The 68000 Wars, Part 1: Lorraine | Hacker News Reader

The 68000 Wars, Part 1: Lorraine

filfre.net

85 pointsmgunes45 comments

I think this bit was really interesting analysis:

"Amiga’s financial difficulties provided the opportunity of a lifetime to a bunch of folks that may have struggled to get in the door in even the most junior of positions at someplace like Apple, IBM, or Microsoft."

It reminds me a bit of the recent realization Google has had (after trolling through vast amount of employment data) they see virtually no difference in performance from people with high pedigrees that people without.

http://www.the-american-interest.com/2013/06/24/google-is-no...

It's a lesson that seems to keep getting relearned over and over again. Hire the individual, not the GPA, not the degree, not the school.

These guys produced an amazing machine virtually on a shoe-string.

It also comes up a lot in some circles that the Amiga was spiritually the follow-on to the Atari 8-bit computers, not the ST. A podcast I've been listening to recently has some wonderful interviews with some of the industry people from that time period and there's some great insight into this transition.

http://ataripodcast.libsyn.com/

Indeed. When I did interviews (and I didn't do a ton, but I did enough) I noticed zero correlation between having a CS degree and being able to code worth a damn. I noticed only a moderate correlation between having a degree from a "good" school and coding ability, but nowhere near enough to warrant the overwhelming degree of credentialism that exists in the industry today.

On a similar note, a lot of the best coders I've known have had degrees in non-CS fields (such as physics) or no degree at all.

The ST was a way to get access to a 32 bit address space without spending a very large amount of money.

The Amiga was the fore-runner of the multi-media PC.

It is kind of amazing how close the software looks and works in ports between the two systems, despite the radically different hardware. In many cases, the ST version ran faster.

And with a little hardware and software (Magic Hat IIR), you could turn your cheapo ST into a Macintosh.

32-bit address space, GUI, hi-rez screen, and good-enough games. In 1987 my choice was a Commodore 128 or an Atari ST -- I wanted 80 columns, modern graphics, and more memory and couldn't afford a PC or Mac (nor did I like the PC anyways). I was coming from a VIC-20. The ST fit the niche and lasted me until I went Linux on a 486 in 1992. The Amiga was another step up in the price tier, and less geared around productivity and more around entertainment.

If you were into video at the time, the Amiga was most definitely the more productive machine.

Oh for sure. But if you were into MIDI, the ST was by far the better choice. The Amiga sold more the ST in the end. But neither really made a dent in the market % wise.

Both had a lot of potential and were let down by bad management. Commodore I think did a marginally better job than Atari did in growing their platform. Atari had a last gasp of inspiration with the TT and Falcon and MultiTOS that seemed like they were making up for 5 years of neglect. But then they packed it in.

I still turn my Falcon on occasionally. It is a really nice machine.

Neither understood the fundamental truth that they had to be priced in the C64 range to stay alive. The C64 was replaced market share-wise by the IBM PC and neither company repeated the success of their earlier, cheaper machines.

They needed games and price to keep winning.

That was sort of Jack Tramiel's schtick. Cut-rate, but capable computers, moved in massive volume. It worked wonderfully for him at Commodore, and he really wanted to stick with that at Atari. Commoditization can be a powerful competitive advantage if you can pull it off. It worked brilliantly for the C64 and in Europe with the ZX Spectrum.

Here's a show from that era with an interview with Jack

https://archive.org/details/LowEndCo1985

PCs didn't become affordable for non-business uses until the early 90s. By then you could get a 486 or 386 for about $2000. When you think about inflation, that's a heck of a lot in today's money, but if you take into account that the original IBM PC was $1500 in 1981 dollars when it was introduced, that's a heck of a lot more cash. ($3,873.30 according to http://www.usinflationcalculator.com/) That was really out of the reach of middle/working class people.

I'm sorry, but I don't buy the inflation line when it comes to computers. As an industry, we continually say computer get cheaper every year. I look at the price brackets. We lost the under $100 and under $200 price brackets and are only now getting them back.

In the late 70's and early 80's, my family could get me a computer I could program for under $200 (well under $100 with the Sinclair). After that, I wouldn't have been able to be a programmer.

As an industry, we have failed our youth. We had no answer to uplift the low income like the opportunities the 8-bit computers gave us. Big box stores allowed non-technical parents to get a computer that could be programmed. Magazines had programming articles.

Yes, I agree! That was my point. We lost the affordable when _everything_ went PC/Mac. PCs were very expensive until the mid to late 90s. Even then they were pretty pricey, but not quite as pricey when you account for inflation (my argument above).

We are only getting the price down again now with the ARM revolution.

Anyways, I could build a pretty darn capable PC in the late 90s for about $800-$1000 which is about what an Atari ST would run you in 1986/87.

Oh, ok - I get it now. I'm just a little sensitive to inflation calculations where people say a $200 PC in 80's is just as affordable as a $500 PC now - which is bunk given people's situations and frankly demonstrated by the $99 price point for cellphones and other electronics.

I dream of getting rich enough to build a modern Sinclair / Atari / Commodore that could be sold in retail stores. Something that a kid can write a program and then share it with other kids without purchasing a developer account. One can dream.

Actually, IIRC, the games for STs were easier to write because the screen was not divided in bit planes but the pixels had all their bits together.

ST had interleaved bitplanes. They were actually in many ways a pain to work with. But later people found techniques for working with it that allowed them to do 'chunky' (https://en.wikipedia.org/wiki/Packed_pixel) type graphics to planar conversion quite quickly, and in a way the Amiga couldn't.

See:

http://alive.atari.org/alive8/c2p.php

24bit, at ~70%(?) true m68k speed, because CPU was constantly halted by video chip.

Address space size has nothing to do with access speed, but yes, the video chip and the cpu were competing for access to the ram. This happens in every graphic display interface that does not use dual-ported RAM or some kind of dedicated CPU doing the writing of the video RAM, you're always going to have contention.

The ST was a pretty clever compromise to reduce costs, unfortunately they weren't reduced far enough to give the machine a longer production run.

>Address space size has nothing to do with access speed

hence the comma

> happens in every graphic display interface that does not use dual-ported RAM or some kind of dedicated CPU doing the writing of the video RAM

Not quite what I had in mind. Atari (called 'Agricultural equipment' by Polish Commodore fanboys https://www.youtube.com/watch?v=oMkCboYJ-tw ) had to constantly halt cpu just to fetch pixels. This is on par with how Spectrum did video 8-9 years before it. Amiga on the other hand had a nice separate dedicated 'slow ram' where you were able to actively perform operations (think of it like a very primitive pixel shader - dma with xor/and/not/or on the fly) while CPU was unaffected (running full speed on separate 'fast ram'). Its not like it was a rocket science concept, this is how video was handled on MSX computers/X68000/Sega consoles.

Did I mention sprites? What sprites haha, scrolling? .....

Everything at the time had separate video memory bank, everything but the farming Atari ST. Yes, it was the cheapest m68K machine on the market, but it only made it terrible.

The CPU isn't halted - it's denied access to the bus. A lot of the time, this doesn't matter, which is presumably why they chose to do it that way, because most 68000 instructions won't be accessing the bus at the time anyway. (Supposedly, if you have a long instruction that doesn't use the bus, such as DIVx or MULx, and you get your timing right, you can overlap it with the blitter, which on the ST always runs in some approximation of nasty mode - better you than me though.)

More info (than you probably ever cared to know) here: http://pasti.fxatari.com/68kdocs/AtariSTCycleCounting.html

> It also comes up a lot in some circles that the Amiga was spiritually the follow-on to the Atari 8-bit computers, not the ST.

The Amiga offloads a lot of functionality to the custom chips, much like Atari 8-bit machines (ANTIC is an incredible piece of hardware design). I liked to compare the ST to the minimalism of the Apple II, but I agree, it's more like the C64.

I think of the Atari ST as being much more like the ZX Spectrum. "Minimalism" is a polite way of describing it: you have a CPU and you have an area of memory (with a slightly bizarre layout) that represents what will appear on screen. Like the Spectrum, the CPU is nice enough to code for, but takes its time when doing anything remotely useful.

(Maybe similar to Apple II as well. Very uncommon in the UK - never used it.)

Of course, the ST has a few things going for it compared to the Spectrum: configurable display RAM address, minimally contended RAM, timer interrupts, quite apart from the increase in colours and pixels and the properly bitmapped display. But overall, it's the same sort of thing, to be used in the same sort of way. You'd approach coding for the C64 or Amiga very differently.

Yeah it's funny, the Amiga is kind of a spiritual successor to the the A8, and the ST is kind of a spiritual successor to the Commodore 64 (lead designer on the ST, Shiraz Shivji, was a designer on the C64).

Given the shift in ownership of Atari, it does make sense. Atari was basically the old Commodore owners.

This is a good honest account of the Amiga, I'm glad to see somebody to admit its faults and quirks rather than blindly sing glorious praise to it as is often the case with its former (or present) users.

I was an Atari ST user back then, the "other 68000 home computer". When comparing the ST and Amiga it is easy to get caught up in hardware capability comparisons, but the ST was 2/3rds the price of an Amiga, and shipped with more RAM and didn't have the interlaced video. There was nothing like the ST's hi-rez monochrome monitor (very good quality) for the Amiga. Really the ST was like a budget Mac, but with a higher clock rate, higher monochrome screen resolution, and disk format compatibility with MS-DOS. Here's a similar accounting of the first genesis of the ST: http://www.atarimagazines.com/startv3n1/threeyearsofst.html

The Amiga was a great piece of hardware, and the games for it were generally the best in class. And it was made by some really inspired people. But it only took a few years and that hardware advantage meant little anymore. Mainstream Macs and PCs absorbed the multimedia capabilities. Game consoles from Sega and Sony were able to outplay it, and there were even machines like the Sharp X68000 that blew it away on the hardware front by the late 80s (and even outsold it, albeit only sold in Japan.)

You are right that the advantages the Atari ST and Amiga had were absorbed into mainstream consoles and computers, but I think the time gap is a little longer than a few years. The Amiga released in 1985, but the Sega Genesis/Masterdrive didn't come to the US until 1989, and didn't arrive in the UK until 1990. Windows 3.1 was 1992, first color Mac was 1993, and Sony PSX was 1994.

The first colour-capable Mac was the Mac II in 1987.

I'm glad you brought up the X68000 truly a remarkable machine that's forgotten when people talk fondly about the 68k machines.

I agree with the other comments here that other machines didn't quickly catch up to the Amiga. I remember growing up with a couple friends who had Amigas in the late 80s, early 90s. My parents ended up buying a PC for home because of work stuff sometime in 90-91 and it didn't hold a candle to the Amigas my friends had.

I had to add a Sound Blaster to even get audio (digital sounds were all computed by the CPU so 4ch sound and Amiga quality video almost brought it to its knees) and it really wasn't until I upgraded to an SB Pro and a decent graphics card (this was before 3dfx) that I could say audio and visuals were capable of what an Amiga could do out of the box. But then it took forever for the software ecosystem to catch up. There were popular games coming out at that time that were still EGA and quite a few with CGA and Mono graphics.

Even then, most of my friends had modded and hacked on the A500s to get them faster processors, more memory etc and they were easily as capable as my PC.

I don't think it was really until the mid90s where PCs started to comfortably pull ahead in both hardware and software. I remember getting a GUS MAX and maxing out the memory on it, and finally knowing for sure that I had audio that was clearly superior to anything the Amiga could pull off and it really wasn't until Doom that you could say that a PC could finally do something an Amiga couldn't. By then we were in 1993/1994, 486s were finally becoming consumer level chips, and high-end systems were running 486DX2s or DX4s and clock multipliers were finally putting the final nail in the Amiga's coffin.

But I was on my 3rd or 4th computer by then and my friend were still on their 5-6 year old A500s (though heavily modded by this point). One of my friends toyed around with getting an A4000, but stories of incompatibilities and other issues scared them away and eventually they just ended up with a PC like everybody else. I think the final nail was when Apple switched from Motorola to PowerPC and it became apparent to everybody that the 680x0 line was a technological dead-end and with it the Amiga.

But 5-6 years was an eternity in hardware in those days. If you use Moore's law (18 months) as a generational guide, the Amiga was 3 or 4 generations ahead of the rest of the PC market in terms of hardware and software capability. There's really nothing like that today that's comparable.

I was a huge Amiga fan as a teen, but I can still remember the moment I saw that PC gaming was where the action was going to be. It was ACE Magazine issue 35, August 1990.[1]

When I saw the previews of Wing Commander (Called "Wingleader" in the magazine) and 256 color adventure games like "Rise of the Dragon" [2] I was completely floored. This was the future of gaming. However, I couldn't afford a PC until 1996.

The end of the home computer era was a difficult time for me and my circle of working class nerds. The price of computing shot up massively. As a troubled teenager who staked my future on being a self-taught programmer, this changing-of-the-era came at the worst possible time.

One of the most valuable things about the Amiga (And the Atari ST, in the case of one of my friends) was that kids like me could learn about programming, 3d modelling, music production and productivity software all with a £399 machine plugged into a 70s TV set. My whole life as a programmer and indie game developer is thanks to that machine.

[1]https://archive.org/details/ace-magazine-35 [2] See "movies in your micro" in the above magazine

Yes, this.

I lucked out in that my parents took some of the money set aside and bought my sister and I 486 machines when I graduated from highschool in 1992. This was a significant chunk of change, but I held onto that 486 for 6 or 7 years, running Linux on it. The next machine I got I had to take out a bank loan for, even though I was actually working as a programmer by that point.

It's nice to see now that the up and coming generation has access -- if they want it -- to very cheap but usable machines. Sub $100 ARM machines you can hook up to a TV and a keyboard, maybe sitting in the same niche as those C64s, Spectrums, Atari 8-bits and 68k machines did back then. The difference now is few people are really playing nuts and bolts with their machines like we had to back then.

It's hard to convey the feeling when I first saw an Amiga. In the office, surrounded by monochrome early DOS machines someone had set up an Amiga 1000 showing the golden sarcophagus of Tutankhamen. It's like someone had switched on an electric light for the first time.

This was exactly my experience. I was at college, walking to the bookstore to buy a 286. There was a guy from Creative Computers with table set up demoing an Amiga. I couldn't believe what I was seeing - it looked so much better than anything else at the time. I bought one the same day, and struggled to find a decent word processor for the next 4 years. But the games made it worth it.

Still reading this, just wanted to thank you for posting this! I owned an Amiga back in the day. I wasn't old at all, I was only just barely getting my toes wet in programming, but I knew I was using something special, and when I look back and see an Amiga 500, I still think it is. It's even pretty despite the design's age. Or maybe it's my pretty memories of it speaking.

Making me even more nostalgic about this era, is that it felt like the golden age of computing, so many wildly different concepts. My ZX Spectrum 48K, my Atari 520 STE, then my Amiga 500 Plus. It felt like anything was possible, that anyone had a chance of building a famous new computer with its own operating system and all. And all that felt on countless LAN's. Man, those were the days!

Ah, those were the days. I learned Assembly on A500. You would just load an image into memory, tell the video card to point to that location and it got shown on the screen.

I recall writing a simple program to extract images from games. After reboot, the game graphics would remain in RAM. Using the program I would change the address with keyboard controls until I got something that looked like an image from game on my screen and then I would save it into a file.

And I got docs for programming it via a friend who had access to some BBS.

Kids these days don't know what they have.

Jay Miner was one heck of a designer. I do wish there was a book that focused on the forgotten hardware designers of that era.

Ah, the Amiga -- computing's Lamborghini Countach. Sexy, powerful, revolutionary engineering for its time. Fueler of many a pubescent schoolboy's wet dreams. But also hopelessly stuck in the 80s in terms of design and fiddly enough to put off non-enthusiasts.

As someone who grew up on the Commodore 64 and Amiga, I sorely miss the 68k platform.

The few times I'm still forced to deal with the X86-architecture at assembly-level it just pains me to see the clusterfuck that modern the Intel-platform represents.

The 68k was clean, nice, logical and let you work on the data you wanted to process. The X86 looks like a patchwork of instructions and registers added cumulatively over time, with no design behind them, and the majority of code is moving data in and out of registers, not actually processing the data.

The X86-architecture marked the end for when I cared about knowing how things worked all through the stack. It was just too ugly for me to embrace.

I got my Amiga as a freshman in high school. It was so much better than anything else out at the time, yet never caught on.

Is there anything like that happening today, where there is something significantly better than its competitors, but has no traction. I don't mean something debatable (e.g., Windows Phone vs iPhone or MacOS vs Windows), but something clear cut?

Probably the closest thing I could think of is Tesla, if they go belly up.

Amiga was a major success, it did catch on after the jury-rigged A1000 was put out to pasture. It was Commodore that squandered it, by milking it financially and stagnating technologically.

I wish Acorn Archimedes had had half the opportunity Amiga got.

The Archimedes had processing power by the bucket compared to the 68K machines of the time but it lacked the fantastic software/hardware integration that the Amiga offered.

I'm pretty sure the Amiga couldn't have run Zarch at anything like the smoothness that the Archimedes managed.

And the difference was purely down to the stripped down Acorn RISC Machine chip running at 8 MHz, compared with the 8 MHz 68000 CISC chip.

There are so many if-only's about the Amiga. If only it could have managed 640x480 un-interlaced, and if-only someone with even a modicum of taste had worked on the Workbench GUI. If only they'd used a decent disk drive, like Apple were forced to on the Mac.

Commodore's pivot to selling to Business was doomed when selling a games machine, though the cost-pared A500 was very successful as a gaming computer in (non-UK) Europe.

Never caught on ? No traction ?

That certainly wasn't true in Europe. It seemed like every kid around me in the early 90s had an Amiga 500, a few years after they all seemed to have an Amstrad CPC (wildly popular in France, even though it was a British computer. The Spectrum was the UK darling, and the C64 was the more popular one everywhere else)

A generation of game publishers (mainly in the UK and France) thrived on the Amiga and went on to become household names today (UbiSoft comes to mind)

As the article notes, there were even Amiga magazines on UK newsstands as far as Y2K, years after Commodore's demise.

So yeah, pretty big deal over here.

In some ways, the Amiga reminds me of Techno music : invented in the US, but found its true home in Europe.

I had an AMIGA 500 in the early 90s. Sadly, I only ever used it to play video games, got bored eventually and then sold it. I knew several people who did, too.

I remember part of the reason to sell it was classmates mocking me for having a machine that was only good for playing video games, while they owned real computers, i.e. IBM compatible PCs. In retrospect this seems kind of funny, given that they usually had 286 machines with EGA graphics cards...

Yes, in Europe it did catch on. In the US we used to say, "See how popular it is in Europe!" to our naysayers. Unfortunately, it never got traction in the US.

In the comments of the article some pointed to a Kickstarter in progress right now for an event at the Computer History Museum in Mountain view to celebrate the upcoming 30th anniversary of the Amiga.

https://www.kickstarter.com/projects/890300835/amiga-30th-an...

I got an Amiga in 7th grade! It was a hugeee step up from my previous computer (an Apple II.)

It felt truly amazing for the time (late 80's).

Reply on news.ycombinator.com