We all know ‘captchas’ are bad…
We all know they are a necessary evil…
…but do we know what we actually use them for?
The assumption is we are using them to detect humans and allow them - not true, we are actually using them to try and detect bots and block them.
And this is the captcha’s flaw - because humans are flawed, captcha’s very often trip people up.
But there is something we seem to have forgotten - the browser. Bot’s don’t use Firefox, or IE, or Opera etc. They are custom browsing scripts.
…so here’s the good bit.
Rather than trying to detect the human in the system, could we use javascript to detect the browser, hunt through the DOM for some (and it could be variable) browser feature that the Bot’s don’t have? such as the ability to upload an image? or install a plugin?
If javascript is available - yes.
If javascript is not available, then we could still use an image and alt text - the ‘old fashioned’ way, but a real human user should never need to see the captcha, unless they can’t use javascript, and then there should be no need to scramble the image.
users of screenreaders with javascript enabled would be allowed in, users of screenreaders with javascript disabled would read the alt text of the image and be allowed in.
Bot’s would be detected as ‘not browsers’ and be blocked.
…well thats my theory anyway.
Tags:
Geek