Tuesday, April 16, 2002

The Bloggle Problem Google should recognize weblogs as a document and site structure, and link to the archive url, not the main page of the weblog... Thus it would be better to point to the archive. This would require a convention similar to the robots.txt convention. Google is perfectly placed to initiate the convention.  [john robert boynton: Usability/Usability humor] I see this problem on my site as well. One entirely client side possibity is a javascript which checks the referrer for keywords, does a find on the current page of text, and pops up a local site search for the terms in the google search. John's suggestion that the blogging community would rapidly adapt a convention for pointing to persistent locales is an important observation of the community's growing empowerment, and google's role in that.  See also my rant on Hypertext in the Blogdom 7:59:12 PM    comment

Expanding Targets One of the more exciting papers headed to CHI 2002 this year is work which demonstrates the "user efficiency" boost of one aspect of the "OS X" doc.  Problems with the dock aside, the magnification component has been shown to produce significant, up to 25%, benefit compared to a no-magnification scenario.  It's worth noting that by setting 2 constants based upon the user population & controls, Fitt's law can account for the lion's share of variance in a experimental mouse movement test with different distances and target sizes. Theories of motor control, attempting to go beyond Fitt's law have posited a variety of underlying mechanisms by which the size of and distance from a target affects movement to the target.  The possibility of a target changing in size is not covered by the Fitt's law formula.  The results of the CHI study suggest that increasing a target is effective, even after 75% of the distance is covered. This suggests a real opportunity in common day UIs. The enlargement is particularly effective in difficult situations, like cross screen targeting of a menubar or an iTV interfaces with lower precision pointing device. This result has implications for theorizing about the underlying mechanism governing target acquisition.  Intuitively, it makes sense that a larger target would assist in error conditions, where the user has in most cases made significant strides closer to the goal.  The results suggest that the benefit can be found even in non-errorful conditions, suggesting an extensive feedback mechanism between the visual system and motor control. If this is indeed the explanation, further experimental work should assess whether an situation with more distractors (eg. memory load related to the task in progress) would still provide this benefit.  The paper in question is:Acquisition of Expanding Targets. Michael McGuffin, Ravin Balakrishnan (University of Toronto). CHI 2002, Minneapolis. The paper is available on CHIplace.org until the conclusion of the conference. Access requires registration but the nifty topic map does not.  A future Uzilla test could compare a toolbar in the Mozilla browser with expansion characteristics. 7:40:44 PM    comment