Re: Geolocation spoofing and other UI woes

[Pavel Machek <pavel () ucw cz>]

Hi!

> This may be of some interest to people on the list:
>
> http://lcamtuf.blogspot.com/2010/08/on-designing-uis-for-non-robots.html
>
> In general, there is a class of UI design problems that trace back to
> the failure to account for the inherent limitations of human
> cognition; the specific example exploited by this PoC is the HTML5
> geolocation API (supported by most browsers except for Internet
> Explorer); calls to navigator.geolocation.getCurrentPosition() result
> in prompts that can be clicked in a timeframe shorter than the minimal
> latency required to respond to visual stimuli. A whimsical
> Firefox-specific PoC is:
>
> http://lcamtuf.coredump.cx/ffgeo2/
>
> I reported a number of flaws similar to this one between 2005 and 2007
> (e.g., https://bugzilla.mozilla.org/show_bug.cgi?id=376473); but it
> seems that these problems remain endemic to security UIs across all
> browsers; and the previously implemented workarounds are sometimes
> removed as a perceived usability roadblock. We probably need to do
> better, especially as computers get faster; this would not have been
> nearly as much of an issue ten years ago.

Well, Android actually does something to address this for system
prompts: dialog box pops up with buttons disabled, and they are
re-enabled after second or so.

This should be certainly done for security prompts, and probably for
normal dialog boxes, too.
                                                                Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/