From: Maarten van Dantzich [maartenv@MICROSOFT.com] Sent: Monday, October 04, 1999 5:51 PM To: 'Niklas Elmqvist' Cc: '3d-ui@hitl.washington.edu' Subject: RE: 3D window management > Well, it is true that many of the standard productivity applications > may not be suited for immersive 3D environments. It's not clear to me that 2D apps are necessarily harder to use inside a 3D environment. If it's easy to align your view with the normal of the window so you get the usual canonical view, and if the visual quality isn't degraded too much, there aren't necessarily any big hurdles. That's not to say that 2D apps would be ideal, or that I think we know HOW to bring existing apps into 3D in a fully seamless way, but I don't see any fundemental barriers. Aside from that... Niklas Elmqvist wrote: > Let's face it, before HMDs become commonplace, software companies > like Microsoft will not start to develop 3D-versions of their > office and development packages. There's a chicken & egg problem here: if Windows remains popular and MS doesn't implement any 3D, then HMDs aren't going to become commonplace. Gaming use isn't enough; notice that 3D accelerators are commonplace now and MS doesn't use them in the Windows UI [yet]. Or joysticks. Etc. > Of course, let's not forget that many "normal" 2D programs may be > extended to three dimensions to great effect. Remember fsn, the > three-dimensional file manager which was showed off in Jurassic Park? To WHAT great effect? I've known enough people who owned SGIs and used them full-time, but none of them used fsn for real tasks. It's flashy, but not that useful. I'd point to the Treemap solution out of U/Maryland as a better file system visualization. > How about a 3D-debugger which gives useful information on the spatial > relationships of program components? What spatial relationships of program components? Mine seem to be pretty linear. Do you have some programming lanuague I don't know about? ;-) ;-) Seriously, though--I'd be very interested in seeing code evolution over time. And showing program performance traces over time. Some of this is going on, e.g. Steve Reiss' work at Brown University is interesting. [I don't see any papers on his webpage offhand. I do believe he published something. http://www.cs.brown.edu/~spr ] Something that's deeply worrisome about 3D UI is that even the researchers who are publishing in this area don't seem to use any 3D-UI-based tools on a day-to-day basis for tasks other than the research itself. Do you put on a headmount to read your email? :) The above is meant as a devil's advocate argument to provoke discussion, and admittedly biased against head-mounted 3D. But obviously since I work in this area (3D UI) I do have hope! >> Maarten.