Date: Wed, 15 May 2002 22:37:01 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: Nils Holland <nils@daemon.tisys.org> Cc: freebsd-chat@freebsd.org Subject: Re: The road ahead? Message-ID: <3CE3457D.3EC8885B@mindspring.com> References: <20020516004909.A9808@daemon.tisys.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Nils Holland wrote: > (DANGER: This is long and has not much to do with FreeBSD (that's why I > posted it to -chat). If you're busy, skip this message now. If not, get > yourself a cup of coffee and read on ;-) [ ... ] The "dot com" bubble was a new industry buble, which floated based on the promise of new revenue models which petered out. I don't think that it's really fair to blame this on infatuation with the Internet, per se; a lot of capital ended up going to a lot of places that petered out. You have to blame a lot of ot on the day traders, and computer driven buying/selling. Looking at themarket today, you see the effects of the day traders getting out for the weekend, and getting back in on Monday. If anything, the IPO valuations were responsible for a lot of the "feeding frenzy"... and this is the fault of stock market players -- ordinary people -- and not really about the companies at all. It started failing in March of 1999, when the Federal Reserve raised the prime interest rate, and money became more expensive; recently, the price of money has been going down, and we are once again seeing investments in new venture creation going up. > So far about what has happened. The question, however, is what we can learn > from it. Basically, I believe that the computer industry is in serious > danger - Moore's Law seems to be self-destructing. What I mean by this? > Well, seriously, if I go to a computer shop these days, then I will find a > whole lot of hyper-fast machines, but for an ordinary user, these probably > wouldn't make much sense. If a 500 Mhz machine sits 90% idle while someone > writes a letter of surfs the web, then why should he upgrade to a 2000 Mhz > one? The answer to this is "task based computing". Kirk McKusick likes to say "the number of MIPS delivered to the keyboard has remained constant since 1978". This is both true and false. One of the most obvious places CPU cycles have been soaked up is into human factors developement, and the most blatant example of that is that it's no longer necessary to explicitly request that a spreadsheet be recalculated, as it was "in the old days", where spreadsheets were relatively much more expensive in terms of time to recalculate. A lot of it has also gone to programmers who, frankly, don't know how to code in a resource constrained environment, and, as a result, code things in such a way as to soak up resources that would otherwise be free for other tasks. Couple this with buying the engineers the newest toys that you expect the target market to have on hand, so they can "take advantage of every ounce", and you have a viscious cycle. [ ... DOS -> Windows -> present day ... ] > And suddenly - BANG! Seems that the computer industry has nothing to offer > beyond that. And that's why it's stuck, having financial problems. Incorrect. The financial problems are larely the result of the U.S. government's fiscal policies. Mostly, the rest of the world has very little to offer, since they do not have the strong intellectual property policies of the U.S. to encourage the idea that one can make a return on investment. The ROI factor means that commercially unexpolitable Open Source Sofware is actually a danger to the financial engiones paying for innovation. A lot of people who got into the technology because that's where they thought the money was are hurting: they aren't as good at it as the people who live and breathe it for love, and so they were the first ones "up against the wall" when the industry hit upon hard times. Hopefully, the current troubles have lasted long enough for them to come to the conclusion that the money is in other fields, so they can go provide incompetent service there, instead. > What seems to be > invented these days seems to be only toys, no more revolutions! Does the > new Fisher Price look in Windows XP make computers easier to use or people > more productive, just like the switch from DOS to graphical Windows did for > ordinary users? I guess not. And then - what else is "new" these days? Some > folks would see the ability to talk to your computer as the next big > revolution (which is partly already possible), but I fear that I have to > say that talking would actually slow folks down, compared to having them > enter commands or use the mouse within a GUI. So, another toy, but nothing > new! You need to read "The Innovator's Dilemma". The point I would have you take away from the book is that, if you are not yourself a visionary, then you will only see the greatness of innovations in hindsight. > While the performance of silicon technology may be increasing by whatever > insane factor every year, I don't see too many occasions where such > performance is actually needed - at least not enough occassions in order to > enable computer companies to earn as much money as they did in the past. Then the problems you are trying to apply the computational power to are the same old problems, and not new problems. The applications you are choosing are apparently not important to the future. BTW, "Moore's law" is about transistor density doubling every 18 months; he actually said "every 12 months" and then later revised it to "every 24 months", and we tend to give him the benefit of the doubt, now that we have perfect hindsight, and attribute the average to him. > To come back to the point of most new inventations being "toys", let me > give you another example: There are currently (at least) two companies out > there working on "electronic paper". One of them is E-Ink, and the other > one is something that has recently been spun off from XEROX (I don't > remember the company's new name, so I'll just call it XEROX). A few years > ago, these companies wanted to make us believe that in the future all > newspapers, magazines and books would use their electronic paper - which > must be imagines in a way that you probably have one single "sheet" of that > stuff, and you don't turn the page, but hit some button or so, and the next > page will then be "painted" onto this electronic stuff automatically (or > something like that). Interestingly, it seems to me that E-Ink has now > stopped working on this stuff, using their discoveries more or less for > normal, especially flat and energy-saving displays on mobile devices. > XEROX, instead of revolutionizing our newspapers, as they initially > claimed, seem to sell "signs" to retailers to put in shops (or wherever). > These sings then "update themselves automatically" and always display the > latest information / prices / etc. All significant innovations face an adoption curve. They also are not necessarily able to capture the market they sought to capture, and end up being used for things their inventors did not strictly forsee. In fact, every single significant advance in disk storage technology started out without a market, and ended up finding its niche elsewhere. You should also read Geoffrey Moore's "Crossing the Chasm". > Personally, when I first heard about the stuff E-Ink and XEROX were working > on, I already predicted that they would have no success - at least not as > far-reaching as their marketting insanity said. The reason for this is > simple: I don't see a need for electronic, re-writable paper. After all, we > have real paper, which is inexpensive, "easy to use" and convenient to use > for somewhat static information. On the other hand, we have well-known > computer display technology for "dynamic" information, like web sites. As > such, I didn't (and still don't) see a board market for this e-paper toy stuff. I think you are seriously wrong about this. The "greening" of America has pushed a lot of otherwise not very useful technologies to fruition. Look at "Energy Star Ally" computer equipment, and look at the studies on total lifecycle energy costs. If you are willing to dig for it, look for the University of Colorado, Boulder's report on pollution per mile for oxygenated fuels in vehicles with Oxygen sensors (i.e. vehicles manufactured from 1981 onward), or other similar studies by other researchers. If you aren't willing to accept "green" reasons for such paper, then accept the "Enron" reasoning that it's a lot easier to "shred" documents by refusing to display them. Personally, I can think of a huge number of reasons to have "smart paper". > It somehow reminds me of the prediction made in the 70's, which basically > said that before the next century, offices would work without any paper. > Obviously, this didn't happen. Most predictions like this don't happen, and > if they do, then mostly a whole lot different that originally imagined. We were also supposed to have flying cars. Just because one prediction fails doesn't mean that they all will. The "paperless office" was a seriously idiotic idea; we never had the tools, until the last year Clinton was in office, to treat electronic signatures as real signatures, making electronic documents into legal documents. The main reason for paper trails is legal. I think that it *is* happening, actually. Try to get a printed copy of a software manual from most software vendors, these days; or have H&R Block do your taxes, and then tell them that you don't want to pay your tax bill (or receive your refund) through direct payment/deposit, electronically, and have them look at you funny for wanting to send (or be sent) a check. > Bottom line (and at this point I really want to stop wasting your bandwidth > and precious time): I guess that looking at the computer and electronics > company, "all the good ones are taken" or "everything that can (sanely) be > done has been done". Of course, the future may bring the one or the other > new interesting development, but I don't see many real revolutions anymore, > as we seem to have reached a point where going any further does no longer > provide any increased benefit. > > Comments to this - well - rather free-style "essay" are welcome ;-) You're seriously wrong. So seriously wrong that RIAA and MPAA are paying off congress-critters left and write to jam through legislation to prevent the possibilities they see, with their limited vision, from happening. I'll leave it at that. 8-). -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3CE3457D.3EC8885B>