Amiga 2000 EATX PCB | Hacker News Reader

Amiga 2000 EATX PCB

github.com

137 pointsdoener55 comments

The Amiga was great back in the 80s, so far ahead of anything else. In the early 90s trouble arrived, and starting 1992, it was essentially technically dead since PCs had caught up. The newest AGA chipset on the Amiga 1200 and 4000 was a disaster and a nightmare to program (being planar rather than chunky) whilst VGA mode 13h was a dream come true. Nobody really bothered with extracting the most out of AGA resulting in lackluster games that looked terrible compared to what was coming out on the PC w/ VGA (nevermind SVGA later).

The Paula audio chip was essentially unchanged since the first Amiga (4 channels, 8bit samples) and was no match for the Soundblaster 16 (and later the GUS). All of that meant that for the first time, games that were coming out on the PC were way ahead of anything the Amiga had to offer and the momentum had fully shifted. Moreover, they were artistically novel and windows into the future.

Ultima Underworld, Ultima 7, Privateer, Day of the Tentacle, X-Wing, Wolfenstein, Doom, System Shock, Myst were all groundbreaking PC games released in 1992-1994 that never came out on any Amiga. Even games that did come out for both systems 1992 onwards, were usually a lot better on the PC (Dune 2, all Lucasfilm and Sierra adventures, Flashback, Another World, Eye of the Beholders, Syndicate, Theme Park). The Demoscene productions were also a lot better and forging new ground on the PC, AGA demos being limited by the c2p routines that were slow on the 020 and 030 Amigas.

It is still a great machine for hacker types though, nearly infinitely expandable and the OS is a joy to work with. It's just too bad that the models following the A3000 - the last great Amiga -, A1200 and A4000, were absolute disasters that killed the entire platform.

The A1200 and A4000 were essentially too little too late. Dave Haynie reported that there was an A4000 prototype ready in an A3000 chassis almost a year ahead of the A4000. Then there were the production ramp up issues with the A4000 where systems weren't available because chip manufacturing didn't happen due to money problems at Commodore.

There were so many missed oppourtunities. Technically, it would have been possible to ship and AGA machine in 1989 as DRAMs with faster access times were available then (even at an affordable price before the duties were applied), and it might have made a difference, but the march of the PC clones was already growing quite strong. Ah well, at least I had a lot of fun learning on those machines!

A lot of the retro experience is in the physical form. A matching keyboard would be an important addition.

I’m not really sure about going for the Amiga 2000 though… The 3000 and 4000 were much friendlier towards modern hardware, as well as much more capable out of the box. I could probably work from a 3000 running Amiga Unix (thanks, X11).

I know the keyboard is being worked on by someone else, not sure when that will be done though.

Someone's done the A4000 in an ATX form factor too https://www.retrosummit.com/2018/08/21/a4000tx-atx-amiga-mot...

sincere questions. what is it about the Amiga, in 2021, that still has such interest to warrant a project like this? Does it do something a modern Mac or PC can't? Does it have a unique ability that does something better?

Nostalgia is a hell of a drug.

Mostly it's nostalgia. It was such massive leaps ahead at the time.

But there's a lot of quirks that still stand out.

Some of my favorites include datatypes - closed source Amiga programs released 30+ years ago can load things like webp images if they support datatypes, as long as you drop a datatype into your system.

And AREXX. The language is pretty awful, but the pervasiveness of AREXX ports in Amiga applications and how normal users took advantage of it is something I haven't seen since.

The openness (despite the os being closed source) of the platform is another one. Hardware schematics. A composable, modular well documented OS where everything could be patched and replaced. Where people kept experimenting with new file systems because you could just drop a file in and it'd be accessible. Or new device drivers, including virtual ones. Or plugged in new system wide file requesters, because they could.

A lot of it has more to do with the community around it than the machines themselves.

It's definitely nostalgia for me. I still remember the day I got Shadow of the Beast for my Amiga 500, back in 1989. I was blown away by the graphics, especially parallax scrolling, and sound. I taught myself C on that machine (Lattice 5.x?) Some of the things I learned still are with me to this day. Before that I had only used BASIC.

I later upgraded to an Amiga 3000. That was my favorite machine of the early 90's era. Eventually, Linux started taking off and around 1994 I moved on from the Amiga.

I play around with the Amiga occasionally on various emulators. I also got a MIST box some years ago (which is an FPGA-based emulator.) IMO the FPGA stuff isn't worth it given the speed of emulation you can obtain even on a Raspberry Pi.

I think it is the fact that the Amiga platform was in advance on its time. A lot of amiga owners clinged to their Amiga waiting for the next revisions to eat the world. When the PC platform took over it was a step backward. All those projects are a way to rewrite history for the fans: see we could have had an amiga in pc form factor, see we could use the web, see we could have had graphic accelerators...

And the fact that today it is possible to do is a fun in itself.

Lots of nostalgia of course, but the Amiga was a glimpse into an alternative future where the computer was a productivity and creativity device, not primarily a media consumption device, while at the same time being friendly to beginners and intuitive to use. For instance the way GUI and command line worked together, and (starting with AmigaOS 2.0) applications could be wired up to exchange data and provide services to other applications in a consistent way across the whole system is something that went way beyond the UNIX command line and still is unmatched.

Switching from my beloved Amiga 3000 to Windows 95 was like being thrown back into the dark ages for a few years (apart from the games of course, Doom FTW!), until the rest of the world finally caught up (but not completely) with WinXP and OSX.

> For instance the way GUI and command line worked together, and (starting with AmigaOS 2.0) applications could be wired up to exchange data and provide services to other applications in a consistent way across the whole system is something that went way beyond the UNIX command line and still is unmatched.

Dbus actually took quite a bit of inspiration from Amiga REXX ports (via KDE DCOP), so we've actually gotten that back in recent years. But so many Amiga features are still lacking in modern OS's.

The problem for me with dbus is perceived complexity, and lack of community expectations.

Dbus tends to be treated as something developers might use, while AREXX was something regular users used, and expected applications to support.

The technical capability is there, but something is lacking. And that is the case for a lot of things I miss from the Amiga - it's technically easy to replicate datatypes for example.

What the Amiga has which more modern systems lack is:

* a stable - as in non-changing - platform from which to extract as much performance as possible by way of programming prowess instead of throwing a few more gigahertz/bytes at the problem

* a compact and rather elegant operating system which' state can be kept in the head of a single person, this makes it possible to reason your way through most problems

* the combination of the above created a thriving demo scene which, if they want to keep active, need access to compatible hardware so they can be sure their exploits can be demonstrated on "real" Amiga hardware

The same is true for e.g. the Commodore 64, the Sinclair ZX-Spectrum and a host of other popular systems. The Amiga was revolutionary at its time and as such attracted those who were looking for a machine to explore hence it gained a large following. While the absolute performance parameters fall in the dust compared to modern hardware [1] it still remains an impressive demonstration of what can be done with a relatively slow CPU combined with the custom circuitry and the OS which made the Amiga different from e.g. the Atari ST.

[1] pulling down from the top of the screen running some program to reveal the workbench (desktop) on an Amiga 500 (512K, 7MHz 68K CPU) preceded the Android notification shade (which Apple later copied into iOS) by a few decades, using hardware less powerful than what is integrated into the SIM card in that same device. On earlier versions of Android (1.x without hardware compositing, tested on a Qtek S200 which originally ran Windows Mobile) this was quite laggy...

For people interested in building/fixing computers these old systems are wonderful. Even without training you can teach youself enough about electronics to reason about how they work, well enough to fix them.

Sure much of it is nostalgia and those who owned Amigas in the 80’ and 90’ now have the time and funds to tinker.

I still wonder about the custom chip. Could you just send a handfuld to China and have them reversed engineer? Sure an FPGA is easier and cheaper, but many want real hardware. The custom chips are almost the only thing you can’t get as a new part.

It was before my time, but based on what I can read and see on YouTube, it strikes me what made the Amiga special in its day was its pile of custom chips that aided graphics, audio and kept assembly costs down by integrating tons of IO and glue logic into a chipset. Everything else seems to built down to cost keeping the overall system price from going into the stratosphere.

And while those custom chips are fine, they seemed almost targeted at sprite-based gaming and cost reduction. Again, I’m just looking at it through other peoples nostalgia, but it seems like it just wasn’t that remarkable of a machine for general purpose usage.

At the tine it was the only game in town for "general purpose" use in a sense in that everything else lacked applications for entire large subsections of use that could compete with what an Amiga could do out of the box.

E.g you could draw higher quality art on an Amiga than machines with far fewer colours or no bitmap graphics at all. You could compose music on an Amiga that was not achievable on any other computer in it's price class without extra peripherals.

And so on.

It's simply false to suggest the primary function of the custom chips was cost cutting - there was nothing that provided what they did when they were introduced. Making it cheap enough was certainly also critical, but making it cheap enough is irrelevant without making it possible first.

I think the problem with looking back at this without a very clear timeline is that things did move very fast. In '85 it was astounding and revolutionary. By '87 it started seeing some competition, and without considering that most of the competition was too expensive it starts looking less impressive. Then prices for PC cards kept dropping. By '91 it was getting dated, and Commodore was desperate to survive and get AGA and AA chipsets completed. By '93 it was all over.

In '88 the custom chips would have looked like just cost cutting if introduced then, but when they were introduced they were expensive and extravagant compared to what was on the market.

And by 2000, sound cards and 3D accelerators on the PC removed all the advantage, with BeOS looking like a possible replacement for the Amiga generation, oh well.

I guess those ideas now live on macOS and Windows platforms, to some extent.

It was when the A3000 came out that Macs started looking more attractive to me. Originally, it was like "4096 colors, cool!" but once high resolution screens became more common, flickering interlace mode and 16 colors was underwhelming.

16 bit color made HAM irrelevant and was more exciting than pre-emptive multitasking and graphic acceleration.

BeOS was almost acquired by Apple to replace the Mac OS, but it wasn't, and it makes you wonder how history would have been different.

Indeed, it would have been mostly C++ based, and not offer the scenario of buying Apple systems as pretty UNIX, as big alternative universe.

Ironically there was a group of engineers at Commodore that was big into trying to merge Amiga OS and UNIX.

"VCFMW 11 - Bil Herd: Tales From Inside Commodore"

https://youtu.be/-Zpv6u5vCJ4

I feel like we hit this point way before 2000. By 1994 you could buy a Pentium machine with CD-quality audio and SVGA graphics on a fast PCI bus. You could browse the web! Just 2 years later in 1996 you have 3D acceleration and sophisticated graphics APIs mainstreamed on PCs: not to mention the arrival of the Pentium Pro and MMX extensions.

My recollection is that the Amiga was 99% dead by 1995, at least in the UK, and by 2000 it was 100% gone.

The PC killed it off. Even before 3D accelerators came along, there'd been years of year-on-year performance improvements, each time at the same or lower prices thanks to competition between suppliers of commodity parts. You also had byte-per-pixel display memory, and much better ALU throughput than the Amiga, so the games would look more interesting even when they weren't the nice high production value stuff you'd now be able to get from US developers.

Doom was basically the end for Amiga gaming. There were still some noteworthy Amiga games after that, but it was the point where chunky graphics modes definitively were shown to be necessary to keep up.

It’s called Amiga Denial Syndrome which is an infectious disease amongst computer enthusiasts. The sufferers of this terrible condition genuinely believe that the Amiga still is the pinnacle of computing achievement and that enlightenment can be obtained by ramming the remains of an A1200 corpse with various accelerators and hack boards hanging off it into a PC case, squinting and pretending it’s the A4000 they couldn’t afford in the 90s. This has an elite subcultural element which provides completely new hardware like this A2000 and even new software. None of these people actually have a working Amiga for more than 5 minutes a month however so have to use a PC running Linux or a Mac (but never evil windows) to support their normal computer usage.

Note this is in jest; as a vintage computing enthusiast I have nothing but respect for this and may be speaking from experience :)

I sold my Amiga 600 in late 90's. I bought it in 1993. My previous computer was an Acorn Electron, which had, like 32k of usable RAM. I didn't get much use out of it. A friend said he has an Amiga, and it had 1M of RAM. I was astonished as to how much RAM that was.

I saw a new vid on YouTube recently with a guy showing how to use the Octamed tracker. By coincidence, the guy that bought my Amiga all those years ago bought it primarily for the tracker.

The Amiga is a surprisingly capable machine. I saw a demo where they booted up Debian on the Amiga. It was slow! It just goes to show how compact the AmigaOS is. The whole OS came on 3 single-sided (?) floppies, and one of them was for fonts. Amazing, a whole OS on less than 2 HD floppies.

It appears that AmigaOS is STILL being released commercially . The latest release was 6 months ago. Amazing, considering that Commodore died in 1994. There is AROS, a free version of AmigaOS, which is still actively development. Development seems to be slow, though.

AmigaOS is available for the MK68K and PowerPC. I saw a PowerPC version awhile ago. They're not cheap, though, which is a pity.

I really thought that an AmigaOS is combo with a Raspberry Pi would be an awesome hit. OS development on Pis has been a disappointment, actually. I was expecting more. Everyone just runs Linux.

Kids these days need 8G just to run Firefox, of course.

The Acorn Election was a budget computer from 1983, that's not really a fair comparison.

If you'd bought a typical Acorn in 1993, it would have come with either 1, 2 or 4 MiB RAM.

You can run the Acorn OS (RISC OS) on a Raspberry Pi.

Well, I'm not sure I see the point of anything beyond an emulator, (is it even slower than the hardware?) but the thing that sticks in my memory about the Amiga is how the Roguelike games tended to have graphical tiles when most ports on contemporary machines didn't. Specifically Hack and Moria, back in the day.

> It’s called Amiga Denial Syndrome which is an infectious disease amongst computer enthusiasts.

Is this related to the Reality Distortion Field?

We were the original fanboys (A1200 here).

Joking aside, they were amazing machines that people, somehow, became emotionally involved with.

Nostalgia for those of us that were there back then of course plays a big role, but knowledge is also important. Had Amiga won the race against the PC back then, today's personal computers would be the equivalent of flying cars. Not many people for example recalls that in the mid-80s we already AmigaOS software for creating GUIs (PowerWindows), that was years before the first Visual Basic was created.

Edit: I forgot that although the OS was fully multitasking, it lacked memory protection and management, so a rogue process could easily crash other processes or the entire system by writing parts of the memory that it shouldn't have touched. Also, there was no such thing as resource tracking, so if I malloc'ed say one kb, I had to remember to free it before exiting, or it would remain allocated until the next reboot. Those were limitations that I found immensely useful when learning to optimize things.

Many Amigans stuck with their aging hardware several years into the early 00's, and I don't think the last of the real die-hards gave upon Amiga Inc. until fairly recently.

It took dedication and many now use it for creating the ultimate retro computer.

Advances for its time is an understatement. In marketing, people are trying to sell you an experience, i.e. iOS is the experience of power and simplicity, Toyota is the experience of affordable reliability, etc. Well Amiga had almost no marketing, but anyone who used one seriously still have a fresh memory of that powerful experience of being in the future, making graphics only accessible (at the time) using 6 digits computer equipment, audio sampling and playback so easy a 12 year old could compose his own tunes, watching babylon 5 and trying to recreate space scenes in lightwave or other 3d software using only a few megs (few MB, as under 10!). Emulating a Quadra in software, faster than the actual Quadra using the same 68040 cpu, playing back video in real time when PCs were starting displaying colors, etc.

For some others, the demo scene, copy parties, gaming, BBSes, first coding, electronics projects, name it.

For me it was all of it, so yeah, powerful experiences that I don’t think I will ever see or live again because of the nature of where and how technology is heading. Last time I got excited with something with the same experience potential was when Oculus launched their kickstarter, received my DK1. Wow. New paradigm shift… but Facebook bought them and that was it for me after the CV1.

So we’re going back to our first love, nostalgic, stable environment and comforting :)

It doesn’t do stuff better than today’s machine, but the fact that its still usable in 2021 and that there’s still a lot of development being done to keep it fast and geeky shows you how powerful the experience was for all of us.

I’m happy I was alive and a kid at that specific timeframe, because being a kid today I would probably not have 80% of the tech skills I got now if it wasn’t for the Amiga, forcing me (as in fun) to understand everything low level and make me a better problem solver and expand from graphics, coding, electronics and creativity to cross-link disciplines.

Now the only last thing exciting I can see before dying is going to space, when prices are down by a notch. Kids today won’t live an Amiga experience, but will probably be able to go to Mars.

I’m glad I won’t be nostalgic over angry bird, an xbox or an iphone!

Now I feel the urge to play that video toaster vhs demo…again. lol.

Games of that era were better to play then and better to play now. Amiga is such an integral part of computing history even though the Mac/PC duopology tries to write it out of history.

When you consider in the early 90s that the Amiga emulated the fastest Mac FASTER than the fastest Mac, you get an idea of the power. Maybe you need to live that time to realize how pathetic today's technology landscape is.

>Maybe you need to live that time to realize how pathetic today's technology landscape is.

Okay, I'm a fan of the Amiga, but this is just too far. Today's computers are "pathetic"? I remember playing with an A500 when I was really young and didn't really know anything other than a few games. I really got into Amiga with the A2000 solely based the Newtek's Video Toaster. Then, computers became all about video for me since. The things we can do now in real-time with video blows an Amiga out of the water. Toaster was all about SD video. We now do more things in real-time with 4K videos.

Your comments act like the rest of the computing world is at a stand still compared to the Amiga of yore.

> Maybe you need to live that time to realize how pathetic today's technology landscape is.

I think that is going a little far. There are some incredible technologies out there right now. What GPUs can do these days is amazing. For another, the phenomenally fast SSDs in current gen consoles like the PS5 will trickle down to PCs. On the flip side of the coin, you have very capable microcontroller platforms based on chips like the ESP32 for a few dollars, not to mention the latest offerings from the Raspberry Pi foundation. Really low cost boards like these put the technology into the hands of anybody that might be interested in learning something with next to no investment required. Even the Amiga couldn't make that claim back then.

It felt like things advanced faster back then. The Amiga was a huge leap over other 16-bit platforms when it was released. Time doesn't stand still, and hardware wise the platform barely evolved between 1985 and 1994. AGA was the only major change to the platform, other than faster processors, which every other platform was also getting. Example: An Amiga in 1994 had the same sound chip ("Paula") as one released in 1985.

The slow, incremental improvements weren't enough to keep up with x86 and SVGA. Commodore really dropped the ball. If the 1200/4000 was released in 1990 instead of late '92 it might've been a different story. The 1200's performance was hobbled by lack of fast memory out of the box and a previous generation processor (68020, which as 5+ years old at the time.)

Early versions of the OS were also primitive and unstable. That didn't change until 2.0 which wasn't generally available until late 1990.

Absolutely, but even if Commodore had rolled out new updates to the chipset sooner, experience is that people wouldn't really have used it because everything in that era was so tightly coupled to hardware -- with minimal abstraction and API [part of the joy them, tho] -- that people had to write to the lowest common denominator, or whatever was most popular.

Example: Atari's STe line got a nice blitter and more advanced sound hardware than the original ST about 2 years after the original ST launch, but almost no games used it. It just wasn't practical for game authors to target it when so many people had the original generation. (And hence the next generation didn't sell well because no compelling reason and so on). And in fact the marketplace was small enough that even just targeting either Amiga or ST and programming specifically for one and getting the most out of them wasn't done that much. People just wrote what amounted to a generic 68k planar bitmap graphics game that could be ported easily to either the ST or the Amiga, not taking advantage of the better features of either.

PCs were ugly and not nearly as sexy and fun as those machines from that era, but what they brought to the table was standards and upgradable commodity hardware.

>the Amiga emulated the fastest Mac FASTER than the fastest Mac

I remember when people would post this sort of thing on Usenet, for years after most people had given up on the platform.

By the time the A4000 came out (late 1992), there were Macs that also had 68040s and ran at higher clock speeds, so I don't think it was remotely possible to do what you say. How could a 25 MHz CPU emulate a 33MHz CPU faster than native under any circumstances?

I believe that statement is based on getting a 68060 Amiga accelerator (whatever the fastest one at the time was?)

Up until the A3000 and the competing '030 Macintosh machines the Amiga's graphics and sound processors made up for slower CPUs. However, by the time of the '040 the Amiga was behind with the exception of SD NTSC video production.

For me it's fun being able to play around/make hardware for a system that is still simple enough for one to understand broadly how the whole thing works.

The schematics are available for them & they're documented really well

People still make software & games for it for presumably the same reason, then there's people who are into it for the nostalgia and the games.

This is huge. Even the Raspberry Pi doesn't have open schematics or firmware.

The Amiga era was the last good overlap of "you can know everything about the machine" and "the machine can do useful things".

I think the Amiga was way ahead of its time and inspired a fierce loyalty in its fans. Also, retrocomputing is a thing: working within the constraints of retro hardware has an artistic/hobbyist appeal of its own. This is what compelled me to buy a C64 again (actually a TheC64 retro clone with an ARM board inside. I've no patience for the actual difficulties of plugging ancient hardware into modern peripherals).

There's even an enthusiast's market for new games and software for retro computers. The target seems to be either people who remember them fondly from their youth, like me, or even people who were actually part of the scene back then! It's something like a club for old automobile collectors I guess.

As for uniqueness: everything about the Amiga was unique, especially when compared with PC and Macs (or the Apple II). The Amiga was built out of custom chipset purposefully designed for the vision its creators had; these were not off the shelf chips like with PCs.

For a pretty detailed history of the Amiga, complete with hope, betrayal and tragedy, I recommend you google the Ars Technica series titled "A History of the Amiga". It will provide a glimpse into what was so unique about the system.

Just to save everyone a Google, you can find it here. https://arstechnica.com/series/history-of-the-amiga/ (shameless plug) I actually wanted to write the series in order to answer the question for myself of why the Amiga was so unique, as I never actually owned one. I feel like I found the answer by the end. :)

Retrocomputing is the programmer’s version of cringe. The very border of benign.

No one else has mentioned this but it’s a growing use case for a few in the retro scene:

These older computers are still auditable from a security standpoint. A pre-486 machine doesn’t have binary blobs or sketchy hypervisors that run in ring -1 that are under the control of a third party.

If manufacturers ever cease to provide general purpose machines, one way for the community to bootstrap itself would be from old hardware or FPGAs (if they can be trusted).

Hoping this could push the costs of getting into Amiga hardware lower. I never got to use an Amiga in "the day", and seeing the prices continually creep and creep higher really discourages me.

Just buy a Vampire v4, done.

He's complaining about cost, so a Vampire is not necessarily the best option (euphemism)

A Raspberry Pi 4 with PiMiga is a great way to get your feet wet (and might be all you need eventually)

I kinda wanted to play with a Video Toaster at some point.

The Vampire is a proprietary closed-source FPGA core that has nothing to do with the spirit of the Amiga. Their emulated CPU and chipset has glitches and issues are the order of the day.

A miSTer FPGA is a much better option, being completely open. An original Amiga 500 is not crazy expensive yet, and WinUAE is also an option, being the best and most accurate software emulator for the entire range of the Amigas (a lot more accurate and faster than any Vampire).

Vampire 4 works fine. Done.

Seeing as this needs the original chips taken from an Amiga 2000, I doubt it'll lower much in the way of costs.

I would buy this if it were commercially available. Is anyone looking to manufacture and sell this board? If yes how much would it cost?

All the info is in that link.

Unlikely. This board isn't suitable for mass production -- many of the parts required can only be obtained from another Amiga system.

Reply on news.ycombinator.com