We May Lose or We May Win (But We Will Never Be Here Again)

March 10, 1997

The other night I had this dream. I dreamed that Rush Limbaugh had formed a barbershop quartet and was singing "Windows NT" to the tune of the Doors' "People Are Strange," while my ex-boss's wife undressed for the crowd. I don't care who you are, you will find something in that dream to disturb you. I do wish I could remember some of the lyrics.

This has little bearing on the rest of the column this time, it's just there to get you into the same utterly weeeeerd frame of mind I'm in right now. I've got some personal things going wrong in my life right now, so I'm just a teensy bit paranoid - perfect for writing the Rumor Mill.

First up, the Amiga goes RISC. Phase 5 finally announced release dates and MSRP's for its Cyberstorm and Blizzard PowerPC accelerators for the 32-bit Amigas. They are, as we'd been told, dual-processor cards, using a 68K-series chip to run existing Amiga software, with the PowerPC called only when needed.

But there's a few nifty features these cards have. For one thing, there's a special AGA version of Cybergraphics - Phase 5's retargetable graphics software - that uses the PowerPC for its acceleration and blitting. This is even cooler when you realize that Cybergraphics doesn't use Chip RAM the way graphics.library does - a Power Amiga with Cybergraphics AGA should use less chip RAM than conventional Amiga software.

For another, these boards ship with Mesa and OpenGL, both of which are the 3-D equivalent of graphics.library - they're 3-D programming interfaces, and OpenGL is cross-platform. NewTek, among others in the 3D software business, have expressed interest in having OpenGL interfaces on the Amiga... now they've got it, PowerPC accelerated with no emulation overhead.

But best of all? The A1200 version - a 120MHz 603e, no RAM, and the 030 chip socket empty - will retail for $400.

$400? WHAT?

Yep. You find the 68030 somewhere else - off a cast-off accelerator card, or go to the store and buy one, they only cost about $80, I don't know whether the board uses FPGA or PLCC chips - FPGA is the "metal"-looking one with the pins sticking out the bottom, PLCC is the black one with the legs around the edges (like absolutely every chip in an A1200). (Since the PowerPC is a PLCC-type package anyway, I'd think they'd go PLCC with everything - but the socket is a different story.) But you put that on the board, add your 32-bit RAM, add the SCSI module... and you have ONE HELL of a system.

A 603e at 120 will benchmark just slightly above a similarly clocked Pentium. This then becomes the perfect show-off Amiga - and what are Amigas for except showing off to friends? - this "keyboard" that can kick your neighbor's Pentium's ass AND still run those hardware-munching Black Lotus demos.

These boards ship in May, and if I can find an extra $400 by then, assuming Phase 5 ships on time and at the announced price, I may finally get that 603 in my 1200 that I've been ranting about for months. I just hope the 030 side includes a math unit - I'd hate to have to sacrifice FPU compatibility like that. If it were emulation I might not mind so much.

The A3000/A4000 models are 604e's running at 150, 180, and 200MHz; at the other end of the card lives a socket for an 040 or 060. I can't remember the prices off hand, but they're not much more than for a straight 040 accelerator - they just don't include the 040 processor itself.

There is talk that Phase 5 is going to incorporate a subset of the A\Box - namely the Display List RISC Processor segment (the Blitter from Hell) into future Cybervisions or as an upgrade to certain Cyberstorms. This will be very cool. Not all of the A\Box functionality would be there, but that DLRP would be faster than the AGA blitter at any rate.

But speaking of A\Box, has anyone noticed they have no OS strategy for it?

All that nifty droolworthy hardware, and so far no one seems to know what kind of software they'll run on it. There's newsgroup talk of an "Amiga-isch UNIX" - their spelling not mine - but since they're apparently not cooperating with ProDAD, we can count out P-OS as an option. And UNIX is not AmigaOS. UNIX is designed for multiuser command-line work, and is not particularly optimized for single-user realtime graphical stuff the way Exec was. How they plan to provide compatibility I don't quite know - and my guess is neither does Phase 5.

PIOS is making headlines in an unlikely place: the Mac rumor mills. Beginning with my Macintosh counterpart, Mac the Knife, and his cryptic mention of "Pious Computer making a Mac clone that runs the Amiga OS with a 68K add-in card" - PIOS is now gaining some name recognition in the Mac market. PIOS' first big project, the PIOS ONE, is largely a Dave Haynie design - with a "monolithic" processor card that contains most of the memory subsystem; you swap the entire daughtercard when you upgrade, but by doing so, you replace the entire heart of the machine, not just the processor; a lesson he learned from the A2630 series, isolating the memory and processor from the parts of the bus that change less rapidly. Anyway, it appears the "68K add-in card" Knife mentioned may very well be in the works - that the PIOS ONE may well run the AmigaOS, and if so, it would do so on a separate processor, thus CONCURRENTLY with Mac System 7, the BeOS, POS, or the Linux port they're doing (which is a monolithic Linux port, btw, not a "microkernal" version like the existing MKLinux for Power Mac - so it'll be blazingly fast).

BeOS, POS, PIOS. LAVA, JAVA. BeBox, A\Box, AS/400. Lemme know if you see any possibilities for confusion. (Just threw that last one in for good measure.)

Two more Amiga web browsers are adding frames technology - IBrowse and Finale Development's forthcoming Finale Web Cruiser (not to be confused with Netcom NetCruiser). It's worth noting Finale is also writing a Java virtual machine for the Amiga - specifically a "just-in-time compiling" type, which is the same principle as a "dynamic compiling" processor emulator like Emplant 586 would have been - i. e. faster than the typical "translation" type Java systems. MOCA, as Finale is calling it, will incorporate into FWC, thus running Java applets inline on Web pages the "normal". MOCA will also make use of a PowerPC if you have it - for an even bigger speed increase.

I mentioned last time an idea for putting Sun's Java chips on PCMCIA cards; the big reason for suggesting this was that, after the failure of P'Jami and other Amiga Java ports, it looked like a hardware solution would be easier to come by than a software one. Anyone who's run Netscape 3 under Shapeshifter knows Java virtual machines running on the Amiga probably won't be all that impressive - which brings me to the other reason for suggesting it. I've had to clarify the design for a few people - the PCMCIA bus speed is NOT the limiting factor, because the Java execution takes place entirely on the card; information gets transmitted back and forth only when communication must take place, such as screen redraws, and between AGA and slow modems, watching a Java applet redraw at the speed of a hard drive won't be that bad at all - especially knowing how slow Java runs anyway, seeing screen redraws at AGA/PCMCIA speeds will be an improvement. In any case, those Java chips will run Java faster than a Pentium 150, and cost only $25 per chip. This is the important part - if you're concentrating on Java, a $100 card for a $500 machine makes an excellent investment. The trick then becomes making the rest of the Amiga keep up.

Anyway.

I've for some time been harping on how Apple won't go the way of Commodore because Apple has done what Commodore did not - fire and replace its defective bosses. Well, I heard someone mention a name I'd forgotten, and realized I was wrong: Mike Markkula, Apple's vice chairman.

Dave Winer of DaveNet fame has pointed out that Mike is the "invisible hand" at Apple - he decided when Scully and Spindler were no longer wanted, and will probably make a similar decision about Amelio someday. Markkula is the one Dave blames for Apple's consistent marketing failures. He's probably right.

Apple, in this light, will not get their act together until Markkula is gone.

Meanwhile, though, we've got the rumors that Adobe Photoshop is ALREADY "running" on a prototype of Rhapsody's emulation layer in Apple's labs. We've got Avadis Tevanian (who looks an awful lot like Steve Jobs, have you noticed?) talking about putting Rhapsody on Intel chips - it won't run current Mac software, but it WILL be a Mac OS-like experience on Intel systems. My own opinion of putting the MacOS - or any other OS born on custom hardware - on Intel is simply this: part of the reason the Macintosh works so well is because it doesn't have to run on Packard Bell or Compaq hardware. Remember that Packard Bell was the first company to apply to Apple for a license to make Mac clones back when Apple first opened the door - and were the first company to be turned down because of their lousy hardware. A Mac that has to deal with limited IRQs, COM port conflicts, serial mice, and the various Intel chip backwardsnesses, is going to be less of a Mac.

There's not really that much bad you can say about Intel chips anymore - the Pentium Pro, for one, is an impressive chip, with a whopping 256K L2 cache on a second chip die sitting inside the chip package, connected to the processor die via a wide bus; the PPro has bipolar circuitry added to its primary CMOS layers, again a unique approach matched only by Exponential's CMOS-on-bipolar trick. The problem with Intel chips is that the systems they go in are still based on a 1981 design - the IBM PC. It's been a gradual evolution over the last sixteen years, at each stage adding a new technology only where it can me made compatible with preexisting technology, until now it's fair to say software written for MS-DOS 1.0 will run on a Pentium Pro; PCI is neat, but is implemented so software thinks it's ISA, it still requires you to set IRQ's, and Plug-N-Play only works if EVERYTHING is Plug-N-Play enabled. The newest SVGA 128-bit 1800x1640 graphics cards still support CGA. The PC BIOS still lives at $FF00 - right in the middle of RAM. Evolution on the PC has been incremental, each stage of improvement must be so conservative to avoid breaking older stuff that people continue to code for the older stuff, until even now much of Windows 95 is written in 8088 code.

But anyway. Rhapsody - the MacOS/NeXT hybrid - is gonna look and feel like a Mac. It's gonna use Display Postscript as its imaging engine. It's probably gonna use Open Transport for networking. There's talk that the System 7 emulation is not a "layer" - but a virtual machine integrated right into the OS; perhaps they're gonna do something like the "retargeted OS" approach I've described on occasion, simply redirecting MacOS Toolbox calls to NeXT classes, so existing Mac apps run as if they were Rhapsody-native, with the enhanced UI and everything.

In the shorter term, the next MacOS release, "Tempo" - which WILL be called System 8 after all now that Copland's out of the picture - will include the multithreaded Finder, window-tabbing, spring-loaded folders, Option key menu, and slick look from the Copland screen shots last year. Oddly enough, the existing Tempo screen shots DON'T look as nice as Copland-appearance system hacks like Aaron and Kaleidescope; SmartScrolls and similar technology still hasn't been integrated into the OS; no word on whether Tempo goes the whole "Appearance Manager" route Copland would have gone, allowing the user interface to be modularized and replaced with the "techno" and "kiddie" themes we all saw in the screen shots.

Apple might not dump Newton after all. SOMETHING is gonna get axed March 14, but we're not yet sure what; Gil at least is proud of the Newton MessagePad 2000, the fastest handheld in the world, built around a 167MHz StrongARM processor. Of course, while we're on the subject of fastest, Apple's new PowerBook 3400 line - topping out at 240MHz - are the fastest laptops ever made. Expect their advertising to say something to that effect. (Advertising? What's that?)

But the rumor mill has one final nifty thing to pass along. Apparently part of the decision of which products to cancel lies with - hold the presses - Steve Jobs. His "part-time consulting" role is apparently a little more involved than we were led to believe. As one observer put it, "He's probably even got his own water cooler."

Anyway.

The big news the last couple weeks has been Quake for the Amiga. Keeping in mind that no major-title computer game has been ported to the Amiga since Maxis' SimCity 2000, this looked initially like something very, very cool - "wow! ID finally acknowledged us!" - until I did some research.

Stolen source code of Quake has apparently been floating around for some time. Apparently a similar Mac "port" has been built from these sources. The Amiga version is coded entirely in C, straight from the sources, no optimizations so it's sluggish on an '060, and has AGA support only as an afterthought - version 0.1 required Cybergraphics. It requires the datafile from the PC shareware or registered version of Quake.

AND IT IS NOT SANCTIONED BY ID SOFTWARE.

I like ID Software. They're sick-minded and nuts - witness the BFG-3000 from Doom II ("big fucking gun") and the Hitler Death-Cam from Wolfenstein. They're Amiga developers in spirit if not in practice - check out Doom OS if you think I'm kidding. It may be that this will inspire them to "recapture" the sources and distribute a sellable Amiga version of Quake after all - the program does exist, if only in a rough form, and is playable on an 060, some fine-tuning would at least make it playable on 030's. A Playstation the Amiga certainly isn't, but having at least one major PC game title ported would help us greatly - and if anyone can do it, ID could.

Problem is, until ID take some sort of action, the Quake port you've all downloaded is of questionable legality. Since it uses the shareware datafile, it's not technically "pirated"; it's written from stolen code, but the "author" makes no money from it, and in fact people are _buying_ the PC version for the registered datafile, in effect "registering" the Amiga version. But it's all very, very confusing and gray - much like the various "gray" Macintosh ROMs you see on the Web, "download ONLY if you already have a Mac ROM but no way to read it," it's just close enough to illegal that many people's consciences won't let them do it. Make your own call on this. In any case, getting the PC datafile is a major pain in the arse; to go through that much trouble to evaluate an illegal program just wasn't worth the effort for me at least, not when Trapped runs much faster and is self-contained and legal.

But there's an eerie flipside to all this.

You probably remember some time ago, an illegally made "demo" of Myst popped up on Aminet. Myst, of course, being the CD-ROM title of the year for Macs and PCs a year or so back - quite a sweet little piece of software. The demo, when it appeared, left a lot of people confused - is Broderbund actually gonna make Amiga software again? My buddies at Amiga Format mag in the UK got hold of Broderbund, and found out not only were they NOT interested in doing an Amiga port, but that they were seeking legal action against the people who coded that demo.

But that's not the sort of thing you tell Amiga Format. Ben Vost wrote up the story last year, and the feedback was so intense that Broderbund seemed to pause for a moment and reconsider. The world held its breath - or at least those of us with Commodore hardware on our desks.

Now, the Amiga software outfit ClickBoom has arranged with Broderbund to port and distribute Myst for high-spec Amigas. It's CD-ROM only (so I'm smiling), wants 4MB of RAM, and a high-color Amiga (either AGA or graphics board). No idea when the release date is - but if it costs closer to $30 than $50, I'm there. I've wanted this particular game for a long, long time.

So now there's something of a precedent. Amiga people have shown their determination - that if we want a particular major-label "other platforms only" title on our machines, we will get it, by brute force if necessary. No, these "major label" releases don't run on your A500 with one floppy drive - but they won't run on your C64 either. This is no longer about nostalgia - about keeping worthless hardware "functioning" in the 90's - it's about getting the software we want under the OS of our own choosing. And let the big boys clean up our messes when we're done.

(To those of you who've never upgraded your A500 past 1.2, all I'll say is this: I could just as easily have stayed with the Color Computer and saved almost a thousand dollars over the last four years. I chose instead to go the Amiga route - a decision I have never regretted. I could have stayed with the 500 and saved all sorts of money - but I chose to go A1200 instead, and while it cost me lots of money to bring that 1200 to where it is today, I won't complain. Much of my life currently involves doing things on computers that just can't be done on lesser hardware. For me an AGA Amiga with an 030 and 6 megs of RAM and some sort of hard drive is an absolute baseline. It's certainly not Broderbund's fault that so many of you are running 2MB A1200s with no hard drive - with hard drive prices what they are these days, you're running out of excuses. Upgrade.)

There's something of a downside to this. Myst, when it arrives, won't appear on the shelves at Best Buy. Nor will magazine advertisements for Myst - at least stateside - be likely to carry the line "Amiga version available now." The box when you see it on Amiga dealer shelves will probably have ClickBoom in bigger letters than Broderbund - because they did most of the legwork. If the game doesn't install properly, don't expect any support from Broderbund. It's a third-party port of a Broderbund product, and you can probably count on Broderbund promoting it as such, regardless of licensing arrangements.

But there's a weird sort of fallout I've noticed in all this.

Labyrinth of Time was there first with the graphical, raytraced CD-ROM adventure game. But until Myst got there, with Broderbund's name under it, the PC market didn't quite take it seriously. Look at Sinkha for the Mac - nothing like it exists for the PC. For that matter, look at Dramatica and assorted other Mac software for which no PC equivalent exists. The day Microsoft writes a Dramatica-like extension for Word, though, it'll be hailed as the greatest technological advancement for fiction writing since Mark Twain became the first writer to submit a manuscript written on a typewriter. The PC market has become so top-heavy that a product or concept is ignored until one of the Big Boys picks it up - Netscape and Wolfenstein are notable exceptions, but through sheer luck wound up catching a degree of mindshare. Now they're Big Boys and can get away with things smaller companies cannot. (Ever used Oracle's PowerBrowser? Ever HEARD of it? Didn't think so.)

The thing to consider, though, is this. Let's say you come up with a totally unique concept in software. For example, you're making a multimedia CD-ROM that looks and plays different from anything else out there. You don't work for Blue Byte or Psygnosis, it's just you with a multimedia authoring package, a copy of Lightwave, and a CD-ROM burner. You master the sucker for Windows 95, put it in a snazzy box, and get it into the distribution channels - and it flounders. Nobody notices it on the shelf at Best Buy among hundreds and hundreds of other titles - and with no name brand, and no magazine reviews, no one cares.

So you try again. You remaster it, this time for Amigas with AGA and 4MB of RAM. It goes on the local Amiga dealer's game shelf right between Qwak and Cyberpunks. Then you upload a 1MB demo to Aminet. No one's heard of you, but it's new, it looks nifty, and that demo was absolutely spiffy. EVERYONE saw the entry in the "recent uploads to Aminet" - which means everyone will download it and try it, they all agree it's throughly cool, and if indeed the product is good enough, you've suddenly got an Amiga Format GOLD sticker, your demo is the CU Amiga coverdisk (even better than front page news), you're in all the reviews, of course being on Aminet is a form of immortality itself anyway, you're in Amiga Report, your homepage goes on Amiga Web Directory, and your cool game is the talk of the town.

And you've sold more copies than you ever would have, had you gone PC.

There exists a compromise, that being the Macintosh market, which is a bit more developed than the Amiga market, but not yet so big and political that a small company with a decent product cannot get a foothold. But Macs are a little trickier to code for - especially since "hitting the hardware" is a bigger nono on Macs than on Amigas, between the various motherboard designs, the hardware is often so different even certain OS revisions don't run on certain Macs. But that's almost nothing. Uploading your demo to InfoMac and getting listed on ChezMark will earn you quite a bit of fame, admittedly - at least within the Mac community. Products like Gryphon Bricks, which would never survive in the PC marketplace, are everywhere on the Mac side. And in any case, the latest round of Macs and clones are the fastest personal computers ever made. Power Macs have lots of CPU juice to spare, so at the very least, if the Amiga can't technically support your product's CPU-heavy algorithms, a Power Mac will be more than happy to take up the slack.

But if you can make it work on an Amiga, there are actually viable reasons for a small company or small developer to do Amiga titles.

I for one am already considering a multimedia title like what I described above. That movie project I've mentioned is shaping up to be a little too render-intensive to do as a movie, but won't survive being flattened into a novel or comic. The remaining option I had was to construct a CD-ROM-based multimedia novel from it, and such a product would get lost in Best Buy between Command and Conquer and the latest Microsoft Arcade compilation. But an Aminet demo, a snazzy box on those cozy Amiga dealer shelves, and a $19.95 price tag, not to mention online direct ordering from a Web page (which obviously will get word around, I mean, you're here reading this, right?) and I should be grinning. A Mac port would thus be right around the corner - Supercard and Quicktime make such things kinda easy. But the Amiga version would be able to make use of preemptive multitasking, Anim5 and Anim7 formats, HAM8, and of course, Paula's exceptional bass range. If nothing else, anims that had to move at high speed could be built as SSA, and I could then see about redistributing the SSA Player; those anims are somewhat RAM-hungry, but an A1200 with 6MB and a hard drive makes a decent baseline, no? That, or use the new YAFA format. Either will deliver 30FPS full-screen where Quicktime delivers only about 10FPS quarter-screen. And a CD-ROM can hold LOTS of anims.

While I'm on the subject of that project, I need to voice some gripes about the state of 3-D software. Imagine has some impressive features, but every day I'm finding more reasons to hate it. (It's better than POV for what I'm doing, but the moment someone builds a decent modeler and spline-path animator for POV, I'm switching.) Its user interface is painfully inconsistent and unintuitive. There are some impressive technologies under the hood, but grafting those features into a late-1980's user interface has proven less than optimal. The system is very sluggish at redraws, often performing a redraw when none is needed. It lacks a mechanism for precisely inputting coordinates - something I was spoiled by with AutoCAD - but yet its mouse control is not particularly confident. "Lock" mode works only when adding points, not moving them. There is no "lock" equivalent for rotations - unless you go to the transform requestor. If you've rotated the axis of an object, various operations use the new alignment while others use the "world" axis - meaning your extrude may or may not go in the expected direction. Certain features don't work as expected (Slice) or at all (Ripples). Bones are powerful but nightmarish to set up - and are highly dependent on the sequence of points within the object, something over which the user has no direct control. Certain menu options (notably various bones automation tools) weren't clearly thought out before being added - asking you to click on an axis when in "pick points" mode, for instance.

Imagine has long since outgrown its interface. Numerous features (in Imagine 4 anyway) are still "works in progress." The Spline editor is nothing more than a font editor. The Action editor should, by all rights, have been incorporated into the Stage editor. Some editors require you to "save changes" while others don't. The Detail editor lets you pick groups and subgroups while the Stage editor doesn't - which means that to make objects properly move in relation to each other, you must stage them twice, once in Detail and once in Stage, and then save the various states. Once bones subgroups are made, you cannot tell which faces are grouped and which aren't - and if you double-group or fail to group a face, you could be in for some weird results. Slice doesn't work if face edges are too close together - but isn't smart enough to just merge the points if that happens. Often-used items are buried deep in panels beneath submenus. There are glitches in the brush-map function - that cause part of the brush to be "striped" across the object. So many cool things have gone in under the hood, but no one at Impulse seems to have considered those of us who may want to use these features in production work - the package gets more and more difficult to use with every revision. It's like a Microsoft product - you have to put up with the interface problems and bugs because there's still this nagging notion that it can do what you want, if you swim upstream long enough.

Admittedly much of my frustration with Imagine comes from the very problem of expressing 3-D topology with 2-D input devices. It's not an easy problem to solve. One solution I'd like to see is a generic 3-D input library for use in various 3D software, one that allows you to choose a 3D interface you're familiar with. A possibility there is that future 3D input hardware could be supported with a simple device driver. But that brings me to the next problem.

3-D input devices are still sickeningly expensive. There's the radio-driven "pens" that cost $3000 and are alignment-sensitive (read differently if you hold it horizontally or vertically). There's the CMMs you find in machine shops, the crane-like apparatus used to measure X, Y, and Z coordinates on a stationary object under it - those cost tens of thousands of dollars and require dedicated computer hardware to drive them. There's laser-driven "digitizers" - like what's used to digitize actors' faces for special effects in movies - which cost in the six figure range. Even the "simplest" kind of 3D digitizer - a pointer on a hinged, movable arm - costs $1400.

This is ludicrous. A 3D digitizing rig is more a mechanical problem than an electronic one. As it happens, positional encoders are cheap and plentiful to come by - there are two in the mouse you're using right now, one each for the X and Y rollers. Three are all that are required for a 3D rig - one per axis. The $20 "disposables" you see on the wall at Wal-Mart will deliver 560 DPI - that is, 560 divisions per inch. You could gear the rollers down if you wanted, for even more precision. A CMM-like device could be built from metal or plastic frames, a couple straightedges, components from scrapped mice, and a microcontroller, all for under $350. Just add a microswitch to the "probe" - program the controller to transmit data back to the computer down a serial or parallel interface as sets of coordinates, have a driver on the computer sitting and waiting, and we're all happy.

Alternately, imagine a 3-D mouse attached to an "arm" bolted to your desk. Those discarded mouse encoders could be arranged physically to record the angle at a given hinge, and just articulate this "arm" properly. The receiving computer - or even the microcontroller - can do all the calculations needed to turn the various angles and distances into 3-D coordinates. Such a device could sell for $150. Descent will never be the same again.

In any case, the whole problem is that I'm trying to construct a human face in Imagine. Imagine is great for building space hardware, but organic shapes like faces stymie it. Its spline editor is useless for this. The Forms editor can't quite manage things like noses. The Detail editor is precisely that, details: to build a human face in the Detail editor, you do so point by point, and with no decent way to input 3-D coordinates, that means laying them out in 2-D and then moving them along the Y axis to where they belong. Forget about smooth curves - even with phong shading on, if you have so much as one point out of place your human looks 20 years older, not to mention the lack of decent control over the overall shape of an object means you can't actually make it all look properly human. I don't know that any other 3D package would have a better time with it - certainly Lightwave, with its lack of hierarchic motion, could do the face but would have a miserable time animating it. Real3D would be my preferred choice, except it's not exactly standard, and since it's a full CSG mathematical-surface raytracing engine (like POV on steroids) it's CPU-intensive (read: slow). It does, however, generate the most photorealistic renders I've ever seen, it has a physics model and collision detection, and has a true hierarchic object system, like Imagine except better. But I ain't got no $300 for it.

My hope at any rate is to at least get Anime-looking faces for the project - the whole story would make great Anime anyway, and computer generated it will be cool. (Think the intro to Tekken 2.) But Imagine isn't even smart enough to do that much. Anime heads are no different from realistic heads where Imagine's detail editor is concerned - it's not one of the primitives, so forget it.

I was hoping to get to the Gateway Amiga Show next weekend, but that's not likely, not while my car has a defective front end and four tires on the virge of calling it quits. It's lucky to survive back and forth between work and home; it will not survive a trip to St. Louis. To fix the front end won't do much, the tires are still borderline, and a new set of tires would exceed a month's rent. The car is going nowhere.

I keep promising I'll be revamping this site soon. A message board, for instance, would have some uses here - especially since USENET sucks. Buyout Watch is so far behind it's worthless. The Ideas section is soon to begin growing by leaps and bounds. And I'm considering a "greatest hits" section for things like guest contributors, USENET highlights, and the like. But right now my head is in a mucus-induced haze, and I'm in no condition to write message boards or redesign homepages until this blasted head cold goes away. So... we'll just have to wait till next time for the fun stuff.