Interludes and Examinations

October 21, 1996

Ah, yes. I knew there was something I forgot last time.

We all read about New Star back during the Great Auction of 1995, they were one of the bidding parties on hand for the Commodore sale. Shortly afterward, they licensed the Amiga chipset from Escom, along with VISCorp and Tianjin.

While we'd forgotten about them, they've been busy.

"Amiga back from the dead?

New Star, an Amiga computer clone, hit the streets in Taiwan for less than $200. That's well below the $500 target for a network computer, and the Amiga system includes storage and other features forsaken by network-dependent designs."

New Star is licensed to sell Amiga clones in Taiwan only. This magic $200 superbox isn't coming stateside anytime soon. But it is still quite a cool thing. For $200 you get a 68000-driven, ECS machine with a CD-ROM drive and Kickstart 3.1. Considering Commodore used to think we'd pay $800 for a similar box, I'd say that's a decent deal - used A500s go for about $120 these days, 3.1 goes for $140, and CD-ROMs are all over the price range.

The trick here is that it costs $200. There's this weird notion in the computer industry with regards to the $500 Internet box everyone keeps talking about - this notion that at the $500 price point, people start buying things. People didn't buy the C64 when it cost $500. People didn't buy the A1200 when it cost $500. People didn't buy the Tandy Color Computer when it cost $500. However, the C64's dream price was $175, the A1200 peaked at $350, and the Tandy CoCo sold most of its units between $150 and $300. Now, a $500 Internet box that costs $200... there's a price point.

A couple things to consider, though. A 68000 Amiga with ECS is NOT A WEB BOX. There is no way to make a usable Web box out of it. It's just too resource intensive. You need a hard drive and/or a lot of RAM. You need a decent speed processor to decode the graphics. You need screen modes that will show more than 16 colors in hi-res because Web pages look like crap on low-res screens on a television. (For that matter, most Web pages look crap on a television anyway...) And a very important component of the Internet box that the New Star box apparently also doesn't have? A modem. This is not an Internet computer. This is exactly what it says: a $200 home computer. Not a toy computer, just a home computer that does what the home computer user needs it to do and no more, for a price that's nice.

ESCOM never learned this. Whether VISCorp has learned this or not remains to be seen. Meanwhile, the New Star unit has caught the eye of several industry notables - even John Dvorak, Mr. What's-an-Amiga himself, is impressed by the idea of the Amiga making a comeback in the ultra-low-end market - and also seems to have hope that the "real" Amigas can make a comeback as well. New Star seems to be doing the first Amiga advertising in six years - simply by existing.

So here's the next thing we need: an ultra-cheap 68030 AGA machine in the States. Forget A1200 and CD32 style cases, build the damn monitor in a la the Macintosh TV and the Toshiba TIMM. Detached A4000-style keyboard, inbuilt floppy and 8x CD-ROM (they're cheap now), a 25MHz 68030 or better, decent size hard drive, 32MB RAM (that's cheap now too), a Trinitron multiscan tube, and a TV tuner. Throw in a framegrabber - you've just invented a perfect multimedia machine. Price point: $800. The motherboard would be cheap - the RAM will cost $100 before the end of the year, the CD-ROM $100, the hard drive $150, the rest of the motherboard maybe $50. The monitor tube and such $250 or so for a 15". Casework and power supply $100. Frame grabber hardware and tuner $100. This is all current prices - yes it adds up to about $1000 right there - but it wouldn't be that hard to cost-reduce things, build the frame grabber onto the motherboard, scrimp on the hard drive and CD-ROM, etc. and bring it down to $800 before 1997 rolls around.

I would buy such a machine.

I had fun this weekend. Saturday I spent all day moving stuff - I finally got an apartment, and soon will be signing off permantly from the old Rat's Nest and reappearing at Headquarters Frigate - and so was physically beat from all that. So Sunday, how do I relax?

Drive 180 miles.

To AMICON's Midwest Amiga Expo.

I left the Rat's Nest in Seymour at about 8 AM Sunday. I arrived at the high school in Columbus, Ohio (actually Westerville) at what looked like 12:20 Indiana time, turned out to be 1:20 Ohio time (that wonderful Indiana time zone strikes again - why can't the rest of America do away with daylight savings time and be normal like Indiana?) - just in time for Jason Compton to take the stage and tell us there's no news.

Jason is one of those odd figures in the Amiga industry, he's considered the Good Guy despite the fact that there are people out there who hate his guts. I'm not one of them - in fact, I think he's done a better job with Amiga Report than I ever would have, I'd have ended up pissing MANY more people off. VISCorp hired him because Amiga people trust him - and all of a sudden Amiga people stopped trusting him. When Compton says "VISCorp will build Amiga desktops" there's still a crowd that says "He's lying." When Compton says "There's no news" there's still a crowd that says "When's it going to be over?" This is part of journalism - I remember the start of the Gulf War, where everyone wanted to know what the hell was going on and nobody knew anything because there wasn't yet anything to know.

But anyway. He doesn't come across as the slimy weasel we expect a PR person to be - he's an average Amiga user like the rest of us, he's seen the BeBox, he still likes the Amiga, he hung out around the booths at the Expo like all the other dudes (until a crowd gathered to trade rumors), he just happens to work for a company that might conceivably own the Amiga soon. He's intelligent, he can think on his feet without embarassing himself (like I would have done), he's just rather hamstrung right now by the fact that he doesn't have anything to report.

So the news? Bill Buck is STILL in Germany. The 30 days have expired but they're still talking, so that apparently isn't going to be a problem. The story as Jason tells it is that, in January, Bill went to Germany to talk about licensing the chipset, and discovered ESCOM's financial difficulties, called home and said "I think we can get a pretty good deal on the whole thing," and the rest is history. VISCorp isn't talking RISC yet, because when they've tried to talk to the processor people (Motorola, etc) they've basically not been taken seriously because they don't own the Amiga yet. This makes perfect sense, because most Amiga owners aren't taking Viscorp seriously either until they own the Amiga.

Jason has a "Dead Ed" with him, a nonfunctioning prototype of ED, very possibly the same unit seen on VISCorp's homepage. It's teeny-tiny, has a cable modem-like module in it, has a serial and parallel port, an AT keyboard port, phone and TV and A/V jacks, and supposedly, on the underneath board which Jason says he cannot get to because the top board won't come out without breaking (he tried), are Lisa and her AGA sisters. There's also a Triton video controller on the prototype which he says will probably be left out in the final edition.

The Round Table Discussion got off to a late start, Jason was out on the show floor with a crowd of us, trading rumors and stories, and showing off the Dead Ed, he'd just found an empty table to set up at, when they came on the PA system and said the Round Table Discussion was about to start. I pointed out the table he was at had rounded corners - close enough? - but no, off to the auditorium we all went. "Follow the crowd," someone remarked.

"I thought Amiga people weren't supposed to do that," I added.

Well, partially - Amiga users are unique in that we do follow crowds, but we're very particular about what crowd we follow. We followed Jason to the auditorium because we knew what was there and because we knew we wouldn't become prisoners once we arrived. We follow the Amiga-owning crowd because they're doing cool things we want to do - not simply because they're there, like the PC crowds. We don't follow the crowd because we have to, we follow the crowd because we want to - and we follow the crowd only as long as we want to and not a minute longer.

The panel was very much an Amiga All-Stars - Eric Schwartz (who "draws things" as he put it), Don Hicks (of Amazing Computing or Amazing Amiga depending on how you read it), a couple local Amiga dealers, Jason Compton, Kermit Woodall of Nova Design, and one of the guys from Silent Paws (the laptop dudes - though admittedly you need one hell of a lap). From the introduction on, Jason pretty much owned the show, as most of the questions were asked of him and not the others. But it was fascinating nonetheless.

There was a recurring theme of the Amiga as personal workstation. An audience member brought up the issue of pricing - since the Amiga is a unique thing, it doesn't need to be priced to compete with PCs - but I have to disagree to a point. No, the Amiga doesn't need to compete pricewise with the $1200 133MHz Packard Bell toy computers. But no, the Amiga 4000T with the 25MHz 68040, behind-the-curve AGA, and munged-up 68030 memory architecture doesn't have any business costing $2600 and up. Cost cutting is in order - especially since we know the machines don't have to be priced like low-end Suns. As Jason pointed out, France (and for that matter, Pennsylvania) isn't the cheapest place to make computer stuff. The 4000s should cost about $1500 - not competitive with PCs, but not outrageous either. A1200s should cost $400 - not competitive with used Macs, but not outrageous either. And of course, you talk about workstations and inevitably someone will mention the purple case. (The new SGI workstations, btw, have vertical pillow-shaped neon purple cases, sorta like a Walker on acid.)

The Walker is dead. Officially VISCorp's engineers have said that to complete Walker will take longer and cost more than to build a new design from the ground up that's better and doesn't look like the head of a penis. (The Freudian computer - the more "modules" you have stacked vertically between the base and lid, the more male pride you have - never mind above two modules there would be timing problems from all the wiring length!) AAA is more or less dead, because as Jason said, the existing AAA boards are in the hands of former Commodore engineers! (See Unsolved Mysteries for more on this.)

VISCorp will probably focus its advertising on those magazines that have been cool to the Amiga in the past - Byte for example - instead of those magazines that seem to go out of their way to ignore the Amiga - CEV Video was listed as an example, listing nonlinear editors in a review and never even mentioning the Flyer. When asked why they ignored the Amiga, the editor said "The Amiga hasn't been manufactured in years." As Jason said, there's no point in "rewarding" magazines that have thumbed their noses at the Amiga for years - or for that matter, trying to be noticed on page 499 of Computer Shopper. Also they'll probably stick to large trade shows - the non-Amiga-oriented shows - hoping for the shotgun approach. Another common theme in the discussions was that Amiga owners are Amiga owners because they LIKE using the computers - while PC owners are PC owners because they HAVE to be. People who WANT to use computers and who LIKE computers will be hanging around at the trade shows wishing there was something different... and the Amiga certainly is different.

I didn't get a chance to buy anything at the show. By the time the Round Table let out, the show was closed down and people were packing - Indiana time zone strikes again.

But at least I went. I saw. I didn't buy. Veni, Vidi, Viscorp.

Other news: Apple shocked the pants off Wall Street last week when, holy shit, they made a profit. Pundits had been predicting a loss of about $37 million for 4Q96, but eyebrows went up industrywide when Apple announced profits of $25 million. The joke here in the office was that some Apple accountant was digging through the crap on his desk and found an uncashed check for $52 million.

Meanwhile, Apple's made the routine pricebumps on its Performa 6400 line - you know, the Macintosh Pillow - bringing the $2600 200Mhz model down to $2199 - and has introduced the Performa 6360. I have not seen pictures of the 6360, but I suspect, given the number, it probably looks something like the existing Performa 636. A 160MHz 6360 will go for $1499 with 16MB and an 8X CD-ROM - a price that rivals the best that competitor Motorola has to offer. Apple is finally learning what the clone market means - it means you have to run to stand still, it means more bang for the buck or you lose.

One other product worth noting, which was announced in the same press release: a system I've been jokingly calling the Mac Toaster. It's a 200MHz Performa 6400 with 32MB RAM, a 2.4G hard drive, and Avid Cinema - drive one off the parking lot today for $2699.

This product is interesting because it's a perfect example of how Apple is doing right everything Commodore did wrong. Commodore released an A3000 that was incompatible with the Video Toaster - excusable because the 3000 and the Toaster were in development at the same time. Commodore released an A4000 that was incompatible with the Video Toaster - excusable because... well... it wasn't excusable. NewTek had to bear the brunt of work to make the Toaster work with the 4000. And then due to Commodore's arrogance, NewTek wound up reselling A2000s and A4000s with "Video Toaster" stickers over the Commodore nameplate - because Commodore kept right on ignoring them, never acknowledging the fact that the Toaster was for awhile the main reason to buy an Amiga. Now, all these years later, Apple is going after the old Toaster market with a video editing Macintosh, decked out with a third party product taking top billing.

Whether you like the Mac or not, you have to admire Apple's approach.

I was talking with one of my informants recently, about custom Amiga cases - more precisely, putting an A1200 motherboard and a 9" RGB multiscan screen into a Mac Classic-like case. I kind of haphazardly made the suggestion of a wooden case, so it would look like a 1970's stereo unit - and then thought about it for a moment.

Hard wood. Glued and pegged together. Nicely sanded and carved. Vent slots chiseled into the front in some artistic configuration. Wood grain floppy and CD-ROM bezels. Slots for the power supply and motherboard. Polyurethane. Very COOL looking designer computer. In my case, that could mean a fan-cooled A1200 with drive bays. This is gonna be awhile in the making, of course - I'm halfway through moving into my apartment - but all I'd need is one case and I could end up with a job building them for people. :-)

On October 5, Seymour Cray passed away.

Seymour R. Cray was one of those people who were like Jerry Garcia - so much a part of the modern world that it's impossible to imagine a world without them. Cray's name is literally synonymous with supercomputers - and to a lesser degree, to offbeat computer designers.

Cray was a computer pioneer since the days when everyone who touched a computer was a pioneer. For Control Data Corporation, he built the CDC 1604, the first commercial computer to use individual transistors instead of vacuum tubes. In 1963 the CDC 6600 became the first computer in the world to top 3 MIPS - blowing IBM's 7094 out of the water. (The CDC 6600 was a cool computer for other reasons: it had a Freon-based air conditioning system to prevent overheating.)

Mr. Cray left CDC in 1972 to form his own company, Cray Research. From here came the legendary Cray I, in 1976. (To put the passage of time in perspective, you can buy a $7,000 quad-Pentium Pro system today that will do the work of the 1976 $8.8 million Cray I.) As Cray wrote years earlier while at CDC, when asked to write a 5-year business plan, "Five-year goal: Build the biggest computer in the world. One-year goal: Achieve one-fifth of the above." It took him four years.

But Seymour Cray succeeded at these things and more. He felt a computer should look good - he was as much artist as technologist, foregoing the standard "box" and "giant reel-to-reel player" look of standard 1970's big boys and produced computers that looked like architecture. An oft-quoted anecdote is that certain Cray supercomputers are so fast that no wire inside them can be longer than sixteen inches or the entire computer slows down. (There is a practical reason for the exquisite C shape of the Cray I - component placement.)

Cray was fixated on gallium arsenide. It's a weird substance - it makes for fast processors, if you can actually figure out how to manufacture a chip from it. He intended to use gallium arsenide for the Cray 2, but wound up retreating to silicon instead. (But the Cray 2 was notable in itself - it used Flourinert, a fluid that looks exactly like water but does not conduct - the entire computer was dunked in this stuff for cooling.) The Cray 2 was not the commercial success Cray could have hoped for - but the Cray XMP, a redesigned Cray 1, was a much better story, saving the company from disaster.

Cray himself left the company to form Cray Computer, and begin working on the Cray 3, again aching to make gallium arsenide work. All sorts of custom machinery needed to be made, and $300 million later, he sold the Cray 3 project to another company in 1989. The Cray 4 met with a somewhat worse fate when Cray Computer filed Chapter 11 in 1995 - the advent of cheap processors from places like Intel and DEC were causing the death of the supercomputer industry. Not long after Cray Computer fell, Cray Research itself was bought by Silicon Graphics, makers of high-end graphics workstations who also happened to have an interest in supercomputers. (Earlier this year, Convex succeeded in making processors out of gallium arsenide - and were then bought by Hewlett-Packard.)

Mr. Cray had been in the hospital for two weeks after suffering head injuries in an auto accident. He was 71.