> If you hand an N70 and 8008 to complete  beginners, 
        > train them in how to set every feature  of the camera, 
        > and then test them on setting different configurations
        > --as in, "Now set each camera to shutter-priority at 
        > 1/250th, spot metering, rear curtain synch and self-
        > timer mode" --most of the new shooters will set the 
        > N70 correctly more often than the 8008.  

        It depends on the training.  If you just taught
        them how to set the features -- "push this button,
        turn that knob, until this icon appears" -- instead
        of teaching them what was actually going on, then 
        yeah, the group with the N70 would do better.  It 
        would be a pointless experiment though.  They'd 
        still be clueless about what they just did.

        If you taught them what the settings do, and what
        they're for, (so they had a logical framework to
        work from) you wouldn't see much difference between
        the N70 users and 8008 users.  The difference between
        the groups would be less that the differences among
        users in each group.

        If you brought the group back after a month, and 
        none of them had handled a camera since training, 
        the N70 users would probably be able to remember 
        what to do faster than the 8008 users.  That's the 
        big advantage with the menu-based stuff.  You don't 
        need as much long-term memory.

        If you brought them back after a month, and they'd
        all shot 2 rolls of film a day for that month, the
        8008 users would kick the N70 users' butts in a speed
        drill.  The patterns would be ingrained.  The limiting
        factor would be the interface, and the N70's takes 
        more steps.

        Human brains don't need intuition FOR TASKS THAT 
        ARE REPEATED OFTEN.  After the first dozen or so 
        repetitions, the steps become automatic, no matter 
        how un-intuitive or even counter-intuitive they are.  

        When the operator reaches that level, all that matters 
        to him is how powerful a set of operations are available,
        and how much flexibility does he have in combining 
        them to do what he wants.  There may be a steep learning 
        curve to get to that point -- it's nice that there are 
        alternatives for people who don't want to climb that 
        curve, or who don't use the system enough to retain 
        proficiency -- but people who use any system all day, day 
        after day will tax the limits of ANY interface eventually,
        and a tree structured interface will get in their way 
        much sooner.


        -Don

Reply via email to