»

[desiv]

Originally Posted by Sparky

I wasn't trying to be smarmy just realistic

While I agree, this is more of a "wouldn't it be nice" than anything useful hardware wise, the fact it, you're initial comment was just about the textbook definition of "smarmy."
So you might not have been trying to be "smarmy," but you were anyway..

desiv

[psxphill]

Originally Posted by Rebel-CD32

After reading the whole thread, you think this idea started with the colour of the PCB? I may not have a solid product to market but sharing ideas is better than nothing, and I'm already doing my best to contribute to the Amiga community with my other projects. Have you anything else to add other than smarmy critique?

I came up with the plan of adding 8 meg of chip ram & AGA to OCS amigas.

http://www.youtube.com/watch?v=azRzqI3BJ2A

[Rebel-CD32]

Originally Posted by Sparky

I wasn't trying to be smarmy just realistic and in the back of my mind I know I'd seen this post before somewhere else ....

http://eab.abime.net/showthread.php?t=39114

It's been 2 years to the day since you last bounced your idea ... and the only change I see is you've drawn a picture.

I'm all for ideas, but if you are so keen on it why not give it a go rather than cling onto the hope that someone smarter will do it for you ? Pick up the books, buy a soldering iron, grab the CD32 schematics and see what you can whip up ?

No, not smarmy at all, and neither is this new post where you simply repeat your "advice" in more detail...

As I've already mentioned, I'm already working on a bunch of my own projects which I'm more suited to, and so what if I've wanted the same thing for several years? Four years ago it would have been too expensive to order the expansion connectors due to the high MOQ, but now we can get the connectors made through Jens Schoenfeld with his own custom tooling. So an idea that was on hold has been revisited. I hadn't measured out and designed anything back then, and no one else had either so I quickly drew up the design I had in mind, then as an afterthought played around with the colours since Jens told us he can make the connectors in any colour. My talent is in artwork, and that's where it's needed most at the moment.

Originally Posted by psxphill

I came up with the plan of adding 8 meg of chip ram & AGA to OCS amigas.

http://www.youtube.com/watch?v=azRzqI3BJ2A

It's an interesting idea and very similar to how the IndivisionECS works. You can use high resolution modes in 256 colours (from a palette of 4,096 still) on Workbench and it has a chunky mode you can access if you write new software to take advantage of it. So far SCUMMVM and ADoom can make use of it.

[darksun9210]

well, whatever the situation, i have money waiting for a CD32 expansion that covers these basics of ram, hard disk, clock port? maybe even a PCMCIA slot for an ethernet card?

[psxphill]

Originally Posted by Rebel-CD32

It's an interesting idea and very similar to how the IndivisionECS works. You can use high resolution modes in 256 colours (from a palette of 4,096 still) on Workbench and it has a chunky mode you can access if you write new software to take advantage of it. So far SCUMMVM and ADoom can make use of it.

It's not similar at all to how the indivision works, as that just replaces denise. You're limited by the speed and size of chip ram. The indivision works like the a2024. With it's own framebuffer you can have hires screens with high refresh rates, but the display can only be updated at the speed that agnus feeds it data.

Avoiding chip bus contention you'll have to run the agnus as if it's a 320x200 16 colour display @ 60 hz. A 256 colour display will therefore run at 30hz, hires will cut that to 15hz, doubling the number of scanlines without interlace will be 7hz. Replacing the agnus and chip ram would actually make it useful.

The IndivisionECS allows you to hook up to a VGA monitor, but I would use an old 15khz Amiga monitor or an LCD TV with RGB SCART input.

[Rebel-CD32]

Originally Posted by psxphill

It's not similar at all to how the indivision works, as that just replaces denise. You're limited by the speed and size of chip ram. The indivision works like the a2024. With it's own framebuffer you can have hires screens with high refresh rates, but the display can only be updated at the speed that agnus feeds it data.

Avoiding chip bus contention you'll have to run the agnus as if it's a 320x200 16 colour display @ 60 hz. A 256 colour display will therefore run at 30hz, hires will cut that to 15hz, doubling the number of scanlines without interlace will be 7hz. Replacing the agnus and chip ram would actually make it useful.

The IndivisionECS allows you to hook up to a VGA monitor, but I would use an old 15khz Amiga monitor or an LCD TV with RGB SCART input.

It's similar in that it is such a custom reach-around of an expansion idea that even if it works you still have to start writing support for it in software because nothing that currently exists will be able to take advantage of it.

I'm sure you'll have no problem writing plenty of new software to take advantage of this custom configuration once you've developed the hardware and convinced the Amiga community they need it.

It's a nice idea but is it viable?

[psxphill]

Originally Posted by Rebel-CD32

It's similar in that it is such a custom reach-around of an expansion idea that even if it works you still have to start writing support for it in software because nothing that currently exists will be able to take advantage of it.

Support for the 8mb of chip ram wouldn't take much, you just need to make sure that auto config ram doesn't conflict with it. There are several ways that might work. A new kickstart rom, adding autoconfig ability to the hardware itself (you'll need a hook up to gary anyway for extra address bits, hell you could even replace gary) or inject a resident module in the chip ram on boot up that does it.

Most of the other changes would be AGA compatible. Chunky/hicolour modes wouldn't be, you'd need to add the monitor prefs for those & some changes to graphics.library.

Support for texture mapping in the blitter etc is a bit more of a pipe dream and software that used it is more likely to just be demos.

Software support isn't what would stop this getting it off the ground. You'd need to make an fpga to PLCC board with headers to go to adapters for two chips, so just the component cost is going to be expensive. But it would be awesome.