Where the weird things are

November 25, 1996

Some flashes from a correspondent (my spies always remain anonymous) who was at Cologne:

"First off, VIScorp were originally going to buy AT with stock, not cash. This is obviously a great thing to do. It costs you nothing except on paper and it's easy to arrange. Petro Tyschtchenko told me that he doesn't believe any bank will loan them the money needed for a company that has gone bust twice (he was nearly in tears of frustration when we talked)."

And no, it doesn't look like Viscorp has the financial strength to pull this thing off. Yes, they're still in negotiations, things are still moving... but both animals in this tug of war (VISCorp and ESCOM) are dying piece by piece while the buyout goes on. More on this later.

"Secondly, how do like this rumour: QuickPak seem to be in the running for buying the machine out from under VIScorp... my contact was being very cryptic about another company that wanted to buy the machine. He was talking about them wanting to do a kickback deal (like with mobile phones) with ISPs where punters who wanted to get on the net but didn't have a computer would get an Amiga laden with ready to run internet software and they would have to pay a premium for the first couple of years and then the machine would be theirs..."

An Amiga "laden with ready to run internet software" - hmmm. Already I can hear the "MUI is cool/MUI sucks" flamewars echoing down to us from the future. Amiga Internet software is either MUI or ClassAct - nobody seems to use straight Intuition and Gadtools anymore - and is either polished (Miami, Termite) or UNIX-like and cryptic (AmiTCP, AS225). Ready to run Internet software is a Holy Grail no matter what platform you're on. And keep in mind, despite all the ballyhoo about Amigas multitasking in 256K on a floppy drive, an Amiga Web browser - even ALynx - does not run on said 256K machine. AmiTCP alone wants a 2MB system with Kickstart 2.0 or better - and is happiest with a hard drive. Miami wants MUI - which wants Fast RAM, a 68020 or better, and demands a hard drive. And most Amiga Web browsers - certainly anything more advanced than AMosaic or ALynx - want AGA, a 68020 or better, 2MB chip, 2MB fast, a hard drive, and kickstart 3.0, absolute baseline, nothing less unless you like sleeping in the glove compartment.

"He also talked about this mystery company building a new range of machines (with hardware serial numbers to stop piracy) based around the 68k series for low-end machines and DEC Alphas for high end (and with the low-end Alpha 21164 going for $3000 to manufacturers that would make this amiga pretty bloody high end). The reason these people were talking to my contact is that they want to incorporate a product he makes into all amigas together with either an internal DSP or modem. The way I found out it was QuickPak was from another source..."

Couple things here: yes, the serial number thing is a good idea, but keep in mind on an Amiga you can do k00l things, like, for example, using the MMU to "project" a different serial number to running software. Serial numbers would not STOP piracy - only slow it down - and then it's a matter of time before the pirates remove the 'serial number protection' from the code altogether, the way pirates removed "dongle protection" from Brilliance, LightRave, and several others. Good idea, though - hardware serial numbers have a million and one uses, personalized shareware registrations that enforce use on one and only one machine, tracking stolen computers, personalized security encryption (messages decode ONLY on your home Amiga and nowhere else), etc.

Alphas do run in the several-grand range, but let's also keep in mind a little thing called the Exponential X704. X704s will cost about $1000 a chip, are completely PowerPC 604 compatible, and start at 466MHz and go up. The only issue is heat dissipation - they run like a light bulb at 75 to 85 W, and thus require an extra CPU fan and some thought given to ventilation. The X704 right now uses 0.5-micron feature geometry (the Alphas use 0.35 and 0.25 micron), and is what Byte calls a rather conservative chip design. It's nowhere near pushing the limits. Expect 1GHz 704s in a few years. The chips are using bipolar-on-CMOS technology (a rather neat trick - the CMOS is mostly used just for cache), and is on the same order of complexity as the Intel 80486 (about one and a half million transistors). That's room for growth. The reason I'm rooting for this particular chip, for high-end Amiga RISC systems, is that it's got what looks like a growth path going way uphill. I've never written Alpha or PowerPC assembly language, though, so someone else can tell me which one's the better chip to code for. Another advantage to PowerPC or Exponential is that it makes Mac emulation much easier - but that's another story.

One last goodie from that correspondent:

"Oooh, I forgot to tell you. Petro confirmed that *one* of the things that AT have done is to put the AGA chipset onto one CMOS chip. It works just fine, but other people have told me that it has problems (linked to processor speed apparently). Now that could seriously help with cost reduction! Also more news on the surfer box. The machine was supposed to ship with chip ram on board but no processor - you choose the speed you want with an accelerator card. Also rumoured at the show (but later denied by phase 5) a cybervision for the A1200. Phase 5's answer was - get a tower and a CV64/3D..."

Cost reduction, although definitely a possibility with a single-chip CMOS AGA chipset, wasn't my first thought.

Think about this for a moment. The Amiga chip set has been NMOS - not CMOS - since its birth way back in Reagan's first term. Now, I'm not sure how much you know about electronics, but back when Motorola and others made most of their chips in NMOS, they usually built (or had others build) CMOS versions of their chips. The Motorola MC6809 and Hitachi HC6309. The Motorola 68000 and 68HC000. The Intel 8088 and the NEC V20. Why?

CMOS runs faster.

CMOS draws less power.

Now, due to the timing schemes inside AGA, CMOS might not necessarily help speed (though that may be the "problems" my contact mentioned), but think about the second advantage for a moment:

The biggest hurdle to Amiga laptops, other than screen mode conversions, is power utilization. AGA runs hot and sucks juice. Ask Silent Paw. They have those gigantic Duracell batteries that are a foot long, that get sucked dry inside an hour.

Although I trust my source, and my source trusts Petro, I have a teensy bit of trouble believing they actually did this. Several ex-Commodore guys have mentioned how hard it would be to revamp AGA into CMOS - there are things you can do in NMOS that you have to do completely different in CMOS. And considering that the OCS masks had disappeared by the time Commodore started on AGA, I'd suspect similar things happened to the AGA masks - I seem to remember someone saying the AGA chip images had disappeared before ESCOM bought everything.

And I've already mentioned the idea of a processor-less Amiga with an expansion card slot - an A1200 with no 68020, sell it with an accelerator and you're no longer buying two processors. (Either that, or work on an SMP system so the 68020 can still run low-priority tasks in the background...)

It sucks that Phase 5 aren't making an A1200 Cybervision. Admittedly such a card would either fit only in towers or displace an accelerator - unless they figured out how to squeeze a processor, some RAM, a SCSI controller, and an SVGA chip all with support logic onto an A1200 expansion card. My Microbotics 1230XA is already quite crowded - and the Blizzard 1240 and 1260 units are absolutely packed, not to mention kick out a lot of heat. Perhaps it's safer for all our plastic cases that it was just a rumor.

So let's go look at a war a little closer to the States. Here's Jason Compton:

"I'm glad to be back home, but I very much enjoyed my trip to Cologne, Germany and the Computer '96 show. The two days spent there flew by very quickly, but I was glad to meet with hundreds of Amiga users, journalists, and professionals to discuss the Amiga.

The topic on everyone's mind, of course [the buyout]. At present, there is no publicly consumable news to report. However, these things are true.

1. VIScorp remains dedicated to the acquisition of the Amiga technology from the bankrupt Escom AG, and our representatives continue discussions with Escom trustee Dr. Hembach over this issue.

2. VIScorp remains dedicated to the productive application of the Amiga technology in the event it is successful in the acquisition--namely, the fulfillment of the interactive technology goals the company was founded on, and the support and further development of the Amiga technology as a desktop computer. The interactive technology goals will explicitly involve technology partners, and we anticipate forging strong partnerships for the Amiga desktop computer technology as well."

It's worth mentioning here that Carl Sassenrath probably violated his non-disclosure agreement by saying what he said. But it's something that needed to be said - a situation where the truth would affect more people than just Viscorp.

But it's also worth mentioning that Viscorp blew off everything Carl and others did to build that set-top box the company was founded to deliver.

Yes, anything Viscorp does from here on in will "explicitly involve technology partners" - they're too far down the drain to do anything themselves anymore.

There is a chance - a small one but a chance anyway - that the "technology partners" mentioned could be people like PIOS, Phase 5, ProDAD, Macrosystems, and others.

"VIScorp has dedicated a great deal of corporate energy and effort to the acquisition. It is not VIScorp's intention to deprive, abuse, or alienate the Amiga market. Rather, VIScorp wants to complete the Amiga acqusition as soon as possible and be freed to pursue its technology goals. It is not in our interests nor our intention to mislead the Amiga market, nor to misrepresent the current state of affairs."

"Also, it is explicitly not in our interests to in any way damage or endanger our efforts in the Amiga acquisition. In order to satisfy all of these requirements, VIScorp is obliged to abide by certain rules of conduct which preclude a "blow-by-blow" report of the proceedings of the Amiga acquisition. This would be improper and beyond our ability to provide. We remain committed to the acquisition of the technology and as such are attempting to act in everyone's best interest. We cannot make commitments to property we do not yet own in the form of purchase orders, contracts, and the like."

But I'll also grant that ESCOM showed up at an auction in April 1995 and had Amigas rolling off assembly lines in September. They may not have done anything else right, but they had that first 1200 motherboard quicker than I would have figured.

"It IS within our purview to conduct preliminary discussions with companies regarding the ultimate goals for the Amiga technology, and as such discussions have been conducted worldwide, with such companies as Motorola, PIOS, Phase5, and Schatztruhe. However, it would be premature and improper to make any commitments, nor rule out any other discussions, at this time."

Here's hoping they add Exponential to that list of companies soon.

Acquiring the rights from a bankruptcy to properties obtained, from a bankruptcy a short while ago, which represent the intellectual property of a decades-old computer giant, with the two sales conducted under different codes of law, is not a straightforward affair, as some might suggest. VIScorp has done its best to learn and understand the full implications of the purchase of the Amiga technology and is working to successfully complete the acquisition.

As for the "full implications of the purchase of the Amiga technology" - two things:

1) You will get flamed constantly from unhappy Amiga people whose A1000 won't run Netscape and it's your fault.
2) You will go out of business in April 1997.

What I don't understand is why this is taking so much longer than the initial sale to ESCOM. I know Jason can't say anything, but it's FUCKING NOVEMBER 1996! ESCOM owns the Amiga name brand, the Commodore name brand and logo, the motherboard designs for all existing Amigas, the source code to the Amiga OS and other Commodore software, the casework for the A1200 and others, warehouses full of Amigas, all the documentation, the C64, the C128, all the various PETs and Plus-4's and other 8-bits, the chip images and tapes for OCS, ECS, AGA, and AAA, and the rights to manufacture and sell products based on those technologies and trademarks in many countries around the world. I'm admittedly very naive about these things - but even in Germany, shouldn't it be analogous to writing out a bill of materials (so many A1200s, so many A4000s, some books, some chip masks, some stacks of Intuition source code, some spare chips and cases, some legal documentation saying "We can sell this in this country," and the word Amiga) and hand it over in exchange for a check?

I don't know. You tell me.

As a general policy, VIScorp does not engage in online flame wars. It is a negative-sum enterprise and not in the best interests of a company and the individuals in it who are dedicated to a goal and wish to conduct business towards that goal.

Case in point: Jason's response to the Sassenrath letter: "I don't know, I don't run his life." The bluntest, rudest, cockiest, stupidest piece of PR since SMG's "Phoenix" bullshit. Bad, bad form. Almost as dumb as something I might say.

Yes, VIScorp continues to work with all diligence towards the completion of the Amiga deal, but don't let anyone fool you into thinking that it should be an "easier" process. As has been mentioned before, this sort of process is not merely closed to VIScorp. VIScorp believes it is the front-runner in this acquisition. If it were any "easier" to purchase the Amiga, by now we or someone else would have made it so. The fact that it is November 22, 1996 and we do not have control of the Amiga assets as of yet is a consequence of the way business is and must be conducted in an international sale of technology with major implications.

This isn't the Bahamas. This isn't Commodore. This is ONE PIECE of ESCOM's pie, a piece that ESCOM completely owns, and a piece that is NOT (to my knowledge) already tied up by creditors worldwide as Commodore was. It WILL BE soon if someone (if not Viscorp, then SOMEONE) doesn't get off their ass and make this thing happen. The trustees of the Commodore liquidation were pissed because they didn't get much money - never mind they'd gone MONTHS and MONTHS taking vacation after vacation. Somewhere, someone in Germany at some bank or lawyer's office is wanking at his desk when he should be signing paperwork. And this isn't some office supply company with a warehouse full of staplers - computer technology decays like food in a grocery store, as newer and faster things come out, such that every day the sale doesn't take place is a certain number of dollars and cents lopped off the top of the value of the word "Amiga." VISCorp understands this - because they themselves are losing value at the same rate. ESCOM understands this - they're the ones in bankruptcy. But there's someone somewhere who doesn't - or else this would have been done by September 19 as promised.

Due to my trip, there is some e-mail that has gone unanswered, and I'll be working to rectify that situation. While VIScorp has been open from the start to accepting and considering comments and input from all comers, please try to keep in mind that you and you alone determine through your language how you wish it to be received.

Sorry - just felt the Evil Spirit of USENET come upon me for a moment. :-)

Well, as if that wasn't fun enough - all that and not a SINGLE attempt to refute a single thing Carl Sassenrath said. Consider that confirmation. Flame on, Viscorp. We're ready for you.

Speaking of Carl Sassenrath, here's the latest from him:

"I have received a large volume of email asking what will now happen to the Amiga. Even as one of its designers, I cannot offer you a solid, positive answer. There are many paths from here. We are again back at the crossroads. To me personally, I think the road must be toward a rebirth of those core concepts that made the Amiga what it was, and I encourage all of you to hold on, regardless of how desparate the situation may seem.

Please, my friends, remain the "rebels" that you are. It is up to each of us to see to it that there always remains a choice, else I fear for our future... and the future of all personal computing."

"We are again back at the crossroads" - I HATE THE CROSSROADS! I just want the car to GO! The crossroads sucks - because every time the Amiga gets to the crossroads, some asshole has to stop the car to go pee, while all the semis and motorcycles go blazing past us.

I'm sick of stop-and-go traffic. The Amiga is like my car - when you pull up to a stop sign, you keep your foot on the gas or else the engine dies.

"A lot of people have asked what I will do next. In general, I can say that I am NOT pleased with the direction personal computers have taken over the last decade. These systems are now hundreds of times faster, with hundreds of times more memory. Yet, are we better off? Are we more productive? Have they made life easier?

No. Not in the least. In some areas they've made the situation worse. It reminds me of our federal government: requiring ever more money and resources, but becoming bigger, slower, and dumber."

"My dog ate it" has given way to "The computer crashed." And teachers BELIEVE IT - because the test they were supposed to give that day also got lost in a system crash.

Modern laptops work for two, maybe three hours on a sixty dollar nickel-metal-hydride battery, while the old Tandy 100 and 102 worked for maybe six hours on eight AA's. Set your guy with the Tandy 102 and ROM-based word processor (and 40-by-8 text LCD screen) next to your guy with the Pentium 75 and Microsoft Word (and 800-by-600 color active matrix LCD) on the flight to Detroit and see which one actually gets the most work done - before the battery runs out, that is.

That 386/25 isn't getting slower every year... is it? No. It just seems slower.

This is not a hardware problem. The software is to blame. We have become enslaved by overly complex and twisted software. I'm talking here about systems like MS Windows, C++, TCP/IP, HTML, etc. They are all junk-the work of designers who really don't understand how our minds work or relate to our personal human condition.

As for the rest... speaks for itself. C++ still requires you to mung around with pointers and raw addresses instead of just GIVING you the data you asked for. TCP/IP doesn't even relate to how the DESIGNERS work - hence the current needs for secure IP and a bigger IP address space, amongst other things. (Remember the Internet was supposed to be just an experiment and prototype - NOT the actual "global network.") And Microsoft Windows... is.

SIMPLE THINGS SHOULD BE SIMPLE TO DO! Yet, complex things should still be possible. That is the kind of system I want to use, and that is what I am going to build next. After creating the Amiga multitasking kernel, I didn't stop thinking about operating systems, I just figured that someone else was going to move us forward to the next best thing. It has not happened.

Want to see what's wrong with today's user interfaces? Watch Star Trek IV, where Scotty sits down at a Mac Plus and can't figure it out. (That was originally going to be an Amiga, by the way, but for some reason Commodore botched up that product-placement deal.) He talks to it - and it doesn't respond. He really can't figure out the mouse. The urban legends of people talking about their "4X coffee holders" and "foot pedals" are true - because people really can't figure these things out. And admit it - you'd never have thought of double-clicking unless someone had shown you. (And conversely, watch newbies using Netscape double-clicking links and the Back button and wondering why they aren't where they wanted to go.)

Computers are logical. They are not intuitive. Once you understand the "rules" - a computer is the easiest thing in the world. The "rules" are tough. I'm running five applications and the Finder on this Macintosh I'm writing this on - and yet I see people who can't figure out why everything disappeared when they clicked outside the window (bringing another application to the front). I know how "command pipelines" work - and often, as many UNIX people do, solve complex problems with a "simple" (to me) command pipeline that sends novices out of the room screaming. I know how computer file systems work - and why Windows can have an icon in one place and the program it goes with in another. I know how the Internet works - and why http:// works and http:/ doesn't. I know how e-mail works - and have to explain it to people who complain because their e-mail address is so hard to remember. I know how computers work - and how they don't. But this is no help for the one who has their file locked deep inside some beige box that keeps barfing things like "Device LPT1 is full - try deleting some files" at them - all while that board meeting is in two minutes.

(My favorite: a Macintosh that got unhappy and warned the user that a certain disk might be damaged if the user continued. The disk in question was a CD-ROM. The Mac had no way to know.)

So now it has come time to start afresh. However, this time, I'm going to create it for you and for me, not for some company whose CEO just wants to get rich. It will be wonderfully elegant, pure, efficient, and simple.

I have already begun. The system will develop in two phases. The first is a better means of expression, the second, a better means of operation. If you want more info in the months ahead, link www.sassenrath.com into your web browser site list and visit often.

I think I'm beginning to like this guy.

So that is my journey, my future. I expect it to be a rough road at times, but this is the path I must take. It is my sincere hope that many of you Amigans will join me in the years ahead... with a system for the next century.

Stay in touch and all the best to you in the years ahead.

"Modern" OSes like BeOS and Spring have all sorts of core features like memory protection, preemptive multitasking, and object-oriented design. There's multithreading. There's OS-level network support. There's nifty command lines. There's even an OS-level relational database (in BeOS)!

But not one, NOT ONE of these "modern" systems has a "Recent Applications" menu.

Not the Amiga. Not Windows 95. Not Be. Not UNIX. Not Plan 9. Not Mach. Not Photon. Not Inferno. Not OS9. Not Merlin. Not Windows NT.

Was it technically easier for Apple to put it into the Mac because of cooperative multitasking? Was it simplified because of the forked file system? Was it the standardization of launching applications?

Perhaps.

But the real reason it was done? The real reason a Mac keeps track of things you've done lately?

Because Apple thought it was a good idea.

Not because it was easy. Not because it was hard. Because it was a good idea.

Oh, it's possible enough to add it to any other system - on the Amiga you'd hook into the file-open routines in Kickstart and watch for things, maybe some extra code to separate executables from documents from folders from files in ENV:.

But by looking at the reams of code in the Mac Finder, at the cold hard bits and bytes and pointers and registers and stack frames and MMU tables and task switching mechanisms and subroutines and Toolbox calls and vectors and interrupts, would you come up with the idea of keeping a list of the last ten applications? No. And that's the point. It's not a technical approach at all. It's not needed. It's not particularly simple. But it was done anyway because it's a good idea.

Let me propose a similar "tough-but-good" idea: an UNDO command in Shell.

Think about it. Hammer it around in your head - it IS POSSIBLE. Why has no one done it?

Because it's not a priority.

Never mind every user of a command-line interface since command lines were invented has wished for an UNDO command. Not an Undelete, but a true generic UNDO that fixes whatever it was you did last, even if you don't remember what it was - like the "undo" menu item in any decent modern piece of software.

Obviously I don't expect someone to write a new OS and have its major selling point be the "UNDO button" in the command line window. Truth told, I don't think a 21st century OS even needs a command line - command lines are for geeks like me who're supposed to be making ourselves obsolete by designing systems advanced enough that they don't need me around to fix. The idea is to give you as much power as possible - and while a CLI is nice, as the Mac proves by its lack of same, a properly constructed UI should do everything you'll ever need it to do without ever making you wish for a command line. The Mac fails at this - look how often we have to dig in and rebuild desktops, mess with extensions, zap PRAM, and even get in with ResEdit and mess with type/creator codes and tweaking the innards of programs. The Amiga has an excellent command line, but who but a dweeb would want it? (Me - but I'm a dweeb.) Does the command line help you write better fiction? No, unless your word processor blew up and you needed command-line tools to fix it. Does the command line make you a better programmer? Only if you're used to it. Does the command line make you a better artist? NO - who wants to paint by typing? Does the command line make you a better lover? Only if your USENET reader isn't graphical. Does the command line make your car go faster? No. Does the command line make you smarter? Perhaps - but it also causes obesity, depression, fits of anger, sleep deprivation, vision problems, carpal tunnel syndrome, and a general inability to get laid.

Banning CLIs isn't the answer - just a possible question to consider. Think in new directions. Don't think Enterprise ("computer?") - think Simon. Think Asteroids. Think television. Think Walkman. Think light switches. Think radio. Don't think bicycles or cars. Think food. Don't think sex. Think building blocks. Yes, the Enterprise computers would be better than what we have today - but if you'd never seen the show, would you, if placed aboard the Enterprise with no instruction manual, figure them out? Yes, your bike is easy - but remember how often you fell off it? Would you have figured out the gearshift in your car if someone hadn't explained it to you at some point? Even sex - yes it makes sense once you think about it, but it isn't totally intuitive that it works THAT way... But food, you can look at things and usually tell whether they're edible or not. (McDonalds is an exception.) Your Walkman has four buttons - stop, play, fast forward and rewind - but can deliver life-changing music if desired with a VERY small learning curve. Television - you turn on, tune in, and drop out. Simon - four buttons. Who cares if there's a CPU in there or not? Asteroids - same thing, who cares how many address lines it has? Radio - turn it up and away you go. If you need an instruction manual for any of these things, it's because someone has made it more complicated than it needs to be.

Yes, we do want more power over our toys - we want programmable VCRs, we want watch-one-channel-while-taping-another, but I personally don't see any reason why we need fiftygazillion buttons to accomplish this. A "watch one channel while taping another" button IS POSSIBLE. Same thing with computers - sure, hide the "print" command under a menu somewhere, pop up a whole bunch of stupid questions (HP LaserJet 5 on port LPT1: - how should I know which paper source?), all in time for the paper to jam - fuck that, why not a "PRINT" key on the keyboard? Tap that, it says "Print it this way (|) or this way (-)", "print color or black and white," "OK here we go" - ten seconds later your document spits out the front of the machine, color and in landscape, no drivers, no COM ports, no page setup crap, it just WORKS. Even the Mac cannot do this - although it does have a nifty Dogcow. And we won't talk about the Amiga's printer support - at least not while we still get jaggies when using outline fonts on a laser printer.

But you see my point (if indeed I had one). Carl's right - computer technology 1996 is different from computer technology 1986 only in the names of the players. Yes, Be has an interesting OS, from a technical perspective - but its user interface is sadly lacking, ugly, and incomplete. The Amiga has the best mix of technical OS and user interface - and you still have those damn "cycle gadgets" and a printer system that sucks and still no memory protection. The Mac has nifty usability features - localization, QuickDraw GX, "recent" menus, listview finder windows, automatic file typing, etc - but has cooperative multitasking, unprotected memory, and that damn "window is closed but application is still open" thing that frustrates new users and experts alike. Windows loses in all categories except maybe market share. All modern OSes act basically alike - the modern GUI still carries a steep learning curve for novices, so new users are still gonna have a fight no matter what brand of box they sit in front of.

So here's to Carl, wishing him the best of luck in his quest. He seeks the Holy Grail - a computer that's ubiquitous with an operating system that isn't.

Meanwhile, Viscorp seeks a paycheck that's ubiquitous and expenses that aren't.

*sigh* Back to reality. You know, just because I'm cynical doesn't mean the world doesn't really suck...