<html><head><style type="text/css">body, p, li { font-family: NokiaPureTextLight; }</style></head><body><div id="fenix-user-content">
<p style=" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px;">I now use duckduck go as my default search engine. Its not as detailed as google, but generally I find what I'm after. It always offers to pass on to google/bing etc, so you can single click to there if you're not satisfied.</p>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px;"><br></p>
<p style=" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px;">Euan</p></div><br><div id="fenix-reply-header"><p>On 22/10/12 6:39 Brad Campbell wrote:<br></p></div><div id="fenix-quoted-body"><div>
On 21/10/12 23:23, Nick Bannon wrote:<br>
<br>
</div>
<div>
> Simple Google search results used to use a tracking URL _occasionally_,<br>
> random sampling to get some idea of when otherwise anonymous users chose<br>
> a particular outgoing search result. At some recent stage they turned<br>
> it on pervasively.<br>
><br>
> Meaning that one can't simply copy and paste a working URL out of<br>
> a search result anymore. It adds extra lag to every click on such a<br>
> result, perhaps over a slow UMTS link. I can't believe it improves the<br>
> results they got from occasional sampling that much. It's not even a<br>
> HTTP redirect, it replaces a simple static link with a JavaScript HTML<br>
> file that does the redirect. (!)<br>
><br>
> <script>window.googleJavaScriptRedirect=1</script><script>var<br>
> f={};f.navigateTo=function(b,a,g){if(b!=a&&b.google){if(b.google.r){b.google.r=0;b.location.href=g;a.location.replace("about:blank");}}else{a.location.replace(g);}};f.navigateTo(window.parent,window,"<a href="http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/">http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/</a>");<br>
> </script><noscript><META http-equiv="refresh"<br>
> content="0;URL='<a href="http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/">http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/</a>'"></noscript><br>
><br>
> Madness.<br>
><br>
> Nick.<br>
><br>
<br>
</div>
<div>
Interestingly enough when you mouse over the results, you get the <br>
correct url in the status bar. Its the trickery they play when you hit a <br>
mouse button that is causing the issue. I pulled apart the firefox <br>
plugins xpi that I'm using and it simply removes the 'onmousedown' <br>
handler from each link. Nice and very simple. Works a treat.<br>
<br>
</div>
<div>
Now I must need to figure out how to do that in squid before the html <br>
hits the browser.<br>
<br>
</div>
<div>
On the same note, I've been looking at alternate search engines that <br>
promote no tracking and privacy. duckduckgo and ixquick. Anyone else <br>
using alternate search engines?<br>
</div>
<div>
_______________________________________________<br>
</div>
<div>
PLUG discussion list: <a href="mailto:plug@plug.org.au">plug@plug.org.au</a><br>
</div>
<div>
<a href="http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/">http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/</a><br>
</div>
<div>
Committee e-mail: <a href="mailto:plug@plug.org.au">plug@plug.org.au</a><br>
PLUG Membership: <a href="http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/">http://razzed.com/2009/02/12/analysis-of-google-outbound-link-tracking/</a><br>
<br>
</div>
</div><br></body></html>