Any web page can detect whether you have Google Desktop Search running by loading an image (or perhaps any URL) from http://127.0.0.1:4664/.

?

Cheers.

Another free product that takes longer to display search results than the google product is Copernic. Their web site is at http://www.copernic.com

I’ll take a shot at this. It’s easy enough to load an image from a certain location without really letting the viewer know it:

<img src=”some-normal-location” width=”1″ height=”1″ id=”some-id”>

From here, Javascript can switch the normal image with a faulty one. It can even determine whether an image loaded or not:

function noLoad() { did_not_load = true; }

var badImage = new Image();
badImage.src = “http://127.0.0.1:4664/“;
var myImg = document.getElementById(‘some-id’);
myImg.onerror = noLoad;
myImg.src = badImage.src;

if (did_not_load)
document.write(“Download Google Desktop!”);
else
document.write(“Welcome, Google Desktop user!”);

This may not seem particularly bad, and in fact it may not be (I’m not the security expert here), but it is revealing more information to the website than it should have.

Note, however, that this doesn’t really apply only to Google Desktop users — it applies to the users of any programs that run a mini-web server on port 4664. (This isn’t to say that Google Desktop may not be the only one, however — I don’t know if any other such programs exist.)