ers on-station for long. VNI's system sends one reporter with a hand-held camera, so that it costs little more than putting a print journalist on the scene; the camera lets the journalist do the editing as well.
A recent documentary about a reporter on a cruise and expedition to the South Pole was finished in weeks and cost about 5 percent of what such documentaries usually cost. Look for many others in the future.
With the Niven/Pournelle novels
Lucifer's Hammer
and
Footfall
,
I've made quite a lot of money out of hitting the Earth with large and heavy objects. Chris Chyba of Princeton presented a general model of meteoric impacts; did you know that the Earth is hit with megatons of impact energy every year? Fortunately, most of it burns up at high altitude.
Chyba's model makes it clear that the mysterious Tunguska event of 1908 in Siberia wasn't a small black hole or a comet, but instead a stony asteroid that exploded with
15 to 20 megatons of energy about 10 kilometers above ground zero--fortunately, before the U.S.S.R. had retaliatory weapons poised and waiting. The asteroid had to be stone, because an iron asteroid would have left an impact crater like the one near Flagstaff, Arizona, while an ice-ball would have exploded so high that it wouldn't have had much effect. It would be pretty easy to use Chyba's data and Allegiant Technologies' SuperCard on a Mac to build a general case model of asteroid and comet impacts, and one of these days I may do that.
Another new model: for the first time, I've seen a convincing case for doing something about human action in the atmosphere. NASA's Goddard Institute for Space Studies now understands how fluorine/chlorine compounds, such as Freon, can affect ozone in the high stratosphere. It's complicated, requiring aerosols injected into the upper atmosphere by volcanic action like that of Mt. Pinatubo to act as reaction surfaces for halogen catalytic activity. The bottom line is t
hat the stratosphere is saturated with halogens from human activities, and it doesn't take a lot of new halogen release to sustain high levels.
The best general data source up to now has been the Stratospheric Ozone CD-ROM from Lenticular Press, for Macs only. The CD was compiled before the role of aerosols was understood; even so, analysis shows that ozone depletion tracks volcanic events.
Alas, there aren't comparable data sources for a cost/benefit analysis of what to
do
about this. Banning Freon entirely has some heavy costs, mostly financial in the U.S., but life-threatening in much of the Third World. The costs of ozone depletion and consequent increases in UV levels, particularly in the tropics, have hardly been looked at. There is probably enough data floating around on the Web for a decent analysis, but I haven't seen that done. The important thing is that we're running an uncontrolled experiment on the atmosphere, and that doesn't sound like a very good idea.
On the ot
her hand, the evidence for global warming is where it was a year, or even a decade, ago. The theory has been around since the turn of the century, when Arrhenius made back-of-the-envelope calculations. If you increase greenhouse gases like CO2 (from burning fossil fuels) and methane (a lot comes from the flatulence of domestic cattle), there ought to be consequent warming.
We've certainly added significant greenhouse gases to the atmosphere in this century. However, our most sensitive instruments--earth satellites that can detect the night temperature differences between a full and new Moon--aren't finding new warming.
There's certainly been warming in the past 200 years. In 1776, Lt. Col. Alexander Hamilton dragged cannons captured at Fort Ticonderoga across the frozen Hudson; nowadays, the Hudson doesn't freeze that hard. It's difficult to connect that warming with human industry, because most warming took place before 1900. In this century, most took place before 1950. There's been even less
cost/benefit analysis of global warming and its prevention than of stratospheric ozone depletion.
These matters are important. We rushed into "prevention" of "acid rain" without adequate understanding and wasted billions of dollars fighting the wrong problem. Now that the supercomputers at Los Alamos and Livermore aren't needed for weapons design, perhaps they ought to be turned loose on these potentially costly environmental issues.
Meanwhile, the economists keep looking for the increased productivity the computer revolution ought to have brought about, and they're not finding as much as they expected. I think I know where it went: small computers finally made it possible for governments and companies to collect all the information they want, and much of our productivity increase went into compliance with reporting requirements, enforcement of compliance, and not much study of the resulting reports.
While in Baltimore, I installed Earthlink Total Access for Windows
on my eld
erly Zenith 486 notebook. In about 5 minutes, I had full access to my Earthlink Internet account. It couldn't have been more painless, partly because I used a Supra 28.8-Kbps external modem. PC Card modems are light and convenient, but they still can't beat a good Supra or U.S. Robotics external modem.
I'd also intended to test Zenith's new Z-Note MX Pentium notebook, but unfortunately its screen failed. My son Richard volunteered to do a test. He borrowed the machine and called Zenith without telling them it was an evaluation machine. Being a new model, it was under warranty, so the technical-support people told him to ship it to them. He sent it on Friday; Tuesday morning he got it back in perfect working order. Now my only problem is getting it away from him: he loves it.
Richard doesn't use a desktop, and for complex reasons had to give up his IBM ThinkPad. After exhausting the Windows serial-port file transfer options, he used LapLink for Windows 95 to move all the ThinkPad's files and prog
rams to the Z-Note MX. Now he's experimenting with LapLink's Internet file transfer capabilities. If you simply have to get your files transferred, use LapLink. It always works.
He tells me that Zenith's AccuPad mouse substitute takes getting used to, but once you've used it awhile, you will prefer it to trackballs and eraserheads. The machine also gets hot; at least once he had to shut it down after several hours of continuous use.
More next month. It looks like he'll buy the Z-Note MX, and after he does, I'll get another from Zenith. Laptops are getting so powerful that many businesspeople don't bother with desktops. Zenith's Z-Note MX is good enough for that.
If you have SCSI hard drives and Windows 95, be very careful what you add to the system.
Syquest has dominated the high end of the removable drive market, largely through speed. Their EZ135 drive works fine with Apple SCSI, and we had no problems with it in a Windows for Workgroups machine, or in RacingCow, the Gateway 2000
P5-133XL with Win 95--but RacingCow's hard drive is IDE.
The problem came when we installed it on Pentafluge, the Pentium 60 that seems terribly fast because it is all SCSI, including the boot drive. The Distributed Processing Technology SmartCache III SCSI Host Adapter really speeds up disk operations, and for what I do, disk operations are what limit system speed. When I write, I save early and often.
Installing the EZ135 on Pentafluge looked easy enough: just use a Granite Digital SCSIVue Gold Diagnostic Cable to connect it into the external SCSI port of the SmartCache and turn the machine on. At first glance, it looked as if that had worked; the drive appeared in My Computer, and I could read and write to it. I did some read/write tests before I realized that the EZ135 was the D drive. The D partition of my hard disk had become the E drive, the Maxoptix T3-1300 optical drive that's usually E had become F, and the CD-ROM drive had become G. Since a lot of programs have drive locations mapped
into them, this wouldn't do.
I shut the system down and disconnected the SCSI drive. When I rebooted, Win 95 appeared to load, but the hourglass never went away; I couldn't get control of the system. Hardware reset and then reboot in safe mode by holding down the Shift key as it starts up. That worked. Clicking on Disk Drives in the Device Manager tab inside the System icon in Control Panel revealed that the system still believed a Syquest drive was attached to it. I deleted it. Alas, the Maxoptix T3-1300 seemed to have vanished. You can't do a hardware search from safe mode.
Rebooting produced the eternal hourglass again. I tried booting with floppy disks and with the Norton Rescue disk. I tried curses. I tried evil and potent magic. Nothing worked. In desperation, I did Ctrl-Alt-Del at that hourglass. I found that Explorer, which launches automatically when Win 95 starts up, was hanging the system. I told Win 95 to halt Explorer, and behold, I had an active desktop again.
My Computer s
howed me what was wrong: several of the drives were mislabeled. Back to Control Panel, System, Device Manager, Disk Drives, cursing how hard it is to find things nested that deep. Eventually, I had all the drives properly labeled. Reboot again. This time, Norton Utilities told me I had C drive problems. When I ran Norton Disk Doctor, it said I had over a hundred errors, one in nearly every subdirectory.
Disk Doctor said it had fixed all the problems, but when I ran it again, some remained. I shut down and rebooted in DOS; then in Win 95 DOS, I ran Disk Doctor one more time. It found errors and fixed them. That did it, and Pentafluge was his old self again.
I haven't the foggiest notion of what went wrong. Syquest has some notoriously bad device drivers for the EZ135, but we hadn't installed any. Whatever happened took place at a much lower level. I wasn't particularly eager to have this problem happen again, but it did seem important to understand the situation. Therefore, I got out the old Pion
eer DRM-604X six-pack CD-ROM minichanger and hooked that up the same way I'd installed the EZ135. When I booted up, the DRM-604X installed itself as drives G to L, just as it was supposed to do, and everything worked fine.
The next experiment was to shut down, turn off the DRM-604X, and reboot. Again all went well. The CD-ROM drive remained F, the Maxoptix T3-1300 was E, and the two partitions of the big SCSI hard disk were C and D. I then experimented with a Pioneer DE-UH7101 optical drive, with the same result. It became G without any problems, and the DE-UH7101 plus the DRM-604X worked together, without changing the designations of the C to F drives. All that SCSI stuff worked the way it's supposed to work with Win 95, real Plug and Play.
Finally I gulped hard, disconnected everything else, and reconnected the EZ135. The result was the same as before: it installed itself as the D drive and everything was a mess. I did as little as possible before exiting from the system and disconnecting the
EZ135. Reboot gave me the endless hourglass. Reboot in safe mode; open Control Panel, System, Device Manager, Disk Drives; remove the EZ135 from the drive list; exit and reboot in regular mode. Ctrl-Alt-Del to close the endless hourglass.
Device Manager again, this time to tell it that the CD-ROM drive was F and not G (and I don't understand how it got displaced to G since I had done all I could to nail the drive-letter assignments down). Reboot once more, and rearrange my desktop, which was all scrunched up because safe mode uses 640- by 480-pixel screen resolution and I like 1280 by 1024 pixels with my ViewSonic Professional Series PT810 21-inch monitor. Now I knew the problem was reproducible.
The next experiment: we have a new Pioneer DRM-624X six-pack CD-ROM minichanger. I tried it and got the eternal hourglass. When I got past that with Ctrl-Alt-Del, the system couldn't find the new drives, but everything else worked, and recovery consisted of turning off the machine and disconnecting the
DRM-624X. Understand, the DRM-624X works fine on IDE systems under Win 95 and is a lot faster than the DRM-604X. I'm pretty sure it will work if I install its drivers. The question is, why does the older one work automagically, while the new one does not?
Next a call to Distributed Processing Technology. Their Win 95 drivers are on the Microsoft installation disks, so I didn't need any updates. No one had asked them about support for Syquest removable drives, but they did understand why the drive wanted to be D; the disk cartridge was formatted as a primary partition. Try booting up with no cartridge in the EZ135.
I did that, and this time it didn't displace the D hard drive, nor the E optical drive. However, it did brush aside the F CD-ROM drive and take over its spot. Installing the latest Syquest drivers obtained from their Web site didn't help. The drive worked, but it was determined to be the F drive, and nothing I could do with Device Manager would convince it otherwise.
What if I b
ooted up with the drive installed but turned off? I tried that, and the system blew up; it was back to boot in safe mode and all the rest. Needless to say, we have consigned the Syquest EZ135 to work with the Mac, where it seems better behaved.
I don't know what is going on. Apparently, Win 95 can handle some SCSI devices--particularly older ones--but not others. Perhaps the new Service Pack patches will cure the problem. Perhaps not. The bottom line is, if you have IDE hard drives and Win 95, adding SCSI devices is simple.
If you have SCSI hard drives, be very careful,
and if things lock up, Don't Panic. Just remember to Ctrl-Alt-Del at the endless hourglass and shut down all nonresponding programs. Then get to Device Manager and fix things. It's tedious, but not as tedious as reinstalling Win 95. By next month I'll know more. Meanwhile, be careful.
Eric Pobirs, who helps out at Chaos Manor, got an Iomega Zip/Z100P drive at Fry's last week. This connects to the para
llel port; you then install some driver software, and it looks like another drive. We tested it on RacingCow, and it worked perfectly. It also worked on SuperCow and everything else we tried it on.
The Zip/Z100P is rapidly becoming the standard sneakernet method for large file transfers. It's not as fast as the EZ135, but it's small, lightweight, and, most important, extremely easy to install and use. It works for backup, archiving, and file transfer. While it's slow compared to Ethernet, it's darned fast compared to floppy disks, and when you remove it, there's no evil aftereffect.
If you've invested in a big machine, you should seriously consider getting an Iomega Zip/Z100P not only for backup, but to carry your working environment from machine to machine. Some of you may remember that was the goal of the Tandon systems back in the IBM AT days. The difference is, the Zip/Z100P works, and it's cheap, too. Recommended.
We've been doing a lot of Web exploring lately, and I'm impressed with
the potential of Java. For instance, there's no reason you couldn't get a flash-RAM PC Card, program it with your computer, and insert it into a slot in your car. A Java program would tell the car to configure itself for you, adjusting the seats and mirrors and suchlike to your satisfaction.
And Eric suggests Java for Genomes. Once the human genome sequence is completed in the first part of the next century, you build an organic compounds constructor into your desktop, connect that to a hypodermic needle, and download your health updates as a Java program to instruct your system. Maybe like Larry Niven's autodoc machines it will give you a manicure as well. On reflection, I'm not sure I won't live to see something like it happen. It will drive the FDA nuts.
I was one of the early users of MCI Mail, and I still have my account, but I'd sort of fallen away from using it because it was a pain to read and answer mail on-line. I had for years let Norton Commander for DOS take care of MCI Mail, but f
or some reason, it doesn't want to do that under OS/2 Warp or Win 95. Even if I figure out how to make it work again, it was never faster than 2400 bps.
MailRoom from Sierra Solutions does all that Norton Commander ever did and works at your modem's speed. It can automatically get your MCI Mail at scheduled times, send and receive MCI faxes, keep an address book, and sort incoming mail into folders. There's also a decent off-line editor to build new messages or compose replies.
The only disappointment is that MailRoom won't handle mail from other sources, like BIX or direct Internet mail, so I'm stuck with three mail readers. However, with MailRoom, MCI Mail is the most painless. If you use MCI Mail, you need this program. Recommended.
Oracle and IBM have been pushing the notion
of dumb Internet terminals for the obvious reason that the only things these can talk to are big mainframes running highly sophisticated software. The problem is that the chip industry has already go
ne well past that. You can now put on a single chip a computer complete with wave-table sound and a modem that's more sophisticated than the "Internet computer" was supposed to be. The next generation of game boxes like Sega Saturn will incorporate those and should eat out the bottom of the dumb terminal market. Gateway is coming at them from the top with what amounts to a really smart TV set.
Gateway already makes a dream game machine. RacingCow can take a game like 7th Level's Arcade America and run it with action faster than I can keep up with. Arcade America isn't the kind of game I enjoy playing, but I like watching it played. It's a sort of Super Mario Brothers with incredible detail, and very cleverly done. The P5-133XL I have came with a Gravis Gamepad game controller (also known as a four-button joystick) that most of these arcade-like games recognize and most game addicts are accustomed to.
Arcade America in theory will run under Windows, but in fact installation is difficult. It's als
o less fun: you have to use the keyboard controls (or do weird things with a two-button joystick) because there's no support for the Gamepad. What this game really wants is Win 95.
Coming from the opposite direction is Interplay's Whiplash, a car-racing game that in theory will run in DOS; in fact, even on the IBM Dominator, the fastest 486 we know, it's no fun at all. Whiplash really requires a 90-MHz Pentium or better. One wonders why, given that anyone with a P-90 or better is likely to be running Win 95, Interplay rushed it out as a DOS game that can't use the Win 95 sound interface. On a high-end machine like RacingCow, though, it's pretty spectacular, both sound and graphics.
Over a year ago, I noted that Intel believed the home market--particularly the game market--was their best bet for really high-end systems. That's proven to be true, and the game developers have rushed in. With games comes sophisticated multimedia capability, and, except for text-to-speech, the Intel platforms have cau
ght up with and passed the Mac in that regard. (What this will do to traditional Mac dominance in education isn't clear.)
Tools like SuperCard made it possible to build something as wonderful as Myst on the Mac; but more copies of Myst sold for Intel platforms than for Macs. When Myst first went to the PC, Broderbund Software got about 100 times as many technical-support requests as the Mac version did; but that was before Win 95. Now, the Discovery Channel's Multimedia's Connections, a game built around James Burke's TV series and book, and Interplay's Frankenstein: Through the Eyes of the Monster, both far more sophisticated (if not more interesting) than Myst, run on RacingCow right out of the box. No technical support needed.
Apple has finally come around to making it easier to port games from Intel systems to the Mac, and SuperCard continues to improve in both capabilities and ease of use. So does Visual Basic. Of course, most games aren't developed in either SuperCard or Visual Basic; but
when the hardware gets good enough, that will change. The neat part is that competition is driving the industry toward better hardware and easier-to-use development tools; and that's good.
The
book of the month
is
The Beginnings of Rome
by T. J. Cornell (Routledge Publishing, 1995). As much as you want to know about the founding of Rome. If you like that kind of book, you will like this one a lot. The
computer book of the month
is
BYTE Guide to Optimizing Windows 95
game of the month
is Westwood Studios' Command & Conquer, a tactics/strategy game you can play against the computer or on-line against live opponents. Good action and graphics, and, alas, addicting.
The
program of the month
is Intu
it's TurboTax--it's certainly the most vital to me as I get my taxes done.
I know I promised more on the Intergraph; next month for sure, as artist David Em reports on what he's been able to do with it. Also, starting next month, a Web site of the month.
Product Information
Arcade America..........................$39.99
7th Level, Inc.
Richardson, TX
Phone: (800) 884-8863 or (214) 437-4858
Fax: (214) 437-2717
Internet:
http://www.7thlevel.com
Circle 1065 on Inquiry Card.
When D Equals E
