ATARI ST FPGA

[lotharek]

new project is coming :

 

lotharek.pl/category.php?kid=17

 

 

soon :

 

- pricing - do not ask me about it yet

 

any question concernig technical -- Ask TILL --- till ( put here @) harbaum.org

Edited April 13, 2013 by lotharek

[GadgetUK]

Looks cool, but I dont see the point in one of these when things like FireBee already exist? I have an actual ST and Amiga, and emulators, so why should I get an FPGA of ST / Amiga?

Edited April 13, 2013 by GadgetUK

[Gandor]

Very interesting

[adolobe]

So is this basically a miniturised ST clone?

 

Looks interesting and possible would want one if it is.

 

Would be really cool if it implimented a 68060 much like the Flacon and Amiga upgrades out there

Edited April 13, 2013 by adolobe

[Atari8guy]

If it could do Falcon, that would be very interesting....but from my reading, right now anyways it's limited to 8Mhz.

[GadgetUK]

If it could do Falcon, that would be very interesting....but from my reading, right now anyways it's limited to 8Mhz.

 

I think that's my point. Once it will do say 16+ Mhz STe then I would be interested, but right now its easier to use an emulator or actual hardware.

 

Will keep an eye on this one, it looks promising.

[+CyranoJ]

Firebee (Board only) - 622 euros

Minimig (Board only) - 160 Euros

 

And this is said to be less than the minimig.

 

There is your point. The firebee is way too expensive (and, as a TOS clone, quite incompatible with many games) to attract casual (or new) Atari fans.

Edited April 13, 2013 by CyranoJ

[oky2000]

Nice project!

[swapd0]

It could be nice to include the Jaguar chipset without bugs and maybe a flew cache or write buffer into the blitter, rgb lighting...

[ParanoidLittleMan]

I think that this will be success if price will be not high and level of compatibility will be high .

Full STE features are very welcome, and accurate CPU and other component timing - then old and not so old good SW will work fine.

 

I would not dream for now about some Falcon clone - there is enough thing to solve even with ST clone. And I really don't see as realistic some Jaguar support too. I think that most of people here wants decent ST(E) clone . Will mail Till about timings ...

[AnakiMana]

I would buy one. It would be great to have a boxed one in the living room on the big TV.

[GadgetUK]

Thats a good point, if its small enough I wouldnt mind one boxed on main TV.

[+Gemintronic]

Now all we need to do is have these individual FPGA developers standardize on one C-One like board. That way we can focus on making better cores rather than re-inventing the boards each time. Dunno why the C-One or Minimig didn't become that standard.

[barnieg]

This is why I think open source/ open standards are better for these sort of developments, it makes the possibility of standardising more of a possibility. At least with the MiniMig it has allowed for spin-off projects

 

 

 

Now all we need to do is have these individual FPGA developers standardize on one C-One like board. That way we can focus on making better cores rather than re-inventing the boards each time. Dunno why the C-One or Minimig didn't become that standard.

[Scarlettkitten]

I'm not really interested in these things till they run TT or falcon compatible software, As these are the machines out of most peoples reach.

[+remowilliams]

There is your point. The firebee is way too expensive (and, as a TOS clone, quite incompatible with many games) to attract casual (or new) Atari fans.

 

There's the key. Make it compatible and cheap enough and it's got a chance I think...

 

 

Now all we need to do is have these individual FPGA developers standardize on one C-One like board.

 

 

You want another fpga project to die a horrible lingering death ?

 

[ParanoidLittleMan]

Funny is how people behind projects like this listen not to target users, people knowing something about Ataris. When I said how high level of compatibility is important, Dr. Till started to talk how great Mist is because you can start Xenon in 2 seconds Something nobody with Atari ST experienced yet

All what I see from our short conversation is: there will be no accurate CPU emulation, no accurate video - so only 16 color stuff will work - if some game has own high-color displaying, overscan, you will see garbage. Hardly will see full STE support.

Because it has built-in Satandisk or UltraSatan (rather the first) may be still interesting if price will be low.

Unfortunately, as I see all projects have same problem, all are based on some free 68000 core, which should be improved. Maybe there is better, but not free ?

[TommyIII]

Midi? Would be nice.

[swapd0]

It would be nice if they also include a packed pixel format, I know that the ST don't have it but ii should be very easy to implement.

[harbaum]

Is there software supporting packed pixel formats? Which ones?

 

With the MIST i am trying to create something as simple as possible (to keep things as cheap as possible), compatible to the original ST and at the same time compatible with todays peripherals (VGA screens, USB mice, USB keyboards, SD cards).

 

Unfortunately there are compromises that have to be taken. And sometimes it's the required work vs. benefit ratio that kills a feature. E.g. in order to allow for working overscan the video controller as well as the cpu have to be cycle exact. None of the existing vhdl implementations of the 68000 cpu are cycle exact. We are talking about thousands of hours of work here as this requires knowledge that's not documented anywhere (e.g. the exact timing of bus cycles relative to the instruction cycles) and one would have to build a test rig to analyze a real 68000 CPU beforehand. The same is true for the shifter. The features that are used to achive overscan on the atari st are not documented at all. So in the end the question is: Is it worth spending thousands of hours to make the last 5% of the software work? Or do we want a cheap machine that's able to run 95%? I personally think the 95% already is pretty entertaining ... but that doesn't mean there will be no improvement. The migration from ST to STE is very likely possible with the current MIST board (e.g. the board itself already supports 18 bit colors while only 9 are required for ST and 12 for STE).

 

I hope the MIST will be for the Atari ST what the Minimig is for the Amiga.

 

And if you want a falcon clone there are hades, milan, firebee etc ...

Edited April 19, 2013 by harbaum

[AnakiMana]

And if you want a falcon clone there are hades, milan, firebee etc ...

 

Except that Hades & Milan are not available anymore... and because of their age, many of them are failing.

The FireBee is pretty much all that remains, and it's very expensive for its specs in today's world.

 

The fact is: these are hobbyist computers. If we want the fastest and most compatible with everything in the world today, we buy or build a Windows/Mac/Linux computer.

The Atari platform... the FireBee... is far removed from today's computing world. Nobody wants one except us nostalgic techies and programmers. It's hard for most of us to justify spending $1,000 on something we don't need. If the price was $200-400, I would feel comfortable buying a fairly barebones project that would give me a lot of enjoyment.

[AnakiMana]

I think the MIST is more attractive to more people because of the abandonware game collections freely available for the ST/STE line. There were a lot more people who had those systems, and have fond memories of games and music creation. Instead of just techies and programmers, you're now including gamers. And the price is low enough to attract more casual nostalgic people.

[harbaum]

And the price is low enough to attract more casual nostalgic people.

 

That's the point.

 

And there are only two ways to get the price down: Simplicity and volume. Volume barely exists in retro computing. That's why i decided to go for simplicity. We'll see if this will work. But i think it's definitely worth a try.

[life_is_pleach]

Looks very interesting.

[+Gemintronic]

As long as GFA-Basic runs on this I'm happy.