Where do you draw the line ?

[ijor]

But I am looking to use @ijor's cores as they are more mature than suska and @ijor actively maintains them. As to if he can add STE features.. maybe. I want to add DMA playback from the STE, but not sure what else the STE really has to offer which is of some actual use ?

It is certainly possible to add STE features. Sound DMA might indeed make sense to implement since we are targeting digital audio via HDMI.

[alexh]

not sure what else the STE really has to offer which is of some actual use ?

12-bit palette? Hardware scrolling?

[exxos]

Been thinking about this on and off.

I think the "trend" Atari was doing was of course, integrating the MMU,BLITTER,GLUE into one chip. There are changes with the shifter between STFM and STE as well.

So (again) I think ultimately the shifter (Atari custom silicon) and then ultimately the DMA (again Atari custom silicon) would have all been integrated into the same chip. Atari would not have integrated things like the CPU, ACIA, SOUND IC etc as they simply did not own the designs etc.

As someone mentioned earlier, it is what would be likely for the "next step" which Atari would have taken. I mean there was integrating all sorts on the Falcon with even more custom silicon. But the CPU, SOUND, ACIA etc were all still there.

So I think the bottom line is, we could integrate MMU,GLUE,BLITTER,SHIFTER,DMA into a single chip without getting into the (same) discussion of when Atari is not a Atari anymore kind of thing . It is pretty evident for the custom Atari silicon which was constantly being messed with on the Falcon etc.

So I think that is a reasonable point to "draw the line" at. If anything more is integrated then we basically reinvent the MiST / Suska again.

Of course I am not trying to discourage development of alternative cores like CPU, 1772 etc. As we all know, these chips are becoming harder and harder to find. So indeed that work is also important. As to what will make it into a "stand alone" single chip replacement (like we intent with Sparkalaphobia to be with the blitter) is a whole different topic.

[foft]

I think the older computers offer something things that the all-in-one FPGA solutions do not:
i) Original case + keyboard
ii) Original ports/electrically compatible
iii) Low frequency busses between 'chips', so that things can be extended/adjusted, mixed and matched.
iv) Stability. Reconfigurable cores means hardware behavior changes, often multiple times, after release.

Talking about these in turn:
i) This is getting very close now with 'professional' 3D printing services + people making their own hardware keyboards. https://forums.atariage.com/topic/35255 ... order-one/
ii) Low speed ones can be done with a few pull ups and logic level chips mostly. At least at low speeds of the 8-bits. For faster busses its a bit trickier.
iii) Already Atari was on the path of COMBEL - SSI->LSI->VLSI->ULSI etc. However I think keeping things single function is part of the hobbyist appeal of old products so we should not follow Atari here. So perhaps keeping a socket per chip on a new board makes sense. Then for cost reasons someone can create a multi-chip chip. Where a single FPGA containing independent blocks connects to each socket via a wire.
iv) Beta period of say a year, after which the core is frozen. This final frozen version is the only official version.

[Cyprian]

not sure what else the STE really has to offer which is of some actual use ?

12-bit palette? Hardware scrolling?

also:
- the BLiTTER
- a bigger than 192kB ROM in $Exxxxxx address range;
- LMC audio mixer in addition to SoundDMA;
- two Jag ports;
- input for external genlock clock;

[exxos]

I think with the C64 they might of got cases made and then put a small FPGA emulator in it. Basically what the Atari 2600+ etc is. Getting larger cases made like the ST etc is a whole new world of pain which has been talked to death for years. I don't really like the notion of a emulate in a Atari style box personally. But on another platform it seems to do extremely well.

The point about upgradability and tinkering I think is also a factor. Sure you can change the firmware to add software features to a point. But being able to hack and mod machines I think is part of the ST heritage. Is why I tried to make upgrades easier on the H5 by having banks of 68K sockets.

But I think also the topic of "resources" hasn't really being talked about much either. The problem with Atari world is there just simply isn't the user base to get much done. I mean if I was raking in 20K a month in profit, after a few months I would be able to afford to get proper injection moulds made for the cases. Be able to rent a unit and get some space to ship them etc. But sadly that is not going to happen. And even kick starters , while they tend to be popular on there platforms, I think any Atari project would lucky to get 20 peoples donating a few quid to it.

I basically have to fund everything out of my own pocket and I am not in even full-time work any more. @Icky has also ploughed a lot of time and money into projects himself lately. A lot of people as of course helped in various ways as well. But resources wise, it is only the sales from my store which is allowing funds for projects. Without trying to topic drift again, I am slowing down due to old age and health reasons these days. So it is only really downhill from this point on.

I do have a lot of ideas and things I want to do for the Atari scene. But sadly I just do not have the time anymore. Keeping my store running and general service stuff basically soaks up all my time these days. I mean sure, if someone wants to donate to me a couple million, I can pack in my day job and going to mass production of Atari cases or keyboards or whatever relatively quickly I think.

But to try and stay on topic, "where you draw the line" also has to tally with a "acceptable budget" and a certain amount of realism.

[stephen_usher]

The "The C64" is effectively a Raspberry Pi running an emulator, the VIC=20 and Amiga versions are the same.

To keep original Commodore 64s running people are recreating the original chipset using FPGAs. Someon'e even made a 6510 CPU. All of them are pin-compatible with the original motherboard, or you can buy a recreated or enhanced motherboard and fit the chips, rather like the H5.

Now, at which point does it stop being a Commodore 64?

Is it still a C64 if all the chipset has been replaced by modern reproductions but everything else is original?

Is it still a C64 if it's all the original chips but a recreated motherboard?

Is it still a C64 if then those chips are replaced one by one?

We get into "Ship of Thesius" or "Grandfather's broom" territory then.

[alexh]

Put out a feeler? Prepare a kickstarter for Atari projects? Make up a great pitch with videos and prototypes etc.

I for one would have bought a pre-assembled H5, take it to the next level with HDMI out and integrated ACSI HDD emulation and a case and I'm sold.

Have a long "pre-kickstart" time period to advertise it and drum up support.

Either then is $4M of people's money wanting an "Atari NxST" or there isn't? And you'd only loose a bit of time. I'm happy with Kickstarters that take years from end to delivery. I backed the Spectrum Next 2nd Kickstarter and the Retro Monitor Kickstarter and lots of others that took multiple-years (and in some rare occasions never)

[dad664npc]

Put out a feeler? Prepare a kickstarter for Atari projects? Make up a great pitch with videos and prototypes etc.

I for one would have bought a pre-assembled H5, take it to the next level with HDMI out and integrated ACSI HDD emulation and a case and I'm sold.

Have a long "pre-kickstart" time period to advertise it and drum up support.

Either then is $4M of people's money wanting an "Atari NxST" or there isn't? And you'd only loose a bit of time. I'm happy with Kickstarters that take years from end to delivery. I backed the Spectrum Next 2nd Kickstarter and the Retro Monitor Kickstarter and lots of others that took multiple-years (and in some rare occasions never)

That's a great idea
I'm in

[exxos]

I think kickstarter was already talked about and while I think I tried to sign up I gave up in the end. Was all A few years ago now.It also misses the other side of the problem that there would have to be users to buy the final product as well.

I think it costs something like £75k x 2 (this was pre-covid) to get the molds done. There needs to be money for the plastic, storage, shipping premises etc. So likely need about 200K I think. If anyone wants to start a kick starter then go ahead. I'd be frankly surprised if it got to £500 raised.

EDIT:
Case topic.

viewtopic.php?f=62&t=224