loki@relay.EU.net (Never Kid A Kidder) (05/31/90)
Although `jacking in' may be a long way off, the principal of cyberspace is the representation and manipulation of data in an easy-to-handle form. It started with nasty things like punch-cards, and so far has developed into the DeskTop. We've already seen talk of developing the X environment to improve on this, but we're still only at the conceptualisation phase. The bottom line is this: how *are* we going to represent/manipulate data in "DeskTop II - the sequel"? Gibson certainly got the ball rolling (did anyone preceed him, or was he Numero Uno?), but I always found his descriptions rather open to interpretation. One of the intermediary phases is well represented by Mac's DeskTop V7, where there has been a standardisation of representation for all that system and fonts stuff. I suppose things like HyperCard are also good contenders too. It would seem to me that the tree is a good place to start, as it's pretty well universal now. But are there any better ways to represent it than heirarchical folders/cards/listings? I have thought about representing the tree as a three dimensional thing, and moving around it with some kind of control device, like a joystick, with added controls to search for named objects and so on. For the basic user, of course, there's not always a need for this, unless one decides on a universal format, and then one can maybe break databases into this structure (like the new DeskTop does). Or are there better ways to view a database? (I'm an ignoramus when it comes to this - perhaps I should ask some users!!!) We may want more tools on hand, like a toolbox, which could be manipulated by the DataGlove. Different applications would have different toolboxes, perhaps (I can just see a virtual hardware engineer with a box of `filing system' spanners for tuning the filing system!!!). Would we actually have a virtual joystick, or something, rather than some physical device? It seems to me that with a pair of DataGloves, and an AudioVisual Headset, we could achieve rather a lot of things that would otherwise require specific hardware interfaces. Will this be faster, or will it simply be easier to learn? For example, if you can type fast, you can often type faster than you can write, and switching from mouse to keyboard can slow things down. But then the mouse is only a transitional phase. We might speak the commands at first, and have them confirmed on a display. We might also create command lines using our toolbox (I believe such a thing can be done on Macs already). The data stream is a tried and trusted approach to data manipulation, and perhaps that can be set up in some visual way, and no doubt the stream can be branched in some way using multiple I/O descriptors. The representation of running processes needs to be addressed, and how their connections are shown; indeed we might want to attach to resources for sending/receiving data, and this could be handled in our virtual world by calling up the process and attaching a pipe or something. We'd probably have to go ahead and try some things out before we knew whether it was too slow to work this way. One final thing. I normally visualise the concepts of what I am doing when I type things in, and I presume everyone else does the same, so in a sense, it's already there; if we all had good imaginations, we'd find it even easier. I often wonder whether it might not be better in the long run to somehow improve visualisation techniques and do away with all this virtual world stuff - maybe some kind of acid ... then we'd all have more fun in our leisure periods too ...