(re)build a *NEW* Atari Computer?

[tcropper]

I have a question on this topic for Curt and others - Is Beetle planning to market his upgraded Atari systems, such as the Atari laptop that he built? I think that even that would find an audience that is receptive to it. He might make some $$ in the process. I know that I would seriously consider purchasing an upgraded piece of 1200XL hardware. I would trade my 130XE for it

[tcropper]

I am going to bring this topic back alive a little bit...I did some research, and there is actually a company that is bringing Commodore Computers back alive as a company by selling miniture 'computer' devices that handle audio, video, etc. So that's interesting news.

 

I also happened to see lots of Atari motherboards over the weekend on Ebay for various types of old Atari computers.

 

So why can't Curt or someone similiar get us a Atari 1200XL in a box, a complete re-make, similiar to the Flashback console, that would be a full-blown 'chip in a bottle' , and then some of us could take it from there...

[mos6507]

I think the C= stick and the Flashback2 were representative of a fad in the market that is waning fast. I don't think the market would be very receptive to another x on a stick product at this time. Plus the economy is spinning into a recession. Sure, some die hards will want this, but joe sixpack kind of has to cut the fat out of his finances in order to keep up with his ARM and gas bills. There might be room for a budget GP2X type device. The GP2X was always kind of overpriced and under-distributed.

Edited July 7, 2008 by mos6507

[Wickeycolumbus]

I think the C= stick and the Flashback2 were representative of a fad in the market that is waning fast.

 

maybe the new atari movie will get people back into classic gamming.

[tcropper]

I am not sure these devices are a fad...they are a market niche, same as the regular Atari Computers were back in the 80's...It was usually the computer/science folks that bought them, and then lots of middle class Americans. But now, look at the BBSes of today...there are only 400-odd bbses left in the known world...so that's market niche, not for every Joe six-pack that wants internet.

[peteym5]

With this day and age, most people would not pay more than $100 for a flashback computer, even if it had a hard drive interface/flash rom and stored a bunch of games. Believe it or not, you cannot put every old atari 8-bit game on there and sell it, some still may come after you for copyright infringement even though the computer is over 25 years old. Think Namco is stingy about Pac-Man still and there'sabout First Star Software and Boulder Dash. Might have to check with some companies before selling a retro computer full of games.

 

The other major issue is it cannot run much more than what you can get an used unit on like Ebay, or else it would be a major failure.

Edited July 7, 2008 by peteym5

[Rybags]

The problem is that although the latest gen consoles have played a big part in the retro respark, they've also served to kill off a lot of the demand for redone retro systems themselves.

 

90% or more of people who play retro either via the services offered my M$, Nintendo etc, or via homebrew on consoles or emulators on computers are probably happy with that and wouldn't fork out the extra money for something like a C-64 DTV or Flashback 3.

[machf]

Hmmm... just the other day, I was thinking, if someone would make an updated "A8"-clone, (1) would anybody be interested in buying it, and (2) what things should it incorporate? Something like using 65c816 @ 14MHz (actually, maybe a switch like the old "turbo" ones in the old PCs up to the Pentiums that would allow the original 1.79MHz or 8x faster - the A8s use a crystal at ~14MHz speed and a divisor to get it down to 1.79), with the 65x821 similarly for the PIA, dual POKEYs for stereo output as default, replacing the DIP RAM chips with SIMMs, providing decent video output circuitry, etc.

[MEtalGuy66]

Oh sh*t.... here we go again....

[+remowilliams]

Oh sh*t.... here we go again....

Wheeeeee! Cyclical pointless discussions are fun!

 

[Math You]

The only chance of us seeing another Atari computer is if somebody makes another Flashback product.

 

I remember reading how the old Atari company were designing an Omni Computer that could run 2600/5200/7800 and 800 programs.

 

Something like this could be a hugh success, the Flashback 2 sold over 800,000 units and that was only in the U.S.

 

It could also be made in many different variations such as a console, handheld, educational computer or mini laptop.

 

Edited July 9, 2008 by Math You

[frogstar_robot]

Hmmm... just the other day, I was thinking, if someone would make an updated "A8"-clone, (1) would anybody be interested in buying it, and (2) what things should it incorporate? Something like using 65c816 @ 14MHz (actually, maybe a switch like the old "turbo" ones in the old PCs up to the Pentiums that would allow the original 1.79MHz or 8x faster - the A8s use a crystal at ~14MHz speed and a divisor to get it down to 1.79), with the 65x821 similarly for the PIA, dual POKEYs for stereo output as default, replacing the DIP RAM chips with SIMMs, providing decent video output circuitry, etc.

 

Anything along these lines would require significant engineering for a very little return. Very few people with the needed skills will do this for "Oh! That's Cool!" factor because this thing isn't likely to have a significant financial return. There have in fact been a few 65c816 efforts. None have gotten further than the partially functional prototype stage. And no wonder, there would be few developers for what would essentially be a boutique machine. It would spend 99% of the time in Legacy Compatibility Mode. Ditto for ultra leet VideoBoards that require getting inside the machine and soldering to install. There is another thread about the possibility of a graphical enhancement addon that would be a PBI/ECI or cartport hardware/software combination video overlay type thingy. That likewise has limited viability but since it wouldn't require opening the machine it would be more accessible. Even so, it would have to bundle some homebrew or reworked classic software to do interesting things out of the box. Don't hold your breath for that one either.

 

The very best you can hope for is FB3 type of thing with an implementation of the A8 chipset as an epoxy blob. MAYBE those who do this will be nice enough to expose all the buses for hardware hacking.....assuming it happens at all.

 

On the other hand, there are cool things people like Beetle and Heckendorn are doing today. A cased A8 equipped with a 130XE motherboard, multi-OS roms, 576K RAM, built-in flash memory port, sio2serial hardware, and decent video output wouldn't be too hard to do with "off the shelf" components. Basically, get an out of the box experience for the most common and useful mods that are commonly done today. You wouldn't make much money on a thing like that but it is attainable as a labor of love and a few people would buy them and be able to use them to the fullest.

[peteym5]

One Ideal I always thought about was making a multi-platform Atari flashback, capable of not only running 8-bit computer software, but able to run 2600, 5200, 7800 cart images as well. That is if someone was going to port the chips onto a FPGA and something extra to select which video/audio chipset you want to emulate and map it to the correct memory locations. The 65816 ideal could also help in this case to have a more sophisticated monitor program/operating systems. Can add support for an internal hard drive to store hundreds of games inside, and able to run a cart image from file. One issue is you will need multiple cartridge ports for the different game carts. Maybe have different adapters.

[Math You]

How about using a 200mhz w65C02s cpu!

 

http://www.westerndesigncenter.com/wdc/

[machf]

Oh, I know probably anything in that vein would have to be sold for no less than $200 and few (if any) people would buy it, I was only commenting that I had been thinking about that a few days ago, and then I come back here yesterday and see someone revived a (long) thread talking about right that... I haven't read through all pages of the thread yet, BTW. It's going to make an interesting read, seeing what people have proposed over time. Or maybe just a fun one.

 

What got me thinking about this in the first place was that I was browsing the WDC site for documentation the other day, and found out they have a sample request form where they will send you free samples of their ICs if you intend to use them in developing some prototype that will use them and later intend to start mass production or something. I started to think who would be interested in using a 16-bit processor running at only 14MHz on a computer... that would have been an improvement back then (look at the Apple //GS), but nowadays it isn't much. I guess they're selling them for dedicated systems.

 

Aside from those problems concerning the very low sales/high price and such, there's also the issue of trademarks and copyrights. You couldn't use the name "ATARI", in the first place, you couldn't use the original ROMs, nor the custom ICs (POKEY, ANTIC, GTIA, mainly), so what do you end up with?

 

I agree that something implementing common mods "out of the box" would be a more interesting thing.

[machf]

How about using a 200mhz w65C02s cpu!

 

http://www.westerndesigncenter.com/wdc/

 

I thought I had read something about that somewhere, but haven't found any datasheets indicating it. The 65C02 and 65C816 are limited to 14MHz only. Maybe it was just wishful thinking, or a yet-to-come version...

[peteym5]

100MHZ and 200MHZ are possible with FPGA and ASIC forms of the 6502/65816. Maybe some people misunderstood the webpage.

 

You certainly need permission from the current Atari entity to make a retro game counsel of the 8-bit. Did anyone even contact them regarding this issue and these ideals we have been floating around here? The 8-bit has a strong following along with the 2600.

[ijor]

Aside from those problems concerning the very low sales/high price and such, there's also the issue of trademarks and copyrights. You couldn't use the name "ATARI", in the first place, you couldn't use the original ROMs, nor the custom ICs (POKEY, ANTIC, GTIA, mainly), so what do you end up with?

 

I said it already. The Atari community should ask Infogrames (or whichever company owns the rights) to release the vintage Atari IP.

 

If they would accept to put it in the public domain (which I doubt they would), the better. But at least they should formally allow redistribution and non-commercial usage of the vintage Atari IP. This should include, at least, firmware, system software and IP related to custom chips.

 

This affects not only "FB3 like" products, but opensource FPGA developments and software emulators as well.

 

You certainly need permission from the current Atari entity to make a retro game counsel of the 8-bit. Did anyone even contact them regarding this issue and these ideals we have been floating around here? The 8-bit has a strong following along with the 2600.

 

I'm not sure you would need permission. You certainly couldn't use the brand, and technically you couldn't include the original ROM firmware. For opensource, or for non-commercial distributions for that matter, you might not need to include any ROM though. You could let the user provide it (as is done manytimes in other platforms).

 

But it is too bad we need to resort to that convulted measures. And from a purist point of view, it might still not be 100% legal.

[mos6507]

The ARM chip we're using in Chimera cartridge has about as much horsepower in it (internal RAM aside) as the 68060 processor in the Amiga system I used to edit Stella at 20. And that thing cost less than $10. That's about 60 times as fast as the host 6507 in the Atari 2600. It's surely fast enough to run a 6502 emulation if it could output a display. But Batari is actually getting it to work like a simulated cartridge, eliminating the need for any CPLD at all. It's polling the bus on every machine cycle. This is the kind of thing you'd normally have to rely on programmable logic to do (as well as some ancillary parts to load games into the system) now possible with just one system-on-chip. So my point is, today's off the shelf CPUs are just insanely powerful and cheap. So I think software emulation is really where it's at going forward. GP2X territory. You get a lot more bang for the buck that way.

Edited July 8, 2008 by mos6507

[Math You]

Does anyone know if the Atari operating systems are still protected by copyright. If not, then could they be re-written?

[mos6507]

Does anyone know if the Atari operating systems are still protected by copyright. If not, then could they be re-written?

 

How would you rewrite them? There are already several replacement OSs out there. A ground up rewrite would destroy backwards compatibility with existing software. The OS in the Ataris is more of what we'd consider a BIOS today. It's just the layer between the program and the hardware. It's not a task manager the way something like Windows or OSX is. If you wanted something more advanced you'd have to layer it on top like Diamond GOS tried to do.

[Rybags]

If you're doing emulation, the OS isn't as important.

 

You can have intelligence built in that does context checking of entry into OS routines, and can perform the task required.

 

I don't know the exact number, but I'd guess that there's under 20 undocumented entry points that are commonly used. Regardless, it falls into a similar category of 400/800 to XL/XE compatability.

[ijor]

Does anyone know if the Atari operating systems are still protected by copyright. If not, then could they be re-written?

 

They are of course copyrighted.

 

Yes, you can develop a compatible OS. And actually, there are already free OS replacements available. But there is always some degree of incompatibility.

 

If you're doing emulation, the OS isn't as important.

You can have intelligence built in that does context checking of entry into OS routines, and can perform the task required.

 

I don't agree. I think that an emulator, or a hardware clone, should try to target 100% compatibility. Ideally, you shouldn't be able to detect at all if you are running under an emulator, a hardware clone, or original hardware. Any OS that is different, is not 100% compatible no matter what tricks the emulator performs (because, by definition, I could compare it byte by byte with the original one, if I wanted to).

 

Yes, it is not much different than 400/800 vs. XL/XE compatiblity. That's why most (all?) emulators have the option to select the exact model and OS version.

[ijor]

The ARM chip we're using in Chimera cartridge has about as much horsepower in it (internal RAM aside) as the 68060 processor in the Amiga system I used to edit Stella at 20. And that thing cost less than $10. That's about 60 times as fast as the host 6507 in the Atari 2600. It's surely fast enough to run a 6502 emulation if it could output a display...So my point is, today's off the shelf CPUs are just insanely powerful and cheap. So I think software emulation is really where it's at going forward. GP2X territory. You get a lot more bang for the buck that way.

 

Software emulation vs. hardware cloning is a very interesting debate.

 

Personally, I disagree with wgungfu and others that claim that software emulation is only an approximation. There is no intrinsic reason why software emulation couldn't be as accurate as a hardware clone. And we all agree that current personal computers are insanely powerful.

 

But which is the best solution economically in dedicated hardware, that's another story. I don't know much about the 2600, so I can't comment about it. But for a full A8 emulation, I don't think an ARM7 (which I assume that's what you are talking about) is powerful enough for the task. Not at the level of accuracy that you could achieve with a hardware clone. You'll probably need something more like an ARM9. And for emulating 16-bit computers, you'll probably need something like the Intel Atom or a high end Arm. So we are probably not talking about very cheap processors.

[Retro Rogue]

Gotta disagree there, software emulation is only an approximation, not an exact duplication. And that will always be the problem with software based emulation. I have yet to hear POKEY emulation that sounds as dynamic as the actual thing.

 

Well, in how far does it not approximate the real thing close enough? The software runs, the graphics is identical,

 

Similar, not identical to actual hardware. You can set up a pc running an emulator next to the real thing, and it'll still have a different feel to the graphics because of running through an actual GTIA/ANTIC. All you're doing it trying to map opcodes and certain functionality to modern hardware, its a translator at best, not a 1 to 1 reproduction, you have to go gate level for that.

 

the sound is identical,

 

They're approximations, a PC sound card emulating a POKIE has nowhere near the same tonality, depth, and timber of the actual POKEY, i.e. less dynamic. Same problem with a SID emulator (though SID emulation also suffers from the lack of analog components of course). They sound similar, not the same.

 

where's your problem? Today's PCs are fast enough to emulate the chips on a cycle basis, so there's no problem

 

Once again, emulation is an aproximation. You're trying to map the functionality of one piece of hardware to another. You're not reproducing it at gate (i.e. hardware) level. You'd need to go FPGA for that or take the time to recreate an ASIC. And current PC hardware is just barely fast enough to do a slow gate level emulation of PONG.

 

 

Concerning POKEY, I don't see (or hear) what you mean. Could you please elaborate?

 

See above.

 

If there's anything that doesn't sound like it should, I'm glad to fix it - in atari++ that is. (-:

 

There's nothing much you can do about it, an emulator is fine for its purposes. Its no different than people saying an arcade game running in an emulator is the same thing, when the experience is much different on the actual hardware.

 

 

 

 

Real hardware.

Why? This is a serious question,I don't get the point. I can play the old games and the old software.

 

Its not about just being able to play an old game or old software like its some sort of commodity. Again, that's the same issue with something like MAME vs actual arcade hardware. Its approximations of the actual hardware. If you just want to be able to run some old game or software, that's great, it certainl meets those needs. If you want to be able to experience like on the actual machine, then you need the actual hardware or a reproduction of said hardware. Same thing we did for the FB2 vs. just going with another NOAC port or embedded winbond emulator.

 

Concerning home computers, I mean. A system is defined by the software that runs on it.

 

A system is defined by the hardware and software.

 

Even if you build your own extended Atari machine, which software will exploit its possibilities, other than those you write personally? Besides, the machine will be obsolete at the time it has been constructed. The whole business process that makes custom-made machines like the Atari possible back then no longer exists today.

 

Sure it does, its called a plug and play. We were redoing the Atari 8-bit for the FB3 originally, and were planning for full hardware support. There's also the entire embedded systems market (something also catered to in Circuit Cellar) that thrives on a cottage custom made machine market. The model is still out there, just in different venues then generic home computers.

 

Sure one can dream of "what if the Atari had 16 players instead of four", but at the time you're dreaming, I can emulate the whole thing in software before any piece of hardware gets ready, and neither the hardware nor the software implementatoin will be used by any program. So what's the point?

 

You can create an approximation of it in software via an emulation before the hardware, and test what if's on a trial basis. But that's hardly the same as the actual hardware, where the process eventually has to move to. You're simply describing a common proof of concept as well as a platform development process. I.e. most mobile platforms, etc. provide developers with emulators and simulated environments for application and game development on their PC's, mac's etc. But the application still has to go through testing and running on the actual hardware, and the final product still looks, sounds, and feels different running on the actual hardware. The approximation on the pc's display still will have a different look and feel running on the actual lcd screen of the device through the actual physical display hardware. The approximation of the sound on the PC will still have a different presence when running on the actual hardware through the small tinny embedded speakers. All these things are issues I'm well familiar being in development as well, and issues I have to deal with on a continual basis.