I thought it would continue to be hacked.
Could someone leverage Diego's video manipulation classes & Morphic's ability to convert text to gif to hack together an "enter the text you see in the graphic" security feature? The hardest part would be selecting a range of words.
Any human could edit a page at any time. Automated hacking would be nearly impossible.
All Swikis & SmallWikis would benefit.
Cheers, Darius
We could have 10 or 20 alternating images for the human author to recognized and include as part of the verification process.
Unfortunatly, the bot programmer could easly have their bot try 10 - 20 words on every page. The human hacker could discover all the words after a few minutes at the site.
You'd need a couple thousand words at least, randomly selected.
Cheers, Darius
Except for (at the least):
- people who can't see (or see well);
- people who want (or have to) use text-based browsers.
In principle, yes. In practice, no.
- People who can't see are more likely to be readers than submitters and are unlikely to be Squeak users due to its graphic/mouse dependant nature. Alternatively, they can e-mail a sighted person to post for them.
- People who can't see well could be provided a link to a much larger version of the graphic.
- People who want (or have to) use text-based browsers can download the graphic and view it on their computer. Alternatively, they can e-mail a graphic enabled person to post for them. Also, if they can open Squeak and open Scamper, they can post.
Bandwidth isn't an issue either.
Cheers, Darius
More on the issue:
http://weblog.infoworld.com/udell/2004/12/01.html The CAPTCHA game
"completely automated public Turing test to tell computers and humans apart."
On Wed, 8 Dec 2004 09:11:51 -0800, Darius squeakuser@inglang.com wrote:
Could someone leverage Diego's video manipulation classes & Morphic's ability to convert text to gif to hack together an "enter the text you see in the graphic" security feature? The hardest part would be selecting a range of words.
You wouldn't even need to select a range of words. The system could randomly choose 5 or 6 characters from normal keyboard ASCII-printable text (or even just letters and numbers).
You also wouldn't need any video stuff -- the server would know what the raw text is, and as long as what the person types in matches it, everything would be fine.
Later, Jon
-------------------------------------------------------------- Jon Hylands Jon@huv.com http://www.huv.com/jon
Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com
On Wed, 8 Dec 2004 09:11:51 -0800, Darius squeakuser@inglang.com wrote:
Could someone leverage Diego's video manipulation classes & Morphic's ability to convert text to gif to hack together an "enter the text you see in the graphic" security feature? The hardest part would be selecting a range of words.
I've started working on this -- you can see what I'm doing here:
http://www.huv.com/distort.gif
We can play with the colors, fonts, etc without any problem.
Later, Jon
-------------------------------------------------------------- Jon Hylands Jon@huv.com http://www.huv.com/jon
Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com
squeak-dev@lists.squeakfoundation.org