(pathetic) CAPTCHA implementations...
I've been seeing quite a lot of terrible CAPTCHA implementations out there lately, so here's a thread where everyone can post their findings and maybe make suggestions on how to solve the security issues we see...
Anyway, here's my contribution (found this a few days ago, and saved it because it was so funny):
In a login page's source, the (quite lazy) web developer included:
<script language="JavaScript">
<!--
if (top.frames.length!=0)
top.location=self.document.location;
// Disable browser's Back button on another page
// being able to go back to this page and seeing the same turing image.
window.setTimeout('history.forward()', 1500);
// -->
function initpage() {
document.turingimage.src = '/cgi-bin/sblogin/turingimage.cgi?' + Math.random();
prefillForm();
}
// -->
</script>
...and later on...
<!-- <body> -->
<!--
Use the this 1X1 pixel version of the Turing image if you remove
the visible one from the page. Calling the Turing image is necesary
even if it's not visible.
<img src="/cgi-bin/sblogin/turingimage.cgi" height="1" width="1" name="turingimage" id="turingimage"><br>
-->
So the page basically gives you instructions on how to bypass the CAPTCHA, LOL!
Gotta love when developers believe they're the only ones seeing the source :D