On 20 July 2010 03:14, Moritz Bartl
<tor@xxxxxxxxxxxxxx> wrote:
Speaking on behalf of a good, blind friend: This is not true. Unless you consider him "not normal".
I don't want to get into the intricacies of interface design, and ableism, but some points of note:
-blind people are not normal: they suffer from a disability that differentiates them from others in terms of what they can do;
-the blind are hard to cater for with WIMP-type computer interfaces; adapting interfaces for the blind is often a subsidiary - and difficult - task to basic interface design;
-most good CAPTCHAs like (reCAPTCHA) already incorporate accommodations for blind users;
I agree that it's important to design computer systems and interfaces such that they're accessible to those with disabilities. However, this should not be at the expense of the system's core functionality: we should not allow the great to be the enemy of the good. In this specific case, the point is probably moot.
Spontaneuous idea: I think it might be interesting to use a fingerprint similar to the one caculated by Panopticlick to limit/influence the selection of bridge addresses.
Panopticlick uses a fingerprinting system that's quite effective against individual web users, because of the way that we set our browsers up for functionality. However, a malicious automated crawler can say whatever it wants: whatever resolution, _javascript_, cookies, and flash settings it wants. Using that type of fingerprinting would have little effect on a malicious crawler, but would be extremely inconvenient for normal users.