[Coco] Future CoCo discussion (was CoCoNet)

[jdaggett at gate.net]

Frank I tend to agree with you in that there are those that want emulation 
and those that would like hardware. 

Hardware route using an FPGA is best taken using existing boards. The 
problem there is no one board is "prefect" for a Coco4. Ones that would be 
near perfect also exceed the $500 price tag. Virtex/5 boards can exceed 
$1000.  

designing one exclusively for Coco4 may not be a viable cost alternative. 
Any large FPGA is going to require an 8 layer board as the chip will be in 
BGA package. Better to use a less perfect development board that has 
sufficient expansion capability and add boards to the expansion ports. 


james  


On 28 Dec 2008 at 15:48, Frank Swygert wrote:

> Didn't really mean to hi-jack the CoCoNet thread, so replies will be
> in this new thread: 
> 
> You weren't "shooting me down" James, I really think I understand.
> Those just wanting to play games and such from their youth probably
> won't mind emulation at all. Those true "retro" computer admirers
> prefer the hardware as well. 
> 
> The only partial hardware solution is the FPGA idea. Some of the
> hardware will have to be emulated, but the thing could have a CC3 mode
> and an enhanced mode, including faster hardware. If something like the
> old "512K BASIC" could be implemented in DECB that would be fantastic,
> but then backward compatibility would be lost. I liked the ease of
> programming the CoCo in BASIC, though there's hardly a point now. With
> the FPGA unit you can just "throw some mgHz" at the problem of speed.
> Doesn't help memory access, but that can partially be addressed with
> programming, especially with RAM drives and such that lots of memory
> make available, and faster drive hardware. 
> 
> I wrote a program once that used a RAM drive on a CC3 -- it took a bit
> to load from disc, but ran fast once loaded. To reduce a power
> interruption from causing data loss, data was still written to disc.
> The main program was mostly a menu that called other programs to do
> specialized tasks then return to the menu.
> 
> Emulating the entire system is doable NOW with relative ease and
> little expense. A software package designed to take over a PC entirely
> (or rather mask PC identity) could be run on any old cheap Pentium PC.
> I had wanted to set my old 486 HP Omnibook laptop up that way using
> Keil's DOS emulator, but HP used some strange screen drivers and such
> that aren't 100% compatible without special drivers, and there are no
> DOS drivers, just Windows 95/98. No point in setting it up with all
> the overhead of Windows. 
> 
> I don't find the hardware issue as that important, not when 100% of
> the existing hardware can be emulated. I can only see the absence of a
> cartridge port as a block to those who want to use the emulated CoCo
> for one of it's strengths -- interfacing with the real world. Either a
> ISA/PCI card could be designed and programmed to emulate a cartridge
> port or the PC parallel port could possibly be reprogrammed to act in
> a similar manner. 
> 
> We have to face the same fact that Motorola did when they went from
> 6800 architecture to 68000 -- there can't be 100% backward
> compatibility if you want a lot more power. 80-90% CoCo3 software
> compatibility (which would mean 60-75% CoCo 1/2 compatibility, I
> think) should be an acceptable figure. All the hardware is so old
> (with the exceptions of Cloud-9's stuff, and a few other exceptions)
> in age and technology that there's no real reason to lament hardware
> compatibility. It could be possible with either method, but is it that
> necessary? 
> 
> --------------
> From: James Dessart <skwirl42 at gmail.com>
> 
> On 12/27/08, Frank Swygert <farna at att.net> wrote:
> 
> 
> > > What I find ambiguous is ... that the emulators do the same thing,
> > > only run on a PC also. While the idea is to make the PC a slave to
> > > the CoCo instead of the other way around, it just seems like a lot
> > > of trouble for nothing. I suppose it's good for sharing files
> > > between the two systems, or using the PC to download DSK files
> > > then install on the CoCo. I suppose those who have a CoCo and PC
> > > both set up at and in use at the same time will find it more
> > > useful though. The PC can easily serve the CoCo while still doing
> > > other things.
> >   
> 
> The idea is to avoid too much extra hardware, and yet still allow the
> PC to serve up files to the CoCo...<snip> Perhaps the CoCo has been
> set up with a Drivewire/CoCoNet ROM of some sort, and can boot
> directly from the serial port.
> 
> In the case of CoCoNet, you could even have your CoCo boot directly
> from nightly Nitros-9 builds put up by someone on the internet,
> theoretically.
> 
> As for an emulator, even a dedicated one, with CoCo "compatible"
> hardware, it still wouldn't be a real CoCo, and I don't think the
> enthusiasts of real CoCo systems would be interested. The people who
> are willing to use an emulated CoCo will just use a PC with an
> emulator.
> 
> Not to shoot you down or anything, I'm just trying to explain why this
> is probably a preferable solution.
> 
> -- 
> Frank Swygert
> Publisher, "American Motors Cars" 
> Magazine (AMC)
> For all AMC enthusiasts
> http://farna.home.att.net/AMC.html
> (free download available!)
> 
> 
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco