The end of an era.

September 10, 1996

Admittedly it's a minor thing in today's marketplace, but it's very important to computer historians and hobbyists.

What is the event I'm talking about? Simply this: yesterday, after a torturous year of problems and setbacks, Apple discontinued the 190 and 5300 series PowerBooks.

The 5300 series (as seen in Mission Impossible and Independance Day) had it coming - exploding batteries, software glitches, product recalls that make General Motors run and hide. But the 190 series is somewhat heartbreaking:

The Powerbook 190 series was Apple's last 680x0 Macintosh models.

Apple has done the inevitable, completing the most remarkable transformation of a computer line ever - switching processor design in midstream. It's one thing to go from a 68000 to a 68020, as Apple and Commodore and Atari and everyone else did in the late '80's. It's one thing to go from the AT&T Hobbit processor to the PowerPC, as Be did - before the unit was ever even shown to the public, and before there was any software. But to take twelve years of legacy software and move it from one processor design (680x0 CISC, 32-bit memory bus, 16-bit instruction set) to another completely different design (PowerPC RISC, 64-bit architecture) while preserving compatibility and making it absolutely seamless... I still say this is the one thing Apple has done absolutely right.

And that transition is finally complete. The 190 was a 33/66MHz 68040 laptop, sold in several different configurations with color and monochrome screens. It was introduced in 1995, making it one of the last 68040 systems introduced... and now it's a part of computer history, the last in a proud legacy.

Make no mistake: the Motorola 68000 and its descendants are some of the finest microprocessors ever made. The original 68000 was billed by everyone as a 16-bit processor in 1981 - but inside, it was more like a 32-bit. The 68020 was a truly 32-bit device - 32-bit data bus, 32-bit memory bus, and 32-bit math architecture... plus hyperadvanced features like pipelining (executing instructions in stages, so one instruction can be writing results while the next one is getting started - an instruction that needs no work can execute in zero clock cycles!) and caching (keeping needed data in high-speed memory directly on the processor). The 68020 was, in my book, the perfect microprocessor - a symmetrical instruction set that truly made SENSE (way better than the 6809, and miles better than those damned Intel chips). The 68030 was just an improvement on the design - bigger and better cache, more speed-increasing features, and of course, higher clock speeds (mine at home runs at 50MHz). And of course, the oddball of the lot - the 68040, a 68030 with an integrated math unit and memory-management unit. (problem with the 68040 was that the math unit lacked trigonometric functions, and had to emulate them via an external software library or enabler)

The 68000 series has been used in everything from calculators and game machines (the Sega Genesis) to personal computers (the Macintosh, the Amiga, the Atari ST, the Sinclair Quantum Leap, the TRS-80 Model 16) to printers (the Apple Laser Writer and others - new printers are still coming out with 16MHz 68HC000 processors!) to heavy-duty superboxes (Sun, NeXT, IBM, and even Commodore have all made industrial-strength UNIX systems out of 680x0 processors). And now, with Apple's decisive move, which really was intended more to ease their headaches with these problem-prone laptops than to finalize a transition, the 680x0 series is gone forever from the mainstream desktop computer world, leaving it to the realm of printers, used Macs and Segas, calculators and remote controls, and of course, Amigas.

This makes it all the more important that we get a seamless RISC architecture going on the Amiga - we're the last sorta-major computer platform left on the planet which still uses Motorola 68000 series chips... a dubious honor at best, as the 68060 is a dead-end street and tops out at 50MHz.

Are we having fun yet?

August 30, 1996

He he he heeeee he he hee he he hee he... barf.

Y'know, to be quite honest, when the non-Amiga folk ask me why the Amiga isn't going anywhere these days, I have an answer. This ain't blamable on Commodore anymore. This isn't Microsoft's fault, or Apple's, or IBM's, or SMG's, or Atari's, or Jack Tramiel 's, or Ed Goff's, or Jason Compton's.

Whose fault is it the Amiga is dead in the water? Ours, of course. Ask five people what the Amiga is, or what it should be, and you'll get at least five answers. PCs don't have this problem - the notions of "what a PC-compatible is" are well-engrained in people's brains. Macs sorta have this problem - everyone agrees Apple needs to advertise, but no two people have the same ideas what the commercials should be. The Amiga community and market are fragmented to the horrible extremes - we're witnessing the m odern fall of Babylon, where we finally just piss each other off to the point where we all walk away.

Oh, what brought this sudden bit of negativity on, Squid? Sure, we've long known that no two Amigans have the same opinion of what the Amiga is and means, look at the arguments the Walker turtle-case prototype started, look at the MUI debates, look at the custom chipset debates. But what brought the Squid to this degree of frustration?

Simply the notion that the leaders of the Amiga market ought to be above the bullshit. Here's what Pios had to say not long ago:

"New Product announcements from PIOS Computer AG has been eagerly awaited by both the visitors to our Web-Page (over 46000 since May '96) and the marketplace throughout Europe and the USA.

Unfortunately this delay has been caused by the inability to secure a working agreement with Viscorp over the rights to the AMIGA patents and licenses.

Despite repeated efforts by PIOS management Viscorp have discontinued discussion, without giving any reasons or otherwise. This is despite initial talks and drawing up an agreements early June of this year."

I feel sorry for Jason Compton. He's Viscorp's Amiga PR man now, and he has the unenviable job of putting a decent spin on all the weird Viscorp corporate crap. His explanation for this event is that since Viscorp doesn't yet have the Amiga, they can't li cense it. Which makes sense, but why not just send Pios a fax that says "Sorry - call us in September and we'll talk?" If the Pios message is to be believed, they heard nothing but a sudden deafening silence.

OK, so... now we're stuck. I've heard that Phase 5 has had similar problems getting any kind of meaningful discussions going with Viscorp. We went through the corporate arrogance thing with Commodore, I thought we'd learned our lesson. Viscorp seems to be one of those arrogant American companies who seem to believe that in order to make money at something, they have to make all the money.

But that's not the point of this. Read Pios' statement again. Is this not a corporate temper tantrum? Admittedly it isn't SMG's legendary "the Phoenix is dead, save yourselves" hissy-fit, but it isn't exactly professional either. There are supply problems , and then there is announcing to customers you don't even have yet that you can't negotiate a deal with the company who doesn't yet own the product from which you're spinning off your own product line. I dunno - maybe there was no other way to handle thi s, maybe it is as bad as it sounds.

But it doesn't matter in the long run, because every day that Viscorp and Pios don't talk to each other is one more day we sit around watching Macintoshes run past us in the 240Mhz fast lane.

So what about Phase 5? Oh, this is juicy too - remember when I posed the question, "Just what kind of API are Phase 5 going to make for a Power Amiga?" They've answered it. Basically, the Power-Up boards are modified Cyberstorm Mark II boards with two pro cessors, a 68030/040 and a PPC 603e. The 603 side of the card will probably run at a lowly 150MHz. (grin) 32-bit memory. Zorro III. Basically it's cramming a RISC processor into a bus never designed to accommodate it.

But the software API is what concerns me:

"We searched for a possibility with a loose binding to the underlaying operating system. We achieved this with a library, which enables the loading and execution of PPC tasks. Those tasks are running parallel with the task on the 680x0 CPU and are able to communicate with other tasks, but without the possibility to call OS functions directly. Each application consists of a 680x0 main process (of any size) amd PPC subroutines loaded and started by PPCLoadSeg() and PPCAddTask(). A native PPC scheduler allow s multitasking even for the PPC, and multiprocessing is targeted for future versions."

For everything Apple's done wrong, the one thing it did incredibly right was the transition to the PowerPC. They successfully took ten years of 68000 legacy code and coupled it with a RISC processor and RISC-native software, and made a whole system so sea mless even the operating system itself is still partially 68000 and partially RISC. One processor pretending to be two at the same time. Even 68000 extensions and control panels (the Mac equivalents to commodities and startup-sequence items) work fine (al beit slow) on Power Macs. And best of all, new versions of the OS give you a speed boost, as the interpreter gets faster and more of the OS becomes power-native: software acceleration. The average Power Mac user will never know or care what "68000" means or any of that - except that so-called "fat" programs and "native" programs are really really fast.

So here we are on the Amiga side of the fence with an expensive DSP card that also has a 68K processor on it. Most big-box Amiga owners already have 68030s and 68040s these days, why sell us yet another processor we don't need? What we wanted was the Powe rPC as the main processor, emulating the 68K stuff so we can start writing RISC-only Amiga code, looking forward to the day when the emulation stuff (and all trace of the obsolete 68000-series) can be deleted without hurting new apps. What we got was... a multiprocessor solution even worse than DayStar's.

Now, again, Phase 5 can't do very much until Viscorp gives them something to work with. But again, someone WILL have to define a PowerAmiga API standard and soon. Otherwise, you'll have three different incompatible Power Amigas on the market, fragmenting an already small market, such that all the frustrated would-be buyers end up heading to the PC stores instead.

Of course, there's one way out of all this, which has been proposed before: Let's all, Amiga users everywhere, pool our money and buy the Amiga our damn selves and get it over with. We'll finally control our own destiny for once. But the reason it hasn't happened is quite evident when watching Pios, Phase 5, and Viscorp dick it out publicly: Amiga people each have their own ideas of how it should be done, ideas that someone else will never agree with.

Oh, sure, we could do it - $40 million to buy the Amiga out from under Viscorp, we'd have to shell out between $5 and $10 per Amiga-lover worldwide to really secure the Amiga and have some money left over to actually build the crazy things. But then, once the check is written, who gets to be in charge? Who gets to design new Amigas? Do we go with PPC or Alpha? Do we build set-top boxes? Do we make Stealth Turtles? 3-D chipsets? Custom chips? UNIX Amigas? Built-in Emplants? MUI as a standard part of the OS ? Multiprocessor? Laptops? 15KHz screens? NTSC/PAL considerations? A1200-style cases? CD32? Voila - the makings of a bar fight the same night we all go to the pub to celebrate our victory. Amiga users are stubborn - it's a blessing and a curse, it's why w e stuck with this platform despite the odds, and it's why now, without Commodore's momentum, we're dead in the water because we can't quit fighting long enough to decide which way to go from here.

Other news. Apple is having to squash rumors that they're actually going to license the BeOS and go with it instead of Copland. Admittedly, BeOS for Power Mac way outperforms the Mac OS on the same box. It looks very Copland-like, complete with a Mac-like screen font and Mac-like onscreen buttons and gadgetry. It makes an excellent match for Apple and Power Computing's excellent hardware. There's just one teensy problem: it doesn't run Mac software at all.

This probably won't be a permanent problem. Already Mac emulation has been done on the BeBox, using a horrific combination of UAE and Shapeshifter, courtesy Christian Bauer - keep in mind Shapeshifter normally requires a bigger Amiga than what the Unix Amiga Emulator provides, but CB hacked a special version of Shapeshifter just to show off, essentially a Mac Plus emulator within an Amiga emulator within a BeBox. (A similar problem exists with the Amiga version of UAE - it requires AA, and is therefore no t yet capable of actually running itself, and yes, Amiga people are crazy enough to do just that. The Amiga version of UAE is already proving itself useful on fast machines when trying to run older games and demos that won't run on newer Amigas!)

Point being, Shapeshifter (and Executor, a PC product that runs Mac software as Windows applications) prove that Mac emulation is a possibility. I've often wondered about the possibility of a special Mac emulator that intercepts Mac Toolbox calls and subs titutes Intuition or MUI operations instead, thus "natively" running Mac apps right there on your Workbench. Do you suppose a similar solution could be added to Be? It's against the Be philosophy, of discarding legacy stuff in favor of new technology. But that's exactly what Copland would have done anyway - break all kinds of existing apps to allow for its newer multitasking and memory-protection goodies. And since Copland wouldn't have offered that much in return, Be seems a better choice - especially if a Be equivalent to Shapeshifter (with native BeOS GUI calls) actually runs old apps better than Copland would have.

As to Copland itself, let's face it - it ain't happening. The entire OS is appearing in the Mac market an extension at a time: Open Transport, Aaron, Now Utilities. Part of System 8's mantra was PowerPC Platform compliance - but it looks like Apple's just going to release a version of System 7.5 for PPCP after all. Copland is THAT far behind schedule. As it happens, there are extensions that already do things Copland WON'T do - I just recently found one called Smart Scrolls, that gives the Macintosh propo rtional scrollbars! (And though they act funny in certain applications like Netscape, and under certain conditions cause BBEdit Lite 3.5.1 to crash altogether, they work more reliably than the MUI-based scrollbars in most Amiga web browsers!) Aaron gives me the Copland look NOW, Smart Scrolls gives me the proportional scrollbars NOW, all I'm missing is the multitasking.

I've finally taken the plunge and installed UAE 0.6 on the Power Mac 7100 at work. Tomorrow I'm digging out one of my old Kickstarts, making a Workbench disk image, and heading to the shop to try the fun stuff out. I already checked out the RSI Megademo, it's so slow it's worthless, but maybe Workbench will fare a little better - after the bizarre Mac extension conflicts I've been having lately, I'm perfectly used to slow graphical OSes. :-)

One last thing: Viscorp is having a "contest" to design their new logo. They want the check mark in the V of Viscorp, they might want the Boing ball in the O, they want it done on an Amiga, and there's no cash prize. Keep an eye in the art.marketplace newsgroups for a lot of SCAMS that look awfully similar to this - how do you tell them apart? It's basically "Send us your best ideas, we'll decide which one we like best and not pay you for it." Yes, Amiga folk will happily do it for free - but that must never be construed to mean you should take advantage of us like that. Sounds like a neat trick, though - next time I need some graphic design work done, instead of doing it myself or paying someone else to do it, I'll just hold a contest.

I'm hideously tempted to say "I hope they get sucky designs," but then those sucky designs will appear on new Amigas. So in the interests of Amiga evangelism, I might be working up some ideas - so when they do pick a sucky design it won't be my fault. (They will pick a sucky design - you watch.)

The Internet: What is and what should never be

August 26, 1996

I don't know about you, but I think the Internet is going down.

Seriously. Sprintlink is at it again this week, worse than ever, fluttering so violently that at one point last week the entire East Coast apparently vanished from the Net. Ameritech also has a NAP somewhere in the middle of the backbone - where MCI and Sprint join, I think - and it's such a poor piece of hardware at that location that during peak hours of the day, most of the Internet runs at the speed of a 9600 baud modem across that link.

So basically the infrastructure of the Internet in North America is crumbling under its own weight. More people are getting on the Net - and they're getting faster, with things like ISDN and cable modems, not to mention heftier upstream technologies like ATM - but the Backbone itself isn't keeping up. MCI just installed a terabit switcher not long ago - and they need another one. It ain't been that long, folks, since they replaced the core set of PDP's with Butterfly arrays... y'know, back when all your traceroutes went through ARPANet... and in those days, those BBN Butterfly gates were considered fast.

The Internet is exploding. Quite literally. It's uncontrollable now - you either ride the wave, get out of the way, or you get destroyed. The Internet is multiplying faster than any other technology in history - even microprocessors, with their double-every-eighteen-months record unbroken, lag far behind the Internet's ten-times-as-big-every-year. (Want proof of how bad processors lag behind the Internet? Go to AltaVista, a 300MHz DEC Alpha with 6GB of RAM that can't keep up.)

The worst thing about all this, though, is that the Internet's signal-to-noise ratio is way down. When I first found the Internet in the fall of 1994 (which is ancient history to many modern netsurfers - this was before Mosaic!) there was all this great stuff you could get. You could ftp over to cs.uwp.edu (if you were one of the lucky 25) and download pictures and song lyrics from your favorite musicians! Or go to ftp.wustl.edu and download all kinds of nifty pictures to show off your graphics card. Or for that matter, stay at ftp.wustl.edu and go to the Aminet directory - hell, this was keener than the Fred Fish collection! (And whadda ya know, the Fred Fish collection was online too!)

And then we discovered this thing called the World Wide Web, which was sort of the Microsoft Windows to the Internet's MS-DOS. (A better analogy is the X to the Internet's UNIX interface - which, given the NeXTStep origins of the Web, is accurate on more levels than you might think) I remember seeing the White House page pull up the first time - that gorgeous front-page graphic with the gold-embossed offlinks... I remember checking out Nettwerk Productions' homepage - they were the FIRST record company in the world to have a page on the Web, and a damn good page it was too, full of content and cool pictures and nifty stuff about all their artists (especially Sarah McLachlan). I remember www.wit.com - which was what we all used as our jumping-off point for the Web proper - they had links to all the big FTP sites! (Oh, the coolest part about Mosaic in those days was that you could use it as a souped-up FTP client!)

I remember what I considered at the time the best browser around, Internet Works. It loaded quicker than Mosaic, and had a much nicer interface - your "history" was a series of pulltabs along the bottom of the window. "Web pages" indeed - you really felt like you were going through pages. And I remember when the school decided to replace Internet Works with this sucky new browser called Netscape - an ugly, slow, buggy, horribly crash-prone piece of shit. (This, for history's sake, was Netscape 1.0 Beta... with the "throbbing" N in the upper right corner... made even Mosaic look good.)

Anyway... back to the point. Back then, EVERY page I looked at seemed to have content. But then, back in those days, nobody knew how to write HTML, so anyone writing pages HAD to have something to say first - so they could communicate it to the Internet pros who actually knew how to do this stuff. (I remember back when the White House's forms guestbook was broken because they didn't know you had to have a CGI script behind it...)

Nowadays, browsing the Web is like listening to Beavis and Butt-Head review music videos: "If I turned on the TV, and this video was on, I'd just go... "yep, that's exactly what I thought I was gonna see." Web pages look like Web pages look like Web pages. The buttony graphics, the <H1> </H1> titles, the "Under Construction" sign with the little guy with the shovel, the marbledy-looking "back arrow," the little red dots everyone uses for bullets, the same fifty backgrounds on which text is unreadable, the same dozen links (Yahoo, Alta Vista, Gamelan, Happy Puppy, Stroud's), the "mile-high-by-640-pixel-wide-unbroken-page" pages, the "everything-is-on-the-left" pages, the "Click here" routine, the Java scrolling banners, the pointless animated GIFs they stole from some big GIF repository, the spinning compass, the little red devil, the annoying way people put <blink> around the text they MOST want you to read, the JPEGs compressed with ridiculously low compression settings so they look like they're underwater, the 180K background images, the useless three-second META refreshes that don't let the page load completely before taking you somewhere you never volunteered to go, the 800-pixel-wide graphics at the top, the backgrounds designed for 640-pixel-wide screens... and the near-total lack of real CONTENT on so many pages. Most people on the Web - including COMPANIES who paid some web crew big bucks to build a page - have nothing to say and no ability to say it.

A big-budget example of this is Microsoft's homepage. The biggest software company in the world... and their homepage is unnavigable. There is no internal consistency. There is no overall structure. There is, all through their site, the eerie sense that pages are moving around - because you keep seeing the same three pages but the links aren't yet visited. And no matter how hard you try, the page you WANT to find and you KNOW is there, you cannot find it. But then, graphically their whole site looks lousy anyway - yes, it's supposed to look like a magazine, but then, Microsoft's magazine ads are ugly too - and this is the Web, not a magazine.

Microsoft doesn't understand the Internet. Just like they don't understand the Macintosh. Microsoft sees the Internet as something to conquer. More precisely, they see the Internet as the one place they nearly got their asses kicked (by Netscape) - and they CANNOT bear to lose that battle again. But rather than actually embrace the Internet, they would rather install their own Internet over the top of it. They said "Java is nice, and we'll do Java, but we'd rather you use ActiveX because we make it." They said "UNIX is nice, and we'll be compatible with some of the stuff UNIX does, but we'd rather you use NT because we make it." It's always been like that - way back when, they said, "The Amiga's OS is nice, but we'd rather write our own window-refresh routines for AmigaBasic, because that way it's our window refresh routines and not Commodore's." And they said "The Macintosh OS is nice, but we'll sidestep it any way we can, because that way it's us and not Apple." (End result: Microsoft Word for the Mac acts and feels like a Windows program. And we won't talk about FoxPro.) They said "Netscape is nice, but we'd rather do things our way - and make our browser JUST different enough that you have to switch." (And then, by extension, go to Windows 95 because IE doesn't work on the Mac.)

Speaking of Internet Exploder...

Version 3.0 was released last week, amid much hoopla and fanfare. Oh, woooowww... far out... the cool ads about "Remember the freedom you felt the first time you got on the Web?" (no, but I remember the freedom I felt the first time my Amiga loaded a Web page... and I realized i didn't have to own a PC to do what i wanted to do.) But in the midst of all the hoopla, one nasty little thing went unnoticed...

Internet Explorer 3.0 is crippled. Badly. Literally, it's as if a big piece of the code is missing.

What exactly is broken? Secure sockets layer. Y'know, the mechanism by which Internet packets can be encrypted - and thus cannot be intercepted and opened? The mechanism which lets you transmit your credit card number across the Web to buy stuff, and not worry about some loser hacking it? Their entire SSL mechanism is badly damaged. Depending on the site, and the SSL server being accessed, IE3 may choose to a) refuse to access the site, giving cryptic error messages like "Server was reset", b) refuse to access the site because it doesn't recognize the key (Netscape gives you the option of accepting a "questionable" key anyway - if you're sure you want to trust the site), c) deliver a corrupted page, d) accept the site EVEN IF you specifically told it in its configuration NOT to accept the site, e) accept the site with bizarre error messages, or if you're lucky, f) deliver the site as expected.

This is a serious bug. MCI has already said it will not deliver IE to its InternetMCI customers until this is fixed - several other companies have flat-out said the "final release" of IE3 is just another beta. The rumor I heard is that literally, the day before the final release, someone accidentally forgot to include a big piece of the source code, and the final compile was missing stuff - my source said "Someone's going to get fired over this."

To which I added, "Or promoted."

Meanwhile, on our side of the fence: the rumor got going around that Microsoft was working on an Amiga version of Internet Explorer. Now, this would be cool in the sense that we'd have a brand-name browser - but IE for Amiga would suck with a capital K. The last time Microsoft wrote anything for the Amiga, it broke under any of the following conditions: a processor other than the 68000, a processor faster than 7.16MHz, memory above 1MB, memory that wasn't Chip RAM, memory in the 32-bit space, OSes above 1.3, chipsets made since 1988, hard drives, etc. AmigaBasic is one of about five 32-bit-unclean Amiga programs that never got fixed. Between v1.1 and 1.2 of AmigaBasic, Commodore sent Microsoft a list of reported bugs - it was 130 items long, things like "if you scroll down, the machine hangs."

Internet Explorer for the Amiga would have to require nothing less than a 68030 with 8 megs of RAM to be even remotely usable... and would probably be a dog on anything below a 68060 with a fast hard drive. This assumes they could make ActiveX work on the Amiga. And this also assumes they actually write FOR the operating system instead of around it.

And Internet Explorer is not what an Amiga web browser should be anyway. I've been thinking about this: a REAL Amiga Web browser.

All the existing Web browsers are basically alike - mostly MUI, basically little more than literal copies of Mosaic or Netscape designed to run on 256-color screens. The Amiga doesn't have to be like this... at least not while we have custom hardware nobody else has. It may be slower than the other guys, but it can still do some things better than the rest...

How about displaying a Web page as a HAM or HAM8 scrolling superscreen? No scrollbars - you move your mouse and the page goes with you. No pen-mapping - this is HAM. You could even do nifty things like "frames" - overlapping screens. Or copperlist gradients - what's stopping us from adding Amiga-specific tags? Embedded ANIMbrushes? For that matter, instead of the links being "blue underlined," why not make them "3-D button" - like AmigaGuide? This is OUR computer, this can be OUR browser, and as Microsoft shows, you only have to play 90% of everyone else's game, and have fun with the rest.

But no. Everything's MUI, except AWeb which is incomplete. 256 colors on AGA is dreadfully slow, especially at scrolling. Datatypes eat so much chip RAM it isn't funny - never mind I have sixteen megs of Fast RAM. Even ALynx doesn't work like Amiga Lynx should work - is it that hard to make the browser resize to the window? Even Amiga Tin does that much!

Amiga web browsers do not work on low-end systems. They can - but not happily. They're all designed for OS 3.0 and up. They need 256-color screens to look decent - even on machines that can display 4096 colors in low-res happily. In short, Amiga web browsers are not designed for the Amiga. They kludge their way around what the Amiga does not do well, without taking any real advantage of what it does well.

But I'm not much of a programmer, so I can't solve the problem singlehandedly... someone else gets the dirty work. This is a call to arms for Amiga programmers - make the Amiga web browser everyone REALLY wants.

Well, guys? What do you say?

Various and sundry

August 14, 1996

Lots a goodies today... nifty tidbits and rumors from the MacWeek Expo, but I gotta hurry - I'm on lunch break.

First off, an anonymous rumor sent to me - a member of the informant's family works at an advertising agency, and apparently Viscorp has been talking to that agency. Now, this came just hours after I got the latest newsletter from my local Amiga dealer (Digital Arts in Bloomington, y'know, the ones made forever famous by that 1993 confrontation with John Dilulu?) in which they explained precisely why we cannot expect to see Amiga ads on TV. Which means... Viscorp, if they're going to advertise, are going to advertise Ed. Ed is an Amiga but is not an Amiga - to market an Amiga you have to go up against Microsoft's marketing engine. To market Ed you have to go up against any other company who's making set-top boxes - which ain't too many. They're out there, but not storming the market.

Next up, the bizarre rumor that Microsoft is considering buying Netscape. It's been posted to ALL the newsgroups - as threads with Microsoft in the subject typically are these days - you with killfile-capable newsreaders can make USENET a more enjoyable place by killfiling the word Microsoft and the word "MACINTOSH" in all caps - and thereby eliminate a lot of spam. But it kinda makes sense knowing Microsoft's tradition. If they DO buy Netscape, not only does it mean Internet Explorer will become THE standard browser (for you who haven't used it, its cool features do NOT outweigh its terribly slow speed and bugginess and memory wastage) - but also, it will eliminate the ONLY company who has in recent years actually embarassed Microsoft. This means, by eliminating Netscape, Microsoft will eliminate the only reason they have for actually improving their Internet software. Which means IE3 will NOT get any better... think about that for awhile.

And now the goodies. MacWeek Expo this year has the same positive feel as last year's Video Toaster User Expo - the one where we all went home expecting a bright future. (VTU Expo was where they announced the Power Amiga, and a bunch of other goodies - in a single move, erasing any pessimism we felt about the Amiga at the time.) Apple's own spiel was rather Viscorpey, a lot of positive stuff but no broad sweeping Grand Victory Plan... and no Super Duper PC Killer product in the works. They DO, however, have some fascinating stuff on the burners.

Copland, aka System 8, aka The Neverending Story, aka Mac '93, has changed form. They've finally decided there are things in Copland that CANNOT wait until 1997 or 1998 (when the figure they'll finally get the damn thing done - a product that is already three years overdue!) - and so they will release Copland not as Copland but as a series of interim releases. First up is the rumored Harmony, which will probably be something like System 7-and-three-quarters. Sporting 7.5's old crappy cooperative multitasking engine but with the "classic" Copland interface (which the Aaron system extension will give you NOW), Harmony will contain all the Copland stuff that works NOW that does not require Copland's multitasking engine. The actual Copland kernal will appear next year sometime... and the FULL preemptive Gershwin (aka System 9) multitasking kernal may come MUCH later.

Apple is reportedly working on some way cool designer case designs - the new Performa 6400, to my knowledge the first tower-case Performa, is supposed to be a harbinger of things to come. Also on the hardware front, Apple has in the wings a couple of multiprocessor systems - using the same multiprocessor API as Radius' and SuperMAC's own dual- and quad-processor clones. (A multiprocessor Macintosh has little use for anyone except Photoshop power-users, at least until the OS supports multiprocessing. Today, your quad-150 clone basically runs like a _single_ processor Mac, until a multiprocessing-enabled Photoshop plugin starts using the extra processors. A multiprocessor Mac is functionally a single-processor Mac with three REALLY fast DSP chips... but supposedly Copland contains a true symmetric multiprocessing engine - see above.)

Meanwhile, Power Computing, currently the record-holders for the World's Fastest Personal Computer, the screeching 225MHz PowerTower, are already showing off 240-MHz models in both 603e and 604e flavors. The 603e variety, with 16MB and hard drive, is rumored to sell for about $2200... folks, that's the equivalent of a 300MHz Pentium (which does not yet exist) for the price of a 120MHz Pentium. As one of MacWeek's editors said, "I'm glad Power are on our side."

In similar news, Be just announced their latest box - the BeBox now comes in two varieties, the dual-66MHz 603 variety, and a new dual-133MHz 603 variety. The twin-66 has already dazzled a lot of people with its raw speed, and is no slouch - a twin-133 model would be a screecher, especially at $3000 loaded. Expect to see even faster BeBoxes soon, as the faster chips become stable reference... Be seems to like waiting for a chip to settle down before using it.

Possibly the strangest story from the Expo is that of Be's own demo booth. They got a phone call two days before the Expo, from the organizers of the Expo, saying "We just now realized there's been an error - you guys can't do a BeBox demo at the Expo, you aren't Macintosh developers, we're sorry we misunderstood."

So the guys at Be looked at each other for a moment, and said, "We need a Mac product."

The result? Beneath a sign that said "Not yet fit for human consumption," Be had a single-processor version of BeOS Developer Release 8 running on a Power Computing Macintosh clone. Yes, you heard it right - Be for Power Macintosh. And they did it in two days - and they did it on a whim. It isn't a sellable product yet, and probably won't be yet - but they've already decided to call it - ready for this? - "OS Doubler." Between BeOS and MKLinux, that makes TWO preemptive multitasking operating systems that run natively on the Power Mac. (UAE doesn't count...)

Be is studying the problem of making the BeOS support the multiprocessor Macintoshes. Stay tuned.

Java, Java everywhere - Apple says before the month is out, Java will actually become part of the Mac operating system, as part of OpenDoc. Now, nothing really against forward-moving technologies, but my experience has been that Java is slow, unstable, and memory-hungry on a Macintosh - and OpenDoc reportedly has similar problems. Cool - let's combine the two. Raise your hand if you see a problem with this picture.

Java is not the panacea everyone wants it to be. It is slow by design, as interpreted bytecode for a virtual machine - but Mac Java seems particularly nasty. It is memory-hungry. It takes for-friggin-EVER to download an applet - only to find out it won't run properly if at all on a Power Mac 7100 with 28MB RAM. (Part of that may be Netscape 3) There are alternatives to Java, though - the most touted of which is Microsoft's ActiveX, which although fast, is not interpreted, but rather directly RUNS as OLE components on Windows, go ahead and figure out how to translate that to the Macintosh efficiently. Then there's Sun's own TCL/TK, a plug-in for which is available for Netscape Navigator - it has some nice features, but has one noteworthy flaw: it doesn't work. Call me back when you get to v2.0. Right now, the only workable multimedia option for Web pages seems to be Shockwave... at least until all the other big boys get the bugs squashed.

Which still leaves the Amiga in the cold. I suppose it would be possible to individually patch an existing non-Power Macintosh Netscape plug-in to work on an Amiga web browser designed to use them... but you'd REALLY have to know what you're doing to technically make it work, and you may be on the dangerous side of copyright in the process. For that matter, I've often wondered if it's possible to run, not a Macintosh emulator, but a Macintosh OS emulator, on an Amiga - intercepting Mac Toolbox calls and replacing them with Intuition functions, for instance, thus letting us run Netscape more-or-less natively on an Amiga without having to duplicate the guts of the entire Macintosh in the process. Just an idea... and no, I'm not volunteering to write such a thing.

That's all for now, guys - my lunch break is in overtime. Till next time, fellow cephalopods - keep them groovy roomurs coming!