CGX 4 and P96 SDK

[arnljot]

Hi

I´ve been surfing around looking for the Driver SDKs for these just for fun lately.

And I haven´t found anything. The only thing I found (ref1) was that Toni Wilen (ref2) (WinUAE) mentioned that he had the P96 SDK, and that he had asked to it to be GPL´ed.

What´s the chances of tracking up current rights holders, and persuade them to GPL it for AmigaOS 3.x?

Personally I´d like it for educational purposes, and people like Plato42 who are looking into USB video on the amiga would have an easier life.(ref3)

EDIT:
I sent the following short mail to Cloanto

Hi,

From a forum post by Toni Wilen I understand that you are the current lisence holders of the Picasso96 RTG software.

I was wondering if it could be possible to get a copy of the Driver SDK package for my own educational reasons, and if you would consider to GPL or LGPL the software?

Kind regards,
Arnljot Arntsen

[arnljot]

Update:

Today I received a reply from M. C. Battilana of Cloanto where he says that my request is forwarded to Tobias and Alexander, he also says that Cloanto have proposed to them that Picasso96 should be open.

I hope Tobias and Alexander answers as fast as Cloanto did. Very good of Cloanto answer this fast.

[arnljot]

Well, time goes by

I haven't heard anything from the P96 copyright holder.

Also, the issue is once again relevant with the new graphics.library effor by cosmos.
eab thread

It would be awesome to see CGX4 extended to support Mediators and WinUAE.

[arnljot]

Re Tobias and P96.

Please explain to him that both Cloanto and Toni Wilen has the needed SDK, and others have the binaries.

What we need is permission to use it as we'd like from him and Alexander.

[arnljot]

Great, lets hope for a positive reply.

[arnljot]

For classic amiga computing there are some operating system components that should be in the public domain.

RTG, P96 or CGX4
AHI, it is open source!
Miami or AmiTCP, don´t care which
Poseidon USB stack, it´s open now for AROS.

And I think AROS and Poseidon USB stack is the example to look to. Chris has made a fantastic piece of software, and it´s his. And at the same time it´s vital to the community, and it needed to reach to other platforms than those he´d be able to support both as a free time project or as a commercial venture.

So a bounty was set up... Now, I know that there might bee too many bounties out there. But I think that RTG and TCP/IP would be two decent things to have opened up. Not only for classic amigas, but for the whole Amiga ecosystem.

I´m not going to push for this, but if a bounty can be agreed upon with the IP holders then I´ll happily contribute. Because as they say, it´s their IP. And they legally own it. So we should look into an incentive for them to open it up.

[arnljot]

It´s a bit poor that the IP holders doesn´t provide any reasons. Not that they have to, but some reasons are more understandable than others, say if for instance envoy contains lisenced code that he cannot give away.

Perhaps we should just all pay Karlos and Cosmos a bundle of cash to make a new RTG system ;-) And Piru and Plato42 one to backport ArosTCP with a GUI :-D

The developer who reverse engineered the P96 system to develop a gfx card driver could have peaked at the WinUAE/UAE code I guess (perhaps that´s what he did). But it´s not the way it should be.

Also, it would be far better if CGX4 and P96 could be unified. Since P96 is exclusive on some hardware(Mediator/UAE) and CGX4 on other(CVPPC and GRex).

How much code from AROS is usable on classic for a CGX back port?

[arnljot]

Yes, I realise that.

Some documentation could be retreived from the linux world. Some even from the Amiga 68k linux I guess.

More importantly is to select what to support.

I think it should have an abstraction layer for amiga bussboards ("native" Zorro) and PCI bussboards (Mediator, GRex, Prometheus). That would be probably an framework on top of expansion.library, pci.library and openpci.library.

Then which cards. Cybervision family cards(64, 643d, PPC), Picasso family cards(II, 4), Voodoo cards and Radeons should come first. Then other related cards (chipset wise).

And screens should be able to be in a grid (side by side, above/below) and stacked(amiga like).

Back to the documentation side of things. Like I said, if it were to be an LGPL product in the end, much could possably be gleamed from open source.

As for bounty funding, I think both Vesalia and AmigaKit could find it in their hearts to get the ball rolling, perhaps even Elbox and Individual Computers too.

I do think that the users would welcome it too.

This time it could be done propperly with more accellerated rasters too like you mentioned in another thread here

[arnljot]

A merge of south and north Korea is probably more likely than any merge of CGfx and P96 (aka MorphOS and AmigaOS)

I think you might be right. But hopefully we´ll be able to agree on what should be, and then perhaps motivate someone next to build it.

I think that if we could specify a new API which could have emmulation for CGX and P96 on top of it, it would be a relief to the Amiga community.

Remember, AROS has discussed making Cairo the base and have a CGX emmulation layer on top of that. ATM it´s kinda "side by side" which isn´t optimal. I don´t know if Cairo could be suitable as a basis for an Amiga RTG, probably we´d want to start designing all from scratch.

I´m not too familiar with what´s wrong with current amiga TCP IP implementations, besides them not being open. Which stops them from being ubiquitous in Amiga land like Poseidon could be now. That code is 20 years old shouldn´t mean that it´s bad. After all many parts of the Amiga operating system is, and we´re not opening that can of worms now

The reason why I´m brining up RTG as a pressing issue is Comos´ efforts to improve the graphics.library. To make it RTG compliant. It´s a great effort and best of luck to him! This shows however that some Amiga coders have matured to a level where perhaps elefants can be eaten...

[arnljot]

I was thinking about the documentation for things like mediator/grexx, let alone specific cards.

If anyone attempts this, it´s going to be a huge project.

However, I think that the openpci guys(guy?) would be the right one to have an alliance with. There is good support for GRex in openpci, and provisional support for mediator in the latest version of openpci.

I dunno if Elbox will come down from the fence and join the lgpl open source party, or if any qualified coders would welcome them if they would given all the bad blood. But lets face that if any of this comes to fruition.

I think that if this were to go ahead, then it first should have a six month designing stage where in parallel a team source needed documentation and structure it in a wiki for the devs.

Any bounty raised should go to a lead developer or a small number of developers.

I know, I´m a dreamer...

[arnljot]

I have to admit, I rather like the idea of a replacement graphics.library as a solution, in theory it can be implemented more cleanly than existing RTG for classic systems.

I don´t know the Amiga APIs or libraries well enough to understand this concept

How, in this "vision" is support for various buses and cards achieved in your mind? I would think that this would be delegated to a driver that know the chipset. The next problem is that this chipset can reside on a zorro bus, pci bus - or heaven forbid usb bus - LOL

[arnljot]

hehe, I feel the same Gulliver. But at the moment we´re just throwing around some ideas here.

It´s really helpfull to pick Karlos´ brain, because he´s knowledgeable on the subject. Would be fun if also Piru and Plato42 had some dreams and ideas... (fishing expedition here guys )

[arnljot]

Maybe a good place to start is to doodle some design documentation, just for the heck of it. To get a mental picture of what the new system would be like. Perhaps a google code page, or a soureforge page?

Then when we know what it´ll look like we´ll know if we can attract a bounty for it and a coder or three?

[arnljot]

Perhaps you´re right. But I think that it would be easier to gather support if people knew what they were taking a position on. And also pie in the sky dreams could be grounded first so that the project could have a hint of realism.

But either way, the nay sayers could perhaps overrun the project and discourage coders before it takes off.

Would you be interested in participating with design or code? I don´t know Amiga APIs or C coding, but I´d be happy to contribute with sourcing documentation and doing other leg work as needed or instructed to

[arnljot]

What does the CGX superlayers do? That is one of the CGX P96 differences?