kirby urner wrote: > Computer games will often emulate both aesthetics, as does ordinary > fiction. We're accustomed to a mix of first and third person.
You put it perfectly. And I think that is what bothers me about Croquet in a sense as a proposal for "the" interface (especially for kids). The Croquet approach chooses to focus all interaction with the system through a first person 3D world on the computer display, with third person applications perhaps within that (e.g. a simulated sheet of 2D graph paper in that 3D world). I am more accustomed to think of the computer as displaying "third person" information, and then sometimes I'm willing to suspend disbelief for a first person interface within a window on that third person desktop. Even if the window takes up the entire screen (or all screens :-), I know the third person perspective lies underneath it if I stop that application. Ignoring the R&D aspects of Croquet (it's a good effort, breaks new ground, etc.), right now, I don't think computers (and interface technology) are not going to be close enough to 100% reality to make it worthwhile to channel all our interactions with them *first* through a first person 3D view. Certainly, if we get the Holodeck (with smell and force feedback) I'll revise that opinion. :-) And so also, from a perspective of priorities of making Squeak-like or Self-like things happen using Python in a learning (or developer) context, I think the Croquet aspects are less important than the interactive 2D tools aspects. By the way, the seventeenth proof of concept of Prototypes in Python with the rudiments of a tree browser for 2D Morphs (though the browser is in wx): http://svn.sourceforge.net/viewcvs.cgi/patapata/PataPata/proof_of_concept_017.py?view=markup > Anyway, I think the degree of realism needed, and in what dimensions, > is very application-specific. That's why it's dangerous for the OS to > weigh in too heavily with a graphical view. > [snip] > I don't imagine oil executives want cartoony avatars on screen, but > having all present gaze upon a shared view of some new drilling area, > complete with high def representations of rigs, pipelines (current and > projected), equipment and crews, makes the meeting more effective and > productive. You're absolutely right here. I think 3D is a great thing for visualizing data to understand issues and help solve problems, and a great way to communicate whatever the 3D representation chosen (realistic to whatever degree or even totally schematic or even mathematically abstract). And I can even see the value of ActiveWorlds type 3D with avatars for occasional meeting or even specific projects. To be clear, and to agree here, I have nothing against 3D. I was playing with VTK and the Python bindings for it the other day as on possibility for an underlying toolkit for a PySqueak, and can see the real value of it for looking at scientific data in general using Python. And I can see further how working in 3D could be very motivating to learn programming (if the package makes that easy to do, such as VPython or Alice attempts). And I think Croquet is a wonderful attempt to make it easy to do 3D collaborative applications. But, in considering Croquet as the evolution of Squeak as a kids programming environment I see an implicit intent based on my interpretation of Alan Kay's intent (perhaps in error?) that Croquet is what we should all use all the time when we are at the computer. I think that implicit intent is what bothers me, because, perhaps conditioned by experience (?), I think approaching the computer (as it exist now and for the next ten to twenty years) expecting a third person experience is better (i.e. more flexible, less disappointing, more effective for most tasks) than approaching it expecting a first person experience (where that first person experience will be missing things for the foreseeable future). Anyway, a philosophical point of interface design for me to muse over some more. --Paul Fernhout _______________________________________________ Edu-sig mailing list [email protected] http://mail.python.org/mailman/listinfo/edu-sig
