June 23, 1996 VISCorp Joins the Fray
Here's a quickie that I just had to say something about: This is from Viscorp.
"It has come to our attention that several companies plan to build their own "compatible" or "extended" versions of the Amiga without obtaining the proper licensing from VIScorp. These companies will be placing themselves at legal risk, because their systems will undoubtedly infringe on Amiga intellectual property rights, including copyrights, patents, and trade secrets."
They have to be talking about Phase 5 and PIOS - because I can't think of who else they'd be referring to.

So here's my comments:

Read on:
"In the months ahead, VIScorp will be making substantial improvements to the Amiga system architecture, including both the hardware design and the operating system software. These advanced new systems will be developed not only for our upcoming Set Top Boxes, but also for future Amiga Desk Top Computers. VIScorp is investing considerable resources into engineering these improvements."
Heh heh. Characteristically vague. They've been bullied into making this statement, I think - after two months this is the first time someone from Viscorp has used the word "desktop" and "Amiga" in the same sentence. And as one observer on Usenet put it - "They're recycling old Commodore press releases." Now, when I read Viscorp's statements, I don't see a company with its pulse on the future, or a company who cares about the Amiga, or a company on the verge of pushing the computing envelope yet again. With Amiga Technologies, at least, two months after the buyout they'd hired Peter Kittel and begun whispering about RISC Amigas. With Phase 5, we're already drooling and arguing over which nonexistentPower Amiga will be better - Phase 5's or PIOS'. So here comes Viscorp, late to the party, seeing their thunder about to be stolen by companies with an edge, while they themselves really didn't want (and can't) make anything more advanced than an AGA channel-changer, decide they have to at least poke their head into the crowd and remind everyone who really holds the Amiga purse-strings these days, and say something to the effect of "We're making bigger and better thAmigas, but we can't tell you any specifics because PIOS hired our R&D department. But while you wait for us to track down some development staff and try to get the Amiga back to where it was before the buyout crap started, remember we're in charge and don't try to do anything without our permission."

For me, I think, the final proof of Viscorp's appraisal of the situation is the way in which they capitalize "Desk Top" into two words. All this time we tried to tell them we wanted support for our computers and bigger computers just like them, not channel-changers - and when they finally publicly acknowledge it, they think "desktop" is some kind of buzzword.

Is it too late for the Amiga community to gang up and just buy the fucking rights to the Amiga and take control of it once and for all?


June 21, 1996 Emulator Roundup
Got 15 minutes to type while lunch break lasts...

First up, there is now apparently a new software-only Emplant (guess Jim Drew got tired of hearing about how cool Shapeshifter is) - and there's a 1200-specific version of it that seems to use the PCMCIA hardware and other "undocumented features" of the 1200 to do nifty things. Better still, it will accept a wider variety of Mac ROMs, including that of the IIcx like this one we have here at work, the one with the fried SCSI chip. (Ever notice how similar 1989-era Macintoshes are to the Amiga 3000?) So... I'm gonna be doing some investigating into this particular niftiness, if it does what it says it does (unlike Emplant 586) I may just get it... and have my very own Mac at home so I can install all the nifty hacks like Aaron that they won't let me use on the workstation here.

More importantly though, is something called UAE. UAE = Unix Amiga Emulator. Yes, it has finally been done - Workbench booting on a Linux/X11 system. Better still, the emulator has been ported to the PC (ever see "guru meditation" on a Pentium 120?) the Macintosh (Shapeshifter-in-reverse?) and - multiple orgasms - the BeBox. Yes - that's right - boot Workbench on your Sparc10. Boot Workbench on your AIX box. Boot Workbench and run those old A500 demos on your Power Mac 9500. Boot Turrican on your Pentium60. The downsides, obviously it's dog slow, on a P120 it runs literally at about the speed of an A500, the sprite emulation is still a little buggy (but no program but AmigaBasic used sprite collisions anyway - leave it to Microsoft), and obviously no floppy controller manufactured except Paula herself can read Amiga disks, so you ave to use "diskfiles". The upside is the damn thing actually works, and people really ARE running Workbench 3.1 on their Pentiums. No substitute for an A4000T/060 under your desk, but...

In other news, Phase 5 is now selling "recycled" accelerators - recycled and fully tested 40MHz 68040s on Blizzard and Cyberstorms, starting at 599DM, which I think comes out to something like $450 US. (Tell me that's wrong.) Also, the Blizzard T/ERC series marks another first, with the Blizzard 1240T/ERC - yes, you guessed it, the 12 means it's a 1200 card, and the 40 means a 68040, FULL with MMU and FPU, at 40MHz. Of course, there's a reason nobody's done this before - between the heat and the size of the 68040, it won't fit in the 1200's belly-bay. The official stuff from Phase 5 describes it as an expansion board for the "A1200T" - yes, your 1200 has to get naked before this thing will work, that case has to come OFF and the nekkid motherboard either repackaged or left in its birthday suit. But for $450... the money you save can go towards plywood, for a DIY (destroy-it-yourself) tower case. :-) (Why did they never make computers with woodgrain cases? Stereos, VCRs, etc... but never computers.)

That's about it for this lunch break. I gotta go do some programming... oh wait, that's something to mention. Why is there no Code Warrior for the Amiga? Code Warrior Gold 7 will target its executables to Macintosh 68K-series, Power Macintosh, Windows 32-bit, and believe it or not, BeOS DR7... but not Amiga. CodeWarrior makes it so much easier to write cross-platform applications, but we, the third-largest platform in the world, behind Macintosh and ahead of Unix, are in the cold. Perhaps they're waiting for RISC Amigas... but then, so are we.


June 19, 1996 My first day in lights
Whoa! My Amiga page has gone from the least-hit part of my page (five hits a week) to the most-hit part of my page (a couple hundred hits a day) just by virtue of being on "New Amiga Links." A special thanks to Champaign-Urbana Commodore User's Group for including me, David Tiberio and AmiCrawler, of course, all those of you who've visited to make this thing such a success! And, of course, all you within the computer industry for giving me something to rumor-monger about. :-)

Now, on to the first order of business: answering my statements about Pios.

First off, I probably came off a lot harsher-sounding on Dave Haynie than I intended to be. (That's called being too much like the Bandito - sacred cows make damn good hamburger.) I probably said too much based on too little information - but then, any longtime Amiga pundit will be used to that kind of behavior, not as if it's excusable.

Anyway, prior to the last two days' e-mail, I had not even heard of this rumored $500 Pios box. From what I've been able to piece together (you who've emailed me are now honorary members of the Rumor Mill Spy Network - smile and be proud!) Pios is planning an ultra-low-cost PPC-based Amiga-ish thing, with some kind of off-the-shelf-but-way-cool-anyway 3-D chipset, essentially what the 1200 wanted to be but for less. More details I don't really know off-hand - there is NOTHING on www.pios.de, you'd think a technology company like Pios is at least capable of building a webpage. But then, since I work for a web development company that could theoretically build them a bigger and better one, I probably ought to be more tactful in my criticisms of them. :-) (I don't have an agenda, do I?)

As to the custom/standard chip set arguments, I'm just backing off altogether. This is 1996. The only difference between a custom chipset and a standard chipset is the size of the company who makes it. Would anyone here consider the UltraSPARC a "custom chip" on the order of Agnus? Even if you do, it will cease to be a "custom chip" when another chipmaker decides to clone it. If Phase 5 comes up with a way-cool custom 3-D monster-chipset that's ahead of it's time today, yes, it'll be custom and nonstandard - but if it catches on, as we're hoping it will, as Phase 5 is betting it will, other companies will like it, and either a) try to license it, or b) try to clone it.

USENET is livid with discussions on all this. An Amiga in operating-system only is kinda at a disadvantage without software - most of the Big Boys (Softwood and Digita being the exceptions) have jumped ship - so now we're competing with the BeBox on the high end of the non-mainstream scale and with the PlayStation and Pippin at the low end of the non-mainstream scale. A PPCP AmigaOS is gonna be WAY COOL for those of us who already know and love the system - we who will be using PPCP hardware at work will love being able to multiboot AmigaOS and do our Photoshop work on PowerShapeshifter, while running the _real_ fun stuff preemptively on the Amiga side - but will it sell any _new_ people on the Amiga?

For the Amiga to have any identity in the marketplace it needs something to stand it above the operating system crowd. It's preemptive, but so is Unix. It's easy to use, but so's the Mac. It's technically two steps behind the BeBox, at least for now - those two steps being resource management/memory protection and multiprocessing. Even Copland (aka Macintosh 1998) will probably have an edge over the Amiga in some category or another.

And when PPCP finally hits the marketplace, one OS will be much like another.

The spy network tells me that Dave himself is investigating some kind of PCI-based pseudo-custom stuff, probably to put a 3-D pipeline of some kind into one of the Pios boxes. The $500 box will likely not be PPCP - but will have something custom under the hood.

Which leaves us asking two things: why, then, does Dave disapprove of Phase 5's decision to go with custom hardware, and will the Phase 5 Amiga run the same PowerAmiga binaries as the Pios box.

Of course, while this is all going on, Viscorp is still silent and nonhelpful as ever... "Will you make Amiga desktops?" "Uh... yadda yadda yadda... mumble mumble mumble... but we're gonna make these REAL COOL set-top boxes!" "So are you?" "I answered your question already." "No you didn't." "Don't get rude with me! I'll have to hit you with a set-top box."

Viscorp does not yet own the Amiga, due diligence and all. They're still making sure $40 million worth of Amiga guts and stuff are really there to be bought. As with the original Commodore purchase in 1995, it takes awhile.

Meanwhile, BeBoxes are now shipping to developers again, and will apparently be shipping to REAL CUSTOMERS in July. I think they cost about $2300 or so for the basic unit, twin 90MHz 603's with 16MB, CD-ROM, gig HD, and BeOS DR7 and Code Warrior. I know I won't be buying one unless I win that McDonalds Monopoly thing, but it's still nice to dream...

To be continued... (gotta run!) (Remind me to tell you about Mindmeld sometime. Oh, and... UAE. And some Ibrowse and POV3b problems...)


June 17, 1996 Minor goodies
My spy network here at the Rumor Mill has learned a few nifty tidbits. First off, we've heard a possible explanation for why Phase 5 has been so late shipping 68060 cards for the 1200 to the United States. The rumor? It's not been FCC approved. This puts the 68060 Amigas next to the TRS-80 Model I and the C128 as units that had a helluva time going through the FCC. Meanwhile, P5 is supposedly redesigning the 1260 boards, which is consistent with the story of FCC nonapproval - just quietly rework a few things.

In other news, A1200 supplies have FINALLY filled up the two-year backlog in Europe. What this means for us American Amigans is EXCELLENT news: the boys in Germany have decided to start converting the excess warehouse A1200 stock to NTSC - and we will see Amiga 1200s on dealer shelves in the United States beginning July 4. (Cue Star Spangled Banner) Units will probably hit the ground at ridiculous prices in the $800 and up range - but that A1200 will do more things faster than an $800 Macintosh.

And speaking of Macintosh, Apple is planning to bring out a "hybrid" operating system later this year, sort of a System 7.5/OS 8 bastard child, entitled "harmony" - that will incorporate a few of MacOS8 (nee Copland)'s niftier features. Apple is apparently surveying users to decide which Copland feature(s) are most important to add to "Harmony" - and when I filled out that survey, the box I checked was "improved stability." And if I recalled, I saw a Type 11 not long afterward. You stability freaks will be happy to know that Copland will COMPLETELY do away with extensions and control panels - no more inits or cdevs AT ALL which means no more extension conflicts or startup shuffles or Extension Manager or left-shift-key or frantic calls to 1-800-SOS-my-macintosh-is-crashing or techtool or Conflict Catcher or any of that other crap.

Back to the Amiga... sort of. We've learned some more about PIOS. Apparently Dave Haynie has lost his marbles completely. (Perhaps the death of Iggy, his cat, immortalized in the Workbench 3 manuals as the example picture for the printer settings, was too much for him) Pios is not an Amiga company as we first thought, as we were probably led to believe. What PIOS will be making at first is NOT an Amiga clone, or even a custom Amiga box (a la Phase 5) - but as we feared, a completely standard PPCP computer that just happens to be able to host the Amiga operating system alongside MacOS and Windows NT and Solaris and AIX (rhymes with aches, for good reason). Now, it's on one hand, a good thing to be able to run AmigaOS on non-Amiga hardware - it ensures longevity of the OS, even after our old Commodore boxes crumble into dust, not to mention cheaper prices as PPCP becomes cheaper. But it's something else to just port AmigaOS to this standard platform, stand it up alongside MacOS, Windows NT, and every UNIX variant on the planet, and expect users who already have Mac or Windows or Unix software to make the "intelligent" choice. There HAS to be something there besides the OS. Yes, the OS is what makes the Amiga niftier than the rest - but it's an OS with relatively little software (unless you count Aminet).

This is what Phase 5 is trying to provide, by giving us a visionary computer with kick-ass hardware at silly prices, a la the 1985 Amiga. But as time has passed since al these announcements, it's become clear that Pios doesn't quiet understand this, at least if Dave is to be believed. Dave makes the case that custom hardware costs more in the short-term and will be obsolete in 5 years. Folks, your bought-in-bulk Cirrus Logic and ATI and Diamond Stealth Super-Duper-VGA chipsets will be obsolete in 5 years. Already the Picasso II (based on a Cirrus Logic chipset) is considered ancient - and it's 1993 tech. Yes, a custom chipset will be obsolete in 5 yeaars - that's why you build another custom chipset within 5 years. It's what Commodore would/should have done - they wanted a 256-color chipset done by 1987, and actually delivered it in 1992.

The point is this. The oft-cited example is the Commodore 128. Some would say the best 8-bit computer ever made (but I prefer the Color Computer 3 myself) - but very little software came out to take advantageof all the souped-up new features, because the developers knew it would run C64 software just fine. So they kept writing C64 software and sold to both the C64 and C128 markets at once. Our fear is that a PPCP Amiga will be unappreciated - developers will KNOW that if they write their software for ANY of the PPCP OSes, any PPCP user (with the appropriate OSes installed) will be able to use it - so they'll write for the Big Boys (WinNT, maybe Mac) and ignore the rest.

Yes, a PPCP Amiga is a necessity. (lunch break is coming to a close, so I have to write this quickly) But perhaps more important is to give people something that REQUIRES the Amiga OS - something awesome. Some kind of killer app. Software? Nah - some Mac or PC developer will clone it within a week. Hardware. The Video Toaster did its magic in hardware, not software. A souped-up Amiga (like Phase 5's dream machine) with a complete set of video tools inbuilt will draw people to the multimedia stuff first, and then gently lower them into the ecstasy that is AmigaOS. Later, the world realizes "oh wait, this is a cool OS, can I use this on PPCP?" and bada-bing.

Last, Babylon 5 has been renewed for its fourth season. Speaking of excellence...


May 25, 1996 Point - Counterpoint
It's happened. The Amiga community is fragmented. What the pundits predicted when the A2000 came out ("if it doesn't have a keyboard garage, it's not an Amiga"), what the pundits predicted when the A3000 came out ("the people with classic Amigas won't talk to the ones with new Amigas"), what the pundits predicted when AGA came out ("who wants a non-AGA box these days?"), what the pundits predicted when Commodore fell ("Amiga companies won't want to work together anymore") and what the pundits predicted when ESCOM took up the reins ("is there animosity between Escom and CEI?") has finally happened.

The players? Phase 5, PIOS, and Amiga Technologies itself.

The playing field: It's now common knowledge that Amiga Technologies is not the juggernaught it wanted us to think. Amitech has been plagued with financial problems and lack of vision since sometime last year, and the recent layoff of the R&D staff and imminent buyout by Viscorp are just the tip of a red-and-white checkered iceberg. Power Amiga development at Amiga Tech didn't start until real late last year - and proceeded slowly up until last month's layoffs, and stopped again. The impending Viscorp buyout only served to complicate things, while we wait to find out what the inventor of CDTV is going to do with our future.

It is in this environment that Phase 5 announced their "power project" - y'know, a $2000 PPC box with a 3-D custom graphics chipset and Amiga-compatible OS. It is in this environment that Stefan Domeyer and many of the laid-off Amitech staff, along with Dave "A3000+" Haynie and Andy "layers.library" Finkel, formed the new company PIOS, with the intention of building Amiga PPCP boxes and owning 10% of the market share by 1999.

But here's where it gets cool. Earlier this week, Brian Sorli interviewed Dave Haynie the Hardware God and got this from him:

"I have e-talked with a few of the folks I know at VIScorp, but they couldn't really say much. The pios strategy is really counting on the AmigaOS, since that's the enabling OS for low end multimedia systems (pios may offer other OSs on their standard PPCP machine, just because that's easy to do and may let us have product before the PowerAmiga OS is done). I'm hoping VIScorp is open to collaboration on this. The one thing we definitely don't want is a fragmented AmigaOS -- there has to be one API, one binary standard, etc.

(...)

...AT did take too long to start on the Power Amiga stuff (it didn't even start happening until late last November). I suspect everyone at AT was a victim of ESCOM's money problems, even back then (though we didn't know about them). I also think there's a little animosity on the part of phase 5 -- I think they were expecting more an embrace of their stuff from AT than they got. They do, apparently, have a simple way of running an AmigaOS emulator on a PPC with native graphics support, though they didn't have anything ready to show last February. Since Andy and I were advising AT, we insisted in doing things right. Their approach on the software front is kind of a hack, and on the hardware front it's just too much like the old Commodore; at best, they'll wind up with interesting, non-standard, and overpriced machines that can't keep up with the rapid changes in the industry. Computing in the 90s is radically different than it was in the 80s or 70s. Things have just gotten more complicated. Chips are far more complex to design, so you need larger volumes to make them practical at all, and they might last only 1/2-1/4 as long in the market as their equivalents of 10 years ago. IC processes have gotten exponentially more expensive, to the point that only the top IC makers in the world have state-of-the-art fabs, and even these guys are having to get together on the next generation fabs."

Everyone expected more from Amiga Technologies than we got. A year later, we have overpriced A4000Ts, no American 1200s, cheap "souped-up Mosaic" web browsers, and a prototype computer that looks like the head of a penis. Meanwhile, Dave - as always - is trying to do things "the right way" - he knows when to hack and when not to hack. But on the other hand, Dave also once said IBM would be out of business by now. If Phase 5 is at fault for trying to design "nonstandard" hardware, then so is Intel for making the Pentium Pro slower with 16-bit code, AMD and Sun for incorporating multimedia DSP-like instructions into their processors, and any company shipping either DSP cards or accelerated graphics cards. Frankly, a souped-up Phase 5 chipset would be a hot multimedia item - and people would be designing new chips to incorporate those features.

Meanwhile, running 68030 and 68040 code on a RISC processor, while allowing the RISC chip access to the 68000 parts of the operating system, is by definition a hack. And a nasty one too. Look at the Power Mac. Finder on a 6100 is slower than Finder on a Quadra.

But anyway:

"The real question I have about phase 5's plan, all practical bits aside, is this: will they sell you a Power Amiga, or a PhasePower system that, oh-by-the-way runs Amiga 68K binaries. They're talking about writing their own "advanced" OS as the native PPC part, and they're letting the 68K emulator hook into some bits, like their graphics subsystem. That's not the same thing as a real AmigaOS port, and I don't think the Amiga community can support multiple OSs -- an Amiga runs the AmigaOS, and only the owner of that (presumable VIScorp) gets to say just what that is and isn't.

Not to mention the fact that it takes a long time to write a good OS. I know of three new OSs written by experienced OS people (rare, in these days of Microsoft dominance): Scala's MMOS, the 3DO OS, and the BeOS. Scala and 3DO took each took over three years for their OSs, and neither was intended as a general purpose personal computer OS. Be's OS work started five years ago, and they're still clearly in the beta testing phase (with some modules not out of alpha yet). Either phase 5 is adpoting some other, unnamed OS, stealing the AmigaOS, offering up an ugly hack, or they're not shipping any new OS in 1997."

Ouch! Probably true, of course, OSes take awhile to write. Odd, in this day and age, when we can write things like Doom, we can do cognitive programming, neural nets, spline-based intelligent 3-D rendering and the like, and it still takes years and years to write a multitasking operating system that works. Hell, Microsoft, the biggest company in the world, has tried and can't do it. Granted, Be switched processors halfway through the design phase - they started out on ARM if I recall and then switched to PPC.

My own guess is that Phase 5 is going to just rewrite more and more of the OS into PPC code as time goes on. My big question is, sure, you can run Amiga 68040 binaries on either the PIOS box or the Phase 5 box, but what about Power Amiga binaries? Do you have a forward-expandable RISC OS with a 68040LC emulator, a 68040 OS with a RISC API, or a 68040 OS with RISC-accelerated code running entirely within emulation? FWIW, Apple chose the second - emulating pieces of the OS, so the RISC code accesses 68040 library calls and vice versa. Windows NT uses the first approach running on RISC systems - it's a RISC operating system, but all your x86 code runs as a task inside an emulated virtual machine. Right now, the third scenario is all the Power Amiga can possibly be at this early stage - good old Commodore Kickstart V42 with exec.library, graphics.library, and expansion.library reworked to use PPC native functions, and the rest emulated. Essentially the Power Amiga prototypes sitting on a desk at Phase 5 are little more than A3000s with Cybervision cards, with really really really fast DSPs. When the emulator is working, it'll still be just that- an emulator with a few special functions accelerated. How they plan to handle PPC binaries is not clear - and whether PIOS PPC binaries and Phase 5 PPC binaries are cross-compatible is also not clear.

But then there's this, which came across the wire today, from Wolf Dietrich of Phase 5:

"With interest I have read Dave's comments on the current Amiga situation. As we and the Amiga are in a situation where a lot of major decisions for the future of this system are to come, I would like to answer some of his comments to reflect our position and partial different view of things.

First of all, there was no animosity on our part against AT - we just wanted to get things going, and so we started the PowerUP project in late 95 as AT wasn't giving any view or commitment. We had been offering AT all of our support for quite a long time, and continued to do so in a situation where there was no development, no resources, no vision; what Dave believes to be an animosity was simply the great concern that things wouldn't go into the right direction for the Amiga. Meanwhile, all development on AT side is cancelled, so there is no more cooperation as there is nothing left to cooperate in."

Uh-huh. Told ya. In Dave's defense, I think he just used the wrong word - he didn't mean animosity, he meant frustration. And a lot of us share that frustration - things at Amitech were moving at Spindler-era Apple speeds, before stopping altogether. SMG, in their own weird way, blew the whistle on a lot of this.
"But let's get into some technical considerations. First of all, Dave states that our software development is kind of a hack. Funny to hear that, as he has not seen a single line of code, and also was not involved in in-depth discussions about what we are doing. To simply state the facts: For our PowerUp program, which's goal is to develop PPC upgrade boards for existing Amiga systems, we have re-written Exec and Expansion in PPC Natice Code, and two versions of 68k emulators to run the rest of the OS out of the system ROM. This is not a kind of a hack, but simple the first step which we could realize. Our plan to add a PPC native version of CyberGraphX - which has emerged as a standard today - is just a software add-on to increase the performance of those upgraded systems where parts of the OS have to be emulated. This way is not very different from, for example, having a 68040 or a 68060 library to emulate in software what is different in the processor hardware - however, as 68k and PPC have some significant differences, it can not simply be realized by a library or a new setpatch, but needs a completely re-written, but fully function-compatible Exec."
Ah-ha! Approach #2, on the surface: RISC API with mostly 68000 OS. Now, I could go on about the trickiness of writing a RISC exec.library that's 68000 compatible, knowing what little I know about how AmigaOS libraries work, you'd probably need two sets of library calls in there, one each for PPC and 68000, or else you'd have to switch to 68000 emulation every time your Power-Accelerated POV-Ray needed a library call, which slows things down (again, look at the Mac).

I especially love how he calls the difference between high-endian, 16-bit-instruction set, complex instruction set 68000 code, and multiple-endian, reduced instruction set PowerPC code "some significant differences." I'll say - there is not one 68000 instruction that can be directly executed on a PPC without some kind of translation - even if it's simply a change of byte code. And it's still a "hack" - ugly or otherwise. It's the nature of the beast.

But here we go:

"N ow getting to the comments on the hardware design. First of all, I leave it up to the public to judge if our announced systems are overpriced (see http://www.phase.de/ in the news section); also the non-standard argument is missing any fundamental. I know from the technology meetings which AT, Motorola and phase5 joined in the recent months, that Dave's vision of a new computer is a standard PPRP mainboard, with a PPC CPU and a PCI bus and that's it; any idea of adding something specific which would have to be developed had been rejected by him in these discussions. But, all innovative developments today contain some individual parts, mostly in form of FPGAs or ASICs; it's the only way to build something that stands out of the mass markets. We at phase 5 definitely believe that a new Amiga system needs some unique H/W features as it had in the past; just having a ported OS running on a standard PPRP system which also runs MacOS, WindowsNT, and so on, would be the death of AmigaOS simple as there would be no sufficient reason for S/W developers to continue writing their code for Amiga OS. But even if Amiga OS would survive for some time with some application or shareware support, it would be the death of the Vision Amiga which never had been just another PC (no matter if there is a PPC or a Pentium inside).

As Dave's comments on the rapid changes of the industry are concerned: We know these rapid changes, we live - succesfully - in this world. Are chips more complex to design today? Chips are more complex, but sophisticated design tools, powerful design workstations, and comprehensive functionality libraries are available today for ASIC designers. Today it's possible to start *VERY* complex designs on FPGA basis and go to the more expensive silicon in certain stages of the development. IC processes are *NOT* exponentially more expensive than some years ago, at least not for companies who want to do custom designs and get strongest support from various ASIC suppliers in the world, among them the very big names such as Motorola. Yes, even those big ones go together to build new fabs for the next millennium, but were we talking about building a next generation IC fab?"

Whoa! That hurt.

Dave used to be a big proponent of custom hardware. But that was before AAA fell behind the times. Now, he's apparently hung up believing any custom design is not the way to go. What about things like - off the top of my head - DSPs? Multimedia processors? High-speed ASICs for things like serial and parallel and keyboard? Graphics accelerators? Hard drive controllers? Memory interleave controllers? Network chips? PC Card controllers? Internal modem circuitry?

Perhaps it's just a matter of knowing how weak Amiga Tech - and now Viscorp, and probably PIOS - would be in IC development. Yes, Dave's right - you could build a PPCP system from off the shelf parts and run a Power Amiga OS on it - and yes, it'd be niftier than a modern Amiga - but then you cease to be a computer company and become an OS company, much like what the pundits say Apple is destined for. I rather like the notion myself of having ultra-souped-up hardware in the next new Amiga design - stuff that will make the Amiga truly ahead of the curve again.

The argument for or against a "standardized" Amiga hardware platform - PPCP - has been going back and forth for years, even in the Commodore days, when the first RISC Amigas were on the drawing boards, they intended to build a standardized - or standard compatible - system that could host multiple OSes, not just Amiga, but things like Windows NT, on a Hewlett-Packard PA-RISC core. Pleasance and company from Commodore UK also had something similar in mind - an Amiga box that could run Windows NT. And while no one will dispute the value of being able to run the Amiga OS on several kinds of hardware, the notion of Amiga hardware as something totally standard, off the shelf, and conservative, is a little bizarre - and that's what Phase 5 is fighting.

But the questions remain unanswered: will the Phase 5 Amiga run the same RISC binaries as the PIOS Amiga? Bingo - a fragmented market. The fact that Wolf and Dave have just pissed each other off means there will probably not be a single RISC API - two Amiga operating systems, one definitely PPCP and one maybe-maybenot PPCP. Will we get to run PPCP operating systems on that nifty Phase 5 box? Solaris, for instance, with special drivers for the cool hardware extensions? Will developers have to release two different PowerAmiga executables, one for PIOS and one for Phase 5 API? Or will one of the two systems fall by the wayside? Each potentially has its advantages - PIOS would run on any PPCP box, while Phase 5 OS would have the cool multimedia chipset.

In any case, by all indications, Phase 5 will have product first.

In other news, a very weird rumor from Viscorp's developers meeting in France. Unofficially, just word on the streets, is that Viscorp prefers Digital's Alpha processors to the PowerPC. And following that rumor is another, weirder one, that Viscorp wants to ship a 300MHz Alpha-based Amiga for under $4,000. Now, if true, if $4,000 Alpha boxes hit the streets, I have on the highest authority that we at Columbus PBX/Compassworks would replace all our servers and workstations with Amigas... as I'm sure many other companies would too. That's what makes me think it ain't true - it sounds too good to be true, which means it probably is.

And by the way, IBrowse is now up to beta 7... and no, this one's not anything to write home about, it fixes some minor validation bugs and it sends you to the funny red blinking screen less often than before.

In other news, Linux is now available in beta for Power Macintosh. More precisely, Linux is now ported to the OSF Mach microkernal, which exists on PowerPC. In any case, Apple has officially sanctioned this, and you can freely download the fifty-megabyte archive directly from Apple's own FTP servers. We have a copy of this at CPBX, and I'll probably be reporting on it as we play with it - but it's worth noting that some home-grown benchmarks have placed a Linux Mac 6100/66 at a tenth the overall speed of a Linux 486DX2/66... maybe as they fine tune things it'll get faster.

But in any case, you can finally run a pre-emptive multitasking OS on a Macintosh. :-)


May 22, 1996 Take two.

And following hot on the tail of Phase 5's stunning announcement, is Pios. Who is Pios? Stefan Domeyer, former Amiga Technologies president of r&d, is Pios. John Smith, former Commodore UK sales manager, is Pios. Andy Finkel, former Commodore software wizard extraordinaire, is Pios. Dave Haynie, whom I think we know, is Pios.

Pios, basically, is all those people who got sick and tired of waiting for Amiga Technologies to do what they said they were going to do. Several of Pios' employees are victims of the recent Amiga Technologies R&D layoffs - one can't build a RISC Amiga with no R&D staff.

Now, Pios hasn't flat-out said "Here is what our Amiga will look like," the way Phase 5 did. But Pios probably will gain more mindshare among Amiga supporters, due to the names of its employees. VISCorp has Carl Sassenrath, but Pios has Andy Finkel, Software Wizard Extraordinaire. VISCorp has the guy who made CDTV, Pios has the guy who made the A3000 and A4000, A2620/2630, Nyx, the 3000+, the Scarab card, and of course, DiskSalv. A lot of Amiga folk listen when Dave talks, and let's face it, if Dave says Amiga Technologies isn't getting it done, the Amiga community will probably believe him. He's our own Guy Kawasaki, at least while his BeBox is still in the shop.

Pios' mission statement is to become 10% of the computer market by the year 2000. Can they do it? I don't even think the Macintosh is yet 10% of the computer market. But given the right marketing, you can sell anything - though considering their web page, that could be a tall order.

In other news, Aminet has just hit the 30,000 file mark - making it the LARGEST archive of freely distributable software in the world, by far the largest for any platform. I told you Aminet was the biggest and best software archive in the world, but you didn't believe me! Why didn't you believe me? Aminet is now bigger than Simtel and bigger than InfoMac.

[Articles] [Links] [Buyout Watch] [Personalities] [Workbench] [Unsolved Mysteries] [Ideas] [About Squid]

[John's Homepage] [Sarah McLachlan Stuff] [Donna Lewis] [Cabinet of Curiosities]
[About John] [John's Art] [Email John] [Guestbook]