Peace in the struggle to find peace

May 17, 1997

I can still remember clearly hearing the news that Commodore had crumbled. It was the Day of Infamy, April 29, 1994. I was, as luck would have it, at the keyboard of my Amiga, and off in the distance I heard a television in the other room. It was WISH-8 Indianapolis, they'd popped a brief news update in on a commercial break, and I heard Mike Ahern's voice: "...Commodore International is going out of business" - and then I broke the sound barrier from the bedroom to the living room to hear the details.

I can still remember the day of the Great Auction of April 1995. I spent much of the two days on the Net, watching the action from USENET and IRC, watching, praying, cheering and crying. The two favorites, Creative Equipment and Commodore UK, both seemed to disappear early on in the proceedings - CEI joined PC giant Dell's bid, and Commodore UK just went home. Dark horse ESCOM appeared from nowhere, plunked down a sickeningly tiny amount of money, and walked away with nifty things. I didn't like the idea of a PC company getting the Amiga - and said so on USENET in a decidedly colorful message entitled "I have had enough of this shit!" that comp.sys.amiga.misc regulars probably remember (with a cringe) to this day. ESCOM started doing good things, though, picking up such Amiga notables as Peter Kittel and Olaf Barthel, and their best move was spinning the Amiga off into its own division. But that didn't last.

I remember learning that Gateway was getting the Amiga. It was the last week of March, and one of my contacts sent e-mail to a giant makeshift mailing list of Amiga "celebrities" he keeps in touch with (hope I'm not violating any confidence here, but that list includes everyone from local Amiga users group presidents to Eric Schwartz and Jason Compton, along with nearly EVERY major Amiga developer - how honored I was to see my own name listed between Schwartz, Eric and Stuntz, Stefan), saying "here's the scoop, there's gonna be an announcement next Tuesday, try to keep this under your hats till then." Since next Tuesday was going to be April 1, I emailed back asking if this was in fact an April Fools joke that got started early. But then the news items started appearing on Yahoo, CNet, and even Gateway's homepage. It was official. The only question by the end of that day, which I asked on the celebrity list, was do we start affixing Boing balls and checkmarks to every Gateway we find, or do we paint cow spots on our Amigas? And the next day, everyone on that celebrity e-mail list found an attachment in their mailbox - Eric Schwartz's answer to that exact question, the now-famous cartoon of an Amiga in cow spots saying "Don't worry - this is good news, they tell me." Of course, all the Gateway news REALLY meant was that it would be another endless, breathless wait for "German regulatory approval." So we waited. And we waited. And we waited. Our Amigas slowly sprouted cow spots, and our neighbors' Gateway PCs slowly sprouted Boing balls and checkmarks - and still we waited.

As late as yesterday afternoon, I had told someone that if it turned out Hembach, the lead ESCOM bankruptcy trustee, was in any way involved in the regulatory approval, that I would have to switch platforms because the regulatory approval would never happen. You see, once a company gets sold, the bankruptcy trustees stop making money. This is why it took a year to sell Commodore. This is why QuikPak got screwed over so bad. This is why VIScorp will probably go out of business before the year's out. Hembach knew if he sold the Amiga, his money would stop. If any part of the regulatory approval process involved waiting on Hembach to sign a document, it would probably remain unsigned on his desk for months or years, or at least until the document crumbles into dust and is swept away by the janitors.

Apparently this was not the case. Sometime yesterday afternoon, the news we've all been waiting for finally arrived: the regulatory approval went through. It's over - our moment in the sun has arrived. The bankruptcy is over. The waiting is over. The days of VIScorp and Hembach and all the wonderful things they bring - all that is over. The Amiga has a home again. We're alive. We exist again.

But is it too late?

Let's go visit the competition for a moment. You know, of course, one of the most-ventured opinions of why Gateway bought the Amiga was so they wouldn't have to fuck with Windows CE - or with WebTV. Everyone in the computer industry considers Microsoft competition - or will eventually.

I suppose it's time to ask the specific question: has Microsoft won the war?

Ticketmaster - you've heard this by now, probably - is suing Microsoft. Does anyone else find this amusing? "Giants in the sandbox," I suppose is the only appropriate description - one monopoly deserves another.

But that isn't all of it. Ticketmaster is suing because Microsoft is linking to their Web site.

Exqueeze me?

Yes. Microsoft has links to Ticketmaster's Web-based ticket ordering database from somewhere on Microsoft's site. I saw the pages in question, but have no idea where - and am in no mood to go look for it, if it's like the rest of microsoft.com I'll never find it again. Ticketmaster's complaint is that the links go deep inside their system - bypassing a lot of things Ticketmaster doesn't want bypassed - and that Microsoft has built a number of pages around this, featuring Ticketmaster's logo and the whole nine yards. Admittedly the pages in question are little more than Ticketmaster pages with Microsoft fonts and colors. But there's a bigger issue.

The complaint the rest of us have is, is it now illegal to link to someone's Web site without permission?

It's one thing to go after someone who's copied the entire text of a Web page and claimed it as their own. It's kinda gray when someone copies a section of an article without permission, but attributes it properly (as I try to do when swiping press releases). But just adding a LINK to that information instead - according to Ticketmaster that should be illegal.

In my own opinion, it's just politics - Paul Allen, cofounder of Microsoft, owns part of Ticketmaster. He did not leave Microsoft with a smile on his face, as I understand - Bill Gates swindled him into signing over the company to him - and I think this is just his way of tweaking Bill's nose. Only Ticketmaster is big enough to piss off Microsoft and survive it.

Netscape pissed off Microsoft and is paying the price - they're losing browser share so quickly they're considering dropping Navigator altogether and just making Web server software.

WebTV pissed off Microsoft - by having a product they couldn't touch - and has now been assimilated.

Apple hasn't pissed off Microsoft in a long time - Bill has a soft spot for Apple in his heart, and Apple knows exactly how far they can push him - and don't dare go near that line. A few Apple people are fond of pushing that line - sort of a token way to elicit smiles from diehard Mac users - but know where and when they can get away with it. They don't dare openly piss off the Mac's biggest developer.

Sun is currently in the process of pissing off Microsoft - and it is of Sun that I wish to speak.

Sun has become something of a flag-bearer for UNIX. Once upon a time, the "official" UNIX (AT&T System V) was the flag-bearer - but for whatever reason that's fallen by the wayside. (Do you even know who owns the "real" UNIX now? Novell? Santa Cruz? Who knows.) When AT&T System V Release 4 came out it was a BIG deal, almost as big a deal as Windows NT 4.0 has been. (By the way, Commodore UNIX was the first version of AT&T SVR4 to actually ship.) SVR5 came and went several years back - I don't even know if they're still working on System V at all.

At any rate, of the various UNIX vendors out there - a list that includes Silicon Graphics (IRIX), Hewlett-Packard (HP-UX), IBM (AIX), Apple (A/UX, NeXTSTEP), Berkeley (BSD/OS), and the various Linux distributors (Caldera, Slackware, Red Hat) - it seems Sun's Solaris is the most visible.

I'm not entirely clear on why. BSD has for me long been the "real" UNIX - it runs, runs well, and runs on everything. It's standard - Solaris is a bastard child, part BSD, part System V, with just enough uniquely Solaris thrown in to make it a problem. BSD is more widely supported on Intel chips than Solaris - and more widely used.

I think Solaris' big deal is that Sun actually makes the hardware specifically for the UNIX they ship. Of course, this is not an immutable scenario - for years, Bill Gates' own desktop computer was a Sun SPARCstation running Windows NT. But even as custom UNIX boxes, Sun boxes aren't the greatest - they cost a decimal place more than comparably powered Intel hardware, are only somewhat standard, and honestly, aren't that fast. $20,000 will get you a 167MHz UltraSparc system with a 21-inch monitor and 16MB RAM. $5000 will get you a 166MHz Pentium with a 21-inch monitor and 64MB RAM. The Pentium will mop the floor with the UltraSparc - even running Solaris.

And then there's Solaris itself - the native GUI sucks. If you thought the BeOS was ugly... oh well, there's hope, you can run OPENSTEP on Solaris and get the NeXT's beautiful GUI instead of that nasty and weird Solaris look. Of course, GUIs are also a matter of taste - it may very well be that there are people out there who like Solaris' look, just like there are people out there who like Windows 3.1. The fact that Solaris has the scrollbar arrows attached to the scroller itself - a nonproportional one at that - and your mouse pointer gets pulled up the screen with it as you hold down the scroll arrows! - doesn't have that much to do with the underlying OS.

(Did they really ship Commodore UNIX with the Solaris/OpenLook GUI? Ugh.)

But anyway. Sun is slipping in the workstation market, and in the server market. They make decent OSes and software. They have some cool ideas and some really sharp programmers. And a few years ago, they came up with this concept we now know as Java.

Java was, and is, and will probably remain for the foreseeable future, a solution waiting for a problem. They came up with the idea of CPU-neutral "bytecode" initially for embedded applications - and then realized it would be a way to deliver platform-independant executables across the Internet. Then they decided Java could be its own browser and its own OS - and came up with the idea of the Network Computer, the $500 browser box that runs Java applets and hooks to a TV.

Now, Sun is trying to bill Java as a Windows-killer.

Guys, Java and Windows are as much alike as breakfast cereal and Doritos. They do overlap somewhat in features, but are not enough alike to directly compete - they are not even the same class of product, are aimed at different markets, and are NOT ALIKE. What Sun wants Java to be is a platform, an OS - but to do that, they have to MAKE it an OS. Preferably one with a decent GUI and windowing toolkit - that's why people write applications for Windows and not for UNIX, the windowing toolkit is more robust. Java's AWT (Abstract Windowing Toolkit) sucks rocks - and more often than not, fails to work at all. You can write Java applications (not applets) that make use of the host OS's native GUI toolbox - but then they cease to be portable, thus defeating the purpose of writing it in Java in the first place.

The only way Java will take off is if Sun ships those Java processors and they run as fast as they said. A $20 chip that runs Java faster than a Pentium, coupled with the killer Java applications everyone's been predicting, would be a useful thing. Admittedly Corel's writing its office suite for Java, but whether it goes anywhere depends largely on whether you have a Java system that will run it decently fast - and the Intel version already exists and is fast.

But that presents the original problem again: Java is a solution lacking a problem, a cool-looking wrench that fits no nut known to humankind. It's an old habit computer people have, of making the solution and then being saddled with the daunting task of finding the problem to fit it. The Amiga was in a sense invented this way - but it's a paintbrush, not a wrench, the Amiga was designed to be flexible, to solve new problems as they appear, to create new ways to work, and then allow you to design solutions to the new problems created by those brave new worlds only an Amiga-like machine could let you enter. Java, on the other hand, is like some whimsically-shaped tool found at a yard sale: it looks cooler than shit, you just HAVE to have it, although you have no clue what the damn thing does - and will never find a use for it.

Sun knows Solaris won't beat Windows anymore - not on a technical level, not on an efficiency level, not on a market share level, they've lost all three to Windows NT. So now they're trying to bill Java as an answer to Windows. And in case you hadn't noticed, it isn't working.

But watch carefully how Microsoft is reacting to Java.

When Bill Gates decides to go after someone or something, he has a tendency to blitzkrieg them first - firing lots of missiles and bombs in their general direction, softening them up before rolling the Microsoft tank over the top of them. He did this with Netscape - a few well-placed missiles, and finally Internet Explorer 3 and 4, free for the download, the tank that's slowly flattening Netscape.

But there's more to it than just the blitzkrieg. Bill's modus operandi is very chesslike - to set you up, arrange all the pieces, slowly moving pieces you won't notice until it's too late, and just basically getting the game to where it plays itself, and letting you make all the mistakes. He has two things going for him: he doesn't often make mistakes, and he recovers quickly from the ones he does make. By mistakes I don't mean technologically - I could recite a list of Microsoft technological mistakes as long as your arm - I mean businesswise. In a very real sense he's the one who conquered IBM - by leasing PC-DOS instead of selling it to IBM, and then selling it on the shelf as MS-DOS to all the clone makers who wanted it. Similar trickery with Apple - getting on the ins with potential competitors, finding out what they're up to with promises to help them, and then suckering them at the last minute. Gates has made VERY few business mistakes - I cannot currently name any recent ones. He's not an intelligent businessman - just a very sneaky one. He's shown an immense talent for playing the computer industry like a chessboard.

Or like a game of Command and Conquer.

A story I heard, that I found fascinating, is that long about the time Internet Explorer 2 was released, at one of the big trade shows there was a panel discussion, they had Gates there, Marc Andreessen of Netscape, one of the people from Sun, and a few others. And Marc kept mouthing off about how Microsoft was just trying to put everyone else out of business. Bill kept saying "no, we're just trying to make a browser, we don't wanna put anybody out of business." Marc wouldn't shut up, and finally Bill said "All right, if that's how you want it, Microsoft won't rest until Netscape is out of business." Not for a moment do I believe Microsoft ever intended to be altruistic - but by waiting for Marc to say the wrong thing, Bill Gates has the piece of ammunition he needed, the ability once Netscape folds to say "they brought it on themselves."

So now there's Bill Gates complaining about Java, its lack of a GUI toolkit that works, Sun's arrogant licensing arrangements, on and on.

He's painting Sun as arrogant and self-focused, so when he rolls over them he can say it was their own fault.

After watching so many companies fall before Microsoft, not because they have inferior products, not even because they are poorly managed, but because they're arrogant and walk right into the Gates trap every time, I have to wonder why. Is it the drinking water in Silicon Valley? Something's causing all these companies to end up with slightly off-center leadership that likes walking into traps. It's not that Gates is smart, it's that he knows how to shoot at missing scales on dragons as they fly overhead. (Downright prophetic sometimes, that Tolkien.)

I asked if Microsoft had won the war. My own answer is, "Which war?"

At the Microsoft campus in Redmond, there's a courtyard, each stone on the walkway is the box of a Microsoft program. (The first stone, oddly enough, is Microsoft OS/2.) In that courtyard you will find EVERY Microsoft program ever shipped - I wonder if that includes Color Disk EDTASM. Or XENIX for the TRS-80 Model 16. But there's a plaque in that courtyard, which has Microsoft's mission statement: to have Microsoft software running on every desktop in the world.

At the house I have two computers hooked up and operational - the A1200, and a Color Computer 3. The CoCo's BASIC ROMs were written by Microsoft (and Microware Systems). On the Amiga I still have a copy of AmigaBASIC (which doesn't run on the '030 but I haven't bothered to delete it) - and under Shapeshifter, since Netscape 2 is buggy and Netscape 3 is huge, I use Internet Explorer 3 for the Macintosh.

At work, Internet Explorer is on every Macintosh here - along with Word, Excel, or both.

The ONLY computer I ever use, in fact, that doesn't have some kind of Microsoft software on it, is the UNIX box on which users.compassworks.com resides. And eventually that may end up being replaced by Windows NT.

Microsoft still gets royalties for every copy of OS/2 sold.

Microsoft owns more patents than you realize - and many OSes license stuff from them. (It works in reverse far more often, though - there's a copyright screen somewhere in Windows 95 with names like Apple and Commodore listed.)

In other words, I would estimate that 99.9% of all desktop computers currently in use have something on them from Microsoft. The other 0.1% would be the likes of the Atari 8-bits, machines Microsoft NEVER developed for, or the various UNIXes, most of which you wouldn't count as "desktop" anyway. (The days of Microsoft Word for XENIX are long gone.) And actually, Microsoft is porting Internet Explorer to UNIX. (Why? To compete with Netscape. Have you ever played Command & Conquer?)

If that is the war, then Microsoft has long since won - it's now just a matter of getting the other players off the field.

If the war is that every desktop uses Windows, that war is not entirely winnable - there will ALWAYS be someone with an Apple IIc who refuses to get a Wintel machine, and thus will never run a Microsoft OS. But Microsoft already owns about 90% of the OS market.

Something to consider: Microsoft has a specific definition of what a computer is and looks like.

Microsoft, in the last ten years, has settled on the $1500-to-$2500 price range as "personal computers." By writing their OSes and software for that price range, they've managed to convince the entire industry and all the customers of it that nothing less than $1500 is a real computer.

This is bullshit.

I've written novels on a 2MHz Tandy Color Computer 3 with 512K and two floppy drives. (Started most of those at 128K and one floppy.) Total cost, 1997 dollars, if you find one in good condition at a yard sale, about $100. The biggest reason I switched platforms wasn't entirely the lack of power, it was the question of where to get it fixed if it broke, and of course, the need for faster printing speed so I could print those novels. But the point is, the computers we used in 1987 still work, they just don't run Windows 95. (Yes, that CoCo uses Microsoft BASIC ROMs, but Microsoft didn't write the word processor I used.)

Another thing to consider: Microsoft, and much of its customer base, has this delusional belief that everyone, deep down inside, WANTS to run Windows. Invariably the biggest thing wrong with the BeBox or Java or any other not-Intel platform, to ask these people, is that it doesn't run Windows. Microsoft proceeds from this arrogant notion, that everyone WANTS their software. Their advertising seems not-quite-subliminal, as though they're just trying to subtly "remind" you that your mind already WANTS to use their software. (Those ads would certainly not sell anyone on a product by themselves...)

Consider for a moment that the world is full of these simple, handheld, battery-powered or solar-powered computers with LCD screens, that you can buy for a dollar at the corner drugstore. A wonderful vision of the future, no? Well, you probably already own several of those computers: they're called pocket calculators. Simple, disposable computers everyone can use. And NOT ONE calculator on the planet runs Windows. The calculator industry is probably bigger than the desktop OS industry in total unit sales - but does Microsoft code for it? No. Would we want them to? No. Who would buy a calculator that cost $375 because it required a hard drive and 16MB RAM to hold the OS? Why would a calculator need an OS anyway?

Why do people buy calculators? Do the people who make the simple credit-card calculators have to have hundreds of tech support lines open for the people who can't make it say something besides "0."? Hell no - if they did, people wouldn't buy them. Programmable calculators are a different story - but they're specific tools for a specific purpose, cost a lot more, and you don't need an 87-key scientific calculator to balance your checkbook. (If you do, I don't even want to imagine what your checkbook must look like.) Many of these dollar calculators use "real" processors - often surface-mounted things like the Motorola 68HC05 or HC11, NEC V20/V25, various oddballs from Hitachi and Texas Instruments and ARM, and maybe even the occasional 68EC020. (The Intel 4004, the first commercial microprocessor and the grandfather of the X86 line, was developed for an early portable calculator.) They are computers - but people who wouldn't buy "real" computers, buy calculators anyway. Why?

Because calculators don't fit people's definitions of a computer. Calculators, by not being computers, aren't subject to the same expectations we have of a $1500 desktop. We don't expect our $1 solar credit card calculator to run Windows 95 - we expect it to add 2 and 2 and get 4 every time (unlike the Windows 3.1 calculator), but that's it. Anything it does above and beyond that - memory, for instance - is just added benefit, and doesn't interfere with your work if you DON'T know what it is.

Likewise the Playstation and N64, both with 90MHz MIPS processors and custom 3D chipsets that are probably faster and better than the $1500 desktop PC, cost $150 but nobody thinks anything about it. Why? Because by virtue of their being game machines, our expectations are changed. The Playstation does not and will never run Windows 95 - but is selling like gangbusters anyway. Why? It doesn't run Windows 95... but it DOES run Tekken 2. The N64 doesn't run Microsoft Word, but it DOES run Super Mario World. Ask any 12-year-old which is more important in life, term papers or Shadows of the Empire.

Computers don't need to cost $1500 to be usable. Computers don't need to cost $500 to be usable. Computers don't need to be $100 to be usable. You just need to define what it is you'll be using the machine for, and shop accordingly. The computer industry isn't geared toward that right now, however, billing the Intel-based $1500 PC as the solution to everything. Microsoft and every salesperson on the planet has forgotten the old axiom: "one size fits all" fits no one.

I take that back, if you're shaped exactly average, "one size fits all" would fit great.

For some people, and some applications, Windows/Microsoft solutions really are ideal. But at any rate, the ONLY situation in which Microsoft Windows is absolutely positively the only possible solution at all, is if one of your requirements is that every program your company runs must be made by Microsoft. For every other scenario on the planet, to varying degrees there are non-Microsoft solutions; in many of those, yes, Microsoft does have the best solution, but in none of them is Microsoft the ONLY solution.

So here's the conclusions we can draw. Microsoft does not like cheap computers - they've grown comfortable with their OSes on big machines, fitting Windows on cheap computers is no longer possible. Computers can be VERY cheap. Computers can be VERY easy to use. Computers can be VERY reliable. They aren't - because they must fit the expectations people have come to have of $1500 Wintel boxes.

Know any computers that break the mold?

Thought so.

Now, keep in mind we're not there yet. Not while A1200s cost $500 without hard drive, and A4000s cost $2200 with 6MB RAM and a hard drive. But the A1200, in its better days, WITHOUT any cost-reduction engineering, sold for $340. This was 1994. If the rumors are true, and Amiga Technologies really DID put AGA on a single CMOS chip before the ESCOM collapse, you could probably assemble a cheap, embedded AGA Amiga for $200, with a cheap '030, 4MB RAM, A1200-style unbuffered IDE port, connector for an external keyboard, etc. Ideally, one could start making '030-based or '040-based Amiga motherboards for a few hundred bucks that fit in standard PC cases, and get workable systems out the door for $600 with hard drive. Obviously this wouldn't be a permanent solution - those prices would have to come down to $400 and ship with '060s by the end of 1997. But you get the idea - and the only problem is, if it ships in a PC case and hooks to PC monitors and keyboards, people will expect it to work like a PC, and while they'll be amazed how well it does some things, the first time they see a sluggish AGA 256-color screen redraw they'll laugh.

For this reason, the Walker case was perhaps not such a bad idea. It was loaded with flaws - from the not-quite-Amiga motherboard to the vertical stacking expansion system that would cause timing problems and would look like an erect penis when you were finished - and the case itself was dumb-looking to begin with - but it's a step in the direction the Amiga needs to go. The less a shipping Amiga looks like a PC clone, the less people will expect it to have in common with a PC clone - at least until we get big, bad-ass systems again that outdo PCs at everything no matter what the case looks like. Such wacky cases have to be cheap and distinctive - and either have well-thought-out expansion or none at all. One A1200-style slot, or a complete Zorro-III or PCI backplane, or nothing. But it has to be non-PC-like. Cases made like triangles, or cones, or ellipses, or circles, or spheres - I'm partial to a 12-inch-diameter red-and-white-plastic Boing ball with drives in a foldout front panel myself. Have a contest to design the wackiest cases - make a "standard" Amiga motherboard shape and build maybe five or six different oddball case designs to fit it.

I like it. And given Gateway's preoccupation with outlandish designs and outlandish advertisement, they might just go for it.

Anyway.

The other direction to take is at the high end - you either undercut the $1500 mark by a mile, or you ship machines that are so big and powerful no $1500 box could ever touch them. There's a high-end systems market that Microsoft wants but hasn't conquered yet - mostly because Windows NT eats up too much resources and tends to bring $10,000 workstations down to the performance level of $1500 ones. While I mentioned last time the possibility of an Alpha-based Amiga, the PowerPC is already here, and by the middle of June the various PPC boards for the Amiga should finally ship.

PowerPC software has finally started to show up on Aminet. This is not the big deal people make of it - Mac, Windows/DOS, Atari ST, and UNIX software has been on Aminet for years, things like file exchangers, LHA archivers, ROM grabbers, Mac printer drivers that don't require AppleTalk, and the like. But it does somewhat signify that the Power Amiga has achieved a sort of critical mass. The software in question is a PPC blitter library for Cloanto's Personal Paint. Nothing earth-shattering. And actually, the FIRST PowerAmiga software to appear on Aminet was a PPC emulator-disassembler about 2 days before the Cloanto library. The idea is, you'd use the emulator kit to fine-tune PPC optimizations for your Amiga software without having to get the board.

Phase 5 has added two more designs to their PPC board lineup. They've got a souped-up version of their 1200 board, that instead of an '030 socket, it has an '040/'060 socket. (There is no way in hell you could possibly fit such a board inside an A1200 case and close it. These MUST be for tower units.) Cost, without '060: $600 USD. SIMM socket, PPCLoadSeg, the whole nine yards, 150MHz 603e, 50Mhz 68060... the perfect accelerator, has two of the coolest microprocessors ever invented, one at each end of the board. Now if only the software was more advanced.

And the other new board? They said it couldn't be done, they said nobody could do it: a 200MHz 604e board for the A2000. Yes, that's right - a RISC card for a ten-year-old machine. One with a 16-bit motherboard at that. This makes it official - the A2000 is the most versatile, most expandable personal computer ever invented. You can, quite literally, buy an old A2000 still in the box, sealed since 1987, pull that Kickstart 1.2 ROM and put a 3.1 in its place, and plug a PowerUP board into the CPU slot - and you literally make the leap every Power Mac makes figuratively: jumping from a 68000 to a PowerPC. No one has ever made a PPC accelerator for a 68000-based Mac - the smallest Mac that'll host a PPC board is an early-90's 68030 board. You CANNOT put a PowerPC in a Mac SE. But you can put one in an A2000 from the same year.

Then I got to thinking. Ten years ago, a fully tricked-out Macintosh II with a color card, monitor, hard drive, high-density floppy drive, and a couple megs of RAM, might run you almost $10,000. Today, if you spend $10,000 on a Macintosh, it has a 240MHz processor, 32MB RAM, 2.1GB hard drive, an inch-thick active matrix LCD panel, a designer case, a subwoofer, and Apple sends a guy out to your house to set the damn thing up for you.

So here's my wacky thought. If they can make a PPC board for the A2000, they should make one for the A1000.

Hear me out. THEN laugh at me.

Make a very limited number of these boards - say, three hundred - and sell them as part of some big "commemorative edition" package for $3000 a pop, individually numbered, autographed by a whole bunch of Amiga celebrities not unlike the inside of the A1000 case or the back of the Deathbed Vigil shirt. The board would be internal, similar to the DKB Rejuvenator, include 3.1 ROMs and an SVGA chip, but would continue to use the A1000's graphics chips for normal system functions. Externally the machine would be unmodified and would still use the original Commodore monitor if you have one. The custom ROMs would boot up with the old startup tune (something we 1.3-and-later people never experienced) and display the Boing animation while the system boots behind it. There may - I'm not sure about this - be room inside the A1000 to put a low-profile hard drive, if so, include one with the board, and preload it with the OS, along with the best and most famous of Amiga software (DPaint 5, Final Writer, Lightwave, Real3D, Lemmings, Amiga WordPerfect 4.1 if copies can still be had, AmigaVision, so forth) - and sell the thing not just to Amiga cognoscenti, but to the multimedia and computer industry at large. To remind us where multimedia has been - who was there first - and who is still the best.

Just a thought.

I'm told Phase 5 has decided to go with an "Amiga-like UNIX" for the OS on the A\box (still haven't gotten used to the backslash) - and I wanted to say something about UNIX and the Amiga. There's something of a knee-jerk reflex action I see in the Amiga community and elsewhere, the notion that UNIX is what other OSes want to be when they grow up. Apple's going that route with the Macintosh - remember NeXTSTEP is an object-oriented, graphical BSD UNIX. Novell is gradually trying to turn NetWare into UNIX. Microsoft is slowly trying to either turn Windows NT into UNIX or destroy UNIX altogether. And now there's this notion that the Amiga, to survive, must become UNIX.

It's time I drew the line.

The Amiga is its own OS. It lacks a few modern features - memory protection, resource management, multiprocessing - but so do many older Unices (not just XENIX). It IS POSSIBLE to add such features without gutting the parts of the OS that made it beautiful (shared libraries, non-copying message passing, process model, graphics.library, BOOPSI) and replacing or kludging them with 30-year-old UNIX pieces that were never designed for a real-time graphical single-user OS. Sure, there are UNIX elements the Amiga could make great use of - pipes, better shell features, networking, etc. - but never forget which OS is better out of the starting gate.

Apple will learn this lesson the hard way. By going UNIX, they're going to increase resource requirements substantially - forget ever using a Mac with less than 32MB again, or on a CPU less than 150MHz. They'll lose one of the Mac's best features - its unbelievable security on a TCP/IP network - by going with a UNIX model, suddenly the Mac is open to all sorts of UNIX breakins and attacks. Plus it's going to be a pain in the ass for developers - 15 years of development suddenly stops in midstream because, while the Toolbox is still there, taking advantage of new Rhapsody "yellow box" features involves writing to an entirely different API. It may end up being possible for software to "straddle the fence" - running inside the Mac emulation layer but still taking advantage of Rhapsody services - but you're stuck with two vastly different APIs.

On the Amiga we write to shared libraries. They're dynamically loaded and unloaded as needed, you open the library, get a pointer to the library "base" address, and then simply branch into the library at a predefined offset from the library base. These offsets are defined in the various "include" files you get with the Native Developers Kit - "OpenWindow()" is such-and-such many bytes from "intuitionbase", etc. (The only permanent address in the system is Exec itself - just use exec.library to open everything else.) This is partly why it'll be SO easy to recompile applications for pOS - just get a new set of include files, and you can even use your old compiler. Note also these libraries are fully reentrant (unlike Mac Toolbox code), sharable (unlike Windows DLLs), and dynamically unloading (unlike OS/9 modules). UNIX has nothing like it. To many C programmers it'll make little difference - one function is much like another - but the Amiga library model is far slicker, far snazzier, and far more modern than just about anything else going. And if you do any system-level work, it's a damn sight better.

On the Amiga most OS services are dynamic. You can add and remove libraries, devices, handlers, networking stacks, and even entire FILE SYSTEM FORMATS (y'know, FFS, Cross-DOS, Cross-MAC) dynamically without requiring a reboot. On UNIX, most such OS-level changes require at least a restart of some critical OS function - and often it requires recompiling the kernal, or even rebuilding the entire OS around a seemingly minor change.

On the Amiga the user interface runs at super-high priority (+20). This is why things are so friggin' FAST even at 7MHz - the system makes sure the window-dragging code, button-acknowledgement code, mouse-position code, and others all operate NOW NOW NOW. This is a novel concept - one Windows still hasn't learned - that the user's requests should be processed as they are received without delay. Not like Windows or the Mac, where you click on a button and it finally highlights a couple seconds later even if virtual memory is NOT on. The Amiga is a real-time system where user input is the time-critical task being serviced.

On the Amiga, the various user interfaces (shell, GUI) live pretty much at the same level - are all ROM-loaded before you so much as see a screen - and are interchangeable. The whole system is inherently graphical, very nearly from the bottom up, whereas ANY UNIX will by definition be text-based with the GUI as an add-on. Like UNIX, however, the Amiga treats the GUI as just another task - unlike the Mac or Windows where part of the multitasking code is buried several layers out on the GUI code.

Nothing against UNIX. For a network server it is unbeatable. It's fast, text-based, file-driven, and really does form the backbone of the Internet. (I started to say "literally" but realized the Internet does not actually have a backbone in the literal sense - it is not a vertebrate, it does not have a calcite skeleton. A notochord perhaps, but no skeleton.) For industrial-strength transaction processing (like Web serving) it's the best. And POSIX is cool and cross-platform - a useful set of libraries if there's a UNIX tool you simply MUST have on your OS.

But never assume it's the destiny of every OS to be UNIX someday.

Now then, speaking of Rhapsody... Apple's in the middle of their World Wide Developers Conference (sorta like DevCon but without the Lemmings re-enactment - just not conducive to the corporate culture Gil's trying to promote, y'know), showing a demo of Rhapsody Pre-Pre-Beta 0.001 Demo Beta Pre-Alpha Demo, Beta Version to a bunch of Mac developers. Apparently what the developers saw was a standard Mac menu bar at the top of a screen where most of the windows were in the OpenStep style. Mac OS 8 ran in an early version of the emulator - this is how MacWeek phrased it, not Mac apps, but Mac OS 8 itself - and it all went so slick and problem-free that at the end of the demo, someone in the back of the room yelled "SHIP IT!" This is an encouraging sign. Maybe, just maybe, being a pessimist pays - by preparing oneself for the worst, one is always pleasantly surprised.

Better still? Open Transport lives! Apple listens to its users, odd as that sounds - for the first time in recent memory (perhaps in their history) Apple has exhumed a discontinued technology because of public outcry. Open Transport was the biggest tragedy of the Great Apple Surgery of 1997 - it had come so far (from the buggy 7.5.2 pre-betas to the sleek, fast, stable networking protocol we now know and love) and seemed to be perishing before its time. In many ways it's superior to the Berkeley Sockets mechanism Rhapsody was supposed to use. Well, all the developers who spent the last year or more making their networking software cooperate with Open Transport finally yelled loud enough for it to sink in. Apple has decided not to abandon Open Transport yet after all (cue filmclip of Ellen Hancock on the balcony singing "Don't Cry For Me, Open Transport" in her best Evita voice) - they are reactivating the technology, they are shipping a long-awaited bug fix for System 8 due this summer, and they WILL ship the multiple-IP version of OT in the next System version after 8. (Virtual hosting! YEEEEESSSS!!!!!!!) Better still, OT will be available to Rhapsody applications - not sure whether it'll be in addition to, or instead of, the BSD sockets.

Apple's slowly getting it together. Of course, they still need to learn how to communicate with dealers (they're nice to their developers, but to their dealers, they might just as well say "grab your ankles" and saved everyone the legwork) - they still need to grasp that the reason Mac clones are costing Apple money is that Apple hardware costs too much in the first place - they still need to grasp certain basic marketing philosophies. In some British Amiga magazines, you'll find Apple full-page advertisements saying "Why Macintosh?" Guys, this is not your market. Leave the poor Amiga blokes alone - we have the machines we want. Might as well try to sell VI to an Emacs fanatic - or tell me to trade all my Sarah McLachlan discs for Tori Amos. Why is Apple trying to sell Macs to Amiga people? Desperation? How much market share do they expect to gain? Admittedly in Europe the Amiga actually HAS market share - but still. That full-page ad should have been running in PC Week instead of CUAmiga. Believe it or not, such an ad in a PC magazine WOULD make a difference - it would remind people the Macintosh still exists, for one thing, that Apple is still in business, and has enough money to put ads in computer magazines. Psychologically people think if you don't advertise, you don't exist - and Wintel boxes sell to people who make purchases psychologically. The Mac needs to be in that market, selling their way UP the chain instead of down it.

All I can figure, though, is that ad space is cheaper in Amiga mags.

Which points to something fundamentally wrong with Apple's marketing approach, doesn't it? The only "big" magazines Apple ever advertises in? MacUser, MacWeek, MacWorld, Internet Week, some graphic design journals, and Byte every now and again. (This month's Wired cover - of Apple with the crown of thorns and the caption "pray" - doesn't count.) Does Apple think the only market for new Macintoshes is people who're throwing away their old Mac Pluses and A500s? Are they scared former Windows users will immediately spot the under-the-hood flaws in the Mac and go running back to Microsoft? Does it matter anyway if they do? They've BOUGHT the damn machine, which is what Apple wants - and if they don't like the Mac they bought, they'll sell it to someone who WILL like it. The Mac has a cleaner user interface than Windows - yes it has flaws (applications running but with all windows closed, the user thinks the application is closed but it's not) but Windows does too (it's possible to "accidentally" hide your Start bar - not to mention that nasty Windows trick of a window-in-a-window where the inner window can get lost outside the borders of the outer window that has no scrollbars!) - it's just a matter of unlearning bad habits. The latest round of Macs are snot-fast, and while they're not as cheap as PCs, they come with all the pieces PCs charge you extra for, like 24-bit color, fast SCSI, and 16-bit sound. Clone Macs ARE cheaper than PCs - but Apple isn't trying to actively sell those. (Maybe Power Computing and UMAX and Motorola should advertise in PC mags? Hell, maybe IBM should make Mac clones. That would legitimize the Macintosh for businesspeople.)

(How to get around the "application open with all windows closed" problem? Just shadow the entire rest of the screen. On the Amiga this would be equivalent to a program's custom screen that stays open after all windows are closed - like Excellence. So long as the user has a visual clue - and not just the icon in the corner of the screen - they'll figure it out. Are you listening, Apple?)

Again, I mention all this because Apple is still the only off-platform company big enough to take on Microsoft. They survived open warfare with IBM. And again, once Microsoft is down to a manageable size, then we'll take our souped-up Amigas and put the Mac back in its place. :-) But until that day comes, we need Apple - and an Apple who's dying where it stands is less use to us than an Apple whose market share is actually increasing for once.

Apple, at any rate, is too much a part of human history now to be sacrificed. Xerox, Windows, and the Amiga all be damned - two thousand years from now, when they tell the tales of how computers came to be, Apple will still be the name most mentioned. Apple is not a company of guys with ties and horn-rimmed glasses, it's a company of hippies and grandmas, of everyday people. They're human, which is unusual for a computer company. They named their company after a kind of fruit, an organic thing, instead of an acronym or a buzzword or "International Business Machines" or the like. Apple began and remains a company of people who LIKE making technology. Apple very much speaks to the creative freewheeling element in all of us, a computer company that rollerblades instead of taking the bus. Even in Silicon Valley it's a rare thing - they were the first "pop culture" computer company, the first name brand computer to speak to the humans out there instead of the geeks and corporate weasels, the first computer that non-computer types weren't afraid to admit they owned. Apple didn't invent much - but they sold it to the masses, or sold it to those who would sell it to the masses. They impart style into everything they do. Apple is not a company - it's a way of life. Larry Ellison forgot that. Gil Amelio is forgetting it slowly. But let's face it - if Apple ever folds, it will be a world-shaking event. Entire companies will close along with it. Industries will cease to exist - or slow to a crawl while they figure out how to make other systems do what they need. Microsoft will no longer have anyone to steal ideas from - and their OSes will stop improving. And unless Gateway sells LOTS and LOTS of Amigas, if Apple folds, there will be no visible standard-bearer for non-Wintel machines.

Again, make no mistake: once Microsoft is out of the race, or far enough behind that we don't have to worry about them anymore, the Amiga and Macintosh will be competitors again - and the Amiga will probably resume blasting the Mac's ass clean out of the water. But until that day, we must ally ourselves against a common enemy.

The wait is over, folks. But the war isn't. We have just been granted another chance from above, a chance to make computing what it should be, a chance to crack the monopoly wide open, a chance to put the personal back in personal computing, a chance to put the Amiga in its place of honor for the first time in years. We have an owner - and unlike the last two or three, I don't get bad vibes from Gateway. They make decent PCs - they ADVERTISE - they have money and will be trading on NYSE real soon now. The people in charge of the Amiga are people we know - Petro, for instance. We have a long march ahead of us - once the champagne from the celebration is all gone, we must all get down to the dirty business of reassembling the Amiga community and Amiga marketplace from the ashes ESCOM and the year-long bankruptcy have left us with.

So get the word out! Evangelize again. The Amiga is back in business. We WILL have product again soon. We WILL have new OS development again soon. This time around we have voices that are heard by the Amiga's management. This time around our parent company isn't in danger of going away. This time around our destiny is something we CAN control - it is controlled by people who Understand, who Listen, and who are smart enough not to fuck it up. The details matter nothing right now - whether we go PPC or Alpha, whether we go NewIcons or MWB or Uberbench, whether we go SVGA or custom chipset - what matters is that the slow 680x0 Amigas will be a thing of the past, as will be the old 4-color Commodore icons, as will be slow AGA refresh. And I don't think anyone can argue the value in that.

I've waited a long time for the chance to write a positive Rumor Mill. The wait's been worth it.

Next week, that new look I've been promising.