some applications, such as Tomcat or
JBoss. If there is a port conflict, you can
choose a different port for Squid. Now it's time to
test our Squid configuration. Type the following
command as root to start up Squid: /etc/init.d/squid
start.
You should see an OK after Squid starts. If you
get an error, go back and check
your configuration settings to identify the
problem.

Now we're going to set up Mozilla to use
Squid as a proxy server. To do this, go to
Edit→Preferences→Advanced→Proxies.
Choose the Manual proxy
configuration and put in 127.0.0.1 for
your HTTP Proxy and 8080 as the port. If you are
running Squid on a port other than 8080, enter that
value here. Put in the same values for your SSL
Proxy. You should end up with something similar to what is shown in
Figure 4.

Click OK and try to browse any
Web site to make sure that Squid is functioning
properly. A good test for this is to go to
images.google.com and search on the term
penguin. After the page loads up, click refresh in
your browser and notice how much more quickly the images
load the second time. The increased speed is
from a combination of Mozilla's own built-in
caching capabilities and Squid's caching of the
images for you. As a test, I turned off Mozilla's
built-in caching functionality and loaded up the
penguin images page. The initial loading of the
page took 27 seconds on my dial-up connection, while the
second viewing took only a little over two seconds.

Figure 4. Configure your browser to use the proxy server.

Part II of this articles explains how to use AdZapper to block many ads
and Web bugs.

Frank Merenda is a systems architect for John Deere,
where he writes J2EE applications on Linux, as well
as a coauthor of the Linux site, uptime
(uptime.steidler.net). He and his wife Jacq
live with their seven dogs and five cats in rural Jackson
County, Georgia, where they both are helping the
Humane Society of Jackson County
(www.hsjc.com) build an animal shelter. He
can be reached at linux@frankandjacq.com.