Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!wuarchive!emory!ogicse!milton!cgy@cs.brown.edu From: cgy@cs.brown.edu (Curtis Yarvin) Newsgroups: sci.virtual-worlds Subject: Re: so called cyberspace conferences Message-ID: <12979@milton.u.washington.edu> Date: 13 Dec 90 23:56:12 GMT References: <12657@milton.u.washington.edu> <127 <12868@milton.u.washington.edu> Sender: hlab@milton.u.washington.edu Organization: Brown University Department of Computer Science Lines: 109 Approved: cyberoid@milton.u.washington.edu In article <1990Dec13.093343.8402@cpsc.ucalgary.ca> sharp@cs-sun-fsd.cpsc.ucalga ry.ca (Maurice Sharp) writes: > >In article <12911@milton.u.washington.edu> frerichs@ux1.cso.uiuc.edu (David J F r >erichs) writes: >> >>I must agree that these "cyberspace conferences" are quite useless. They >> >>What would be useful is a conference of researchers and theorists to try >>to create some standards for VR and networked VR (my definition of a >>cyberspace is any multiuser WAN). I think many others agree with me that >>the time for bullshit is over and the time to treat VR and its various >>sub catagories like a true field has come. >>We need standardization and Protocols... not fantasy. > >Hiya, > > I must disagree with your assesment. If you look at the literature >on how scientific fields progress, it is NOT time to define standards. >In fact, the types of meetins we are having now are just the type they >should be. > > According to Gaines (see ref at end), there are 6 stages in the >development of an information technology : > Breakthrough - creative advance > Replication - mimick the breakthrough and gaine experience > Empirical - rules of design based on experience > Theoretical - underyling theories found > Automation - theories predict experience and generate rules > Maturity - routine use of technology > > The average time from Breakthrough to Maturity is 70 years ! >Cyberspace and/or virtual reality is at the BREAKTHROUGH stage. What >is worse, the technology does not exist to implement and test most of >the ideas. It is going to be about 20-30 years before there are heavy >underlying theories. Less (3-10) for some empirical guidelines. We just got an off-scale reading on the good ol' bogometer here. 70 years? "Information technology" itself (unless you define it as double-entry bookkeeping & other heavy-duty paperwork) has been around for less than 50 years. I don't know who this Gaines guy is (some freshwater professor of sociology?), but his "stages" are so vague as to be entirely useless. In fact, I don't even know what he means by "an" information technology. So let's try some guesses: (1) A new methodology of hardware design. This probably comes closest to Gaines's stage system. Let's take RISC as an example of a hardware breakthrough. The first RISC machine was the CDC 6600. This was the result of a creative breakthrough by Seymour Cray, around 1970. It was replicated in the mid-1970s by Cocke and others at IBM, culminating in the IBM 803. The "Empirical" stage was pretty much skipped; Patterson was coming up with some "theoretical" results in the early 80s, and RISC microprocessors were beginning to be fabricated. Right now the technology is definitely at maturity. All processors introduced in the last two years incorporate RISC design concepts; although (like the 486 and 040) they may have complex instruction sets, they are highly pipelined and lack vertical microcode. Total time elapsed? 20 years. (2) A new methodology of software design. Software advances have not followed Gaines's model. Rather, they have moved on a slow but steady curve. One important feature of the curve is occasional penduluming: the gradual realization that compilers are good, followed by the discovery that HLLs can be really slow if they're badly designed (eg COBOL); the gradual realization that structured programming c uts down on bugs, followed by the discovery that it could also be quite anal (eg Pascal); the gradual acceptance of object-oriented models, followed by the gradual ostracism of dynamic binding from those models. (3) A new way of thinking about information processing. These have followed the exact opposite of Gaines's path. Let's take symbolic AI as an example. First a brilliant researcher propounds an elegant theory that purports to explain everything. But extending the theory downward into the cold, smelly mud of reality proves difficult; and it eventually sinks under its own weight. Now where does VR fit into this classification? It consists mostly of the first and the third. There are people trying to build it; and there are people trying to concoct theories about it. The examples presented above, as I'm sure Frerichs is quite aware, indicate that we need a lot more of the former, and a lot fewer of the latter. One Eric Pepke is worth a thousand Timothy Learys. VR is going to happen, whether people theorize about it or not; but we don't want to let the theorists suck grant money away from the genuine researchers. >Take a more realistic view of >science (read Thomas S. Khun [sic], that ought to open your eyes). An excellent book - he takes great care to define his terms, which is the most important thing in the philosophy business. But it doesn't apply. "Knowledge technology" is (gasp, shudder) not a science. It's an art. It is not based on hard data, and theories presented in it are not falsifiable - which means they're meaningless. Read Popper. > >Gaines, B. R. From Information Technology to Knowledge Technology. >Future Computing Systems Vol 2 Num. 4. Oxford University Press. > >Maurice Sharp MSc. Student (403) 220 7690 Curtis "I tried living in the real world Instead of a shell But I was bored before I even began." - The Smiths