16-bit C64/128?
in [Commodore]
|
From: Pheuque on 12 Apr 2008 20:12 The only way something like this could fly is if were shrunk to the form factor of a GameBoy. Additionally, it would need at least WiFi as a form of communication. Video output is pointless. in a years time NTSC and PAL will both be dead. A DTV with built in display might find a cool retro niche if it can be made in the >$75 range. A pocket SX-64. But really, no one will take that risk, and with Intel Atom processors coming down the pike @ $30 each, providing an x86 instruction set and processing power equal to about a 500Mhz Pentium, our cell phones will have enough muscle to emulate any 16 bit machine. A real Commodore homage would be a form of Minux or Linux that includes a programing language like Python at power up, and runs on the cheapest hardware out there.
From: Richard James on 12 Apr 2008 20:16 On Fri, 11 Apr 2008 08:37:54 -0700, Harry Potter wrote: > I would like somebody to do what Apple did with the IIgs and make a > C64/128-compatible 16-bit computer. It would have the following > features: > > * A version of the 65816 processor * 16-bit graphics/sound > * hardware-based 1571 burst mode > * compatibility with CMD drives and CBM devices * modular 8-bit > and 16-bit BASICs > * in-ROM compression and text editing * RGB monitor > * math coprocessor > * character and bitmapped graphics modes and sprites > > What do you think? Good idea, or not? Despite the naysayers here that complain about the manufacturers not wanting to make it in bulk has anyone thought you could make a small number of one off units? The cost would be expensive approx $2000 per unit. Each one would need to be manufactured by hand. I would be interested in the future in such a project. I always wanted to make my own computer when I was a kid. I used to design on paper all the specs and draw circuit diagrams. Now that I am over my illness that probably is one of my goals in the next five years. There are a few competitive solutions out there already. Someone has already mentioned the Amiga. I present you with another one. the "XGS 6502" not yet in production but it should be soon as in later this year. However I believe that the "XGS 6502" might be more focused on teaching people how to make computers than as a box that you buy for fun. Also I think the "XGS 6502" is probably lower specced than what you listed. The specifications need special attention. I believe there is a danger of making the system a bit overboard and thus less retro in feel. For example I could easily add hardware 3D graphics support to that list. But is it necessary? Also needed is the actual resolution of the graphics including the number of colours. I believe that as you increase in graphics size you lose the feeling that the system is an upgraded C64. For instance going beyond 320x200 to 640x480 might be a bit too far. You probably don't want to make the sound system using a DAC because that makes the system too powerful. An advanced system like a SID chip is all that is needed. Maybe add some more voices, stereo and hardware mixing and make the output 16bit. I would add more to the ROM than what you listed. I think you could easily embed. * A kernel * A Basic or 2 * A Assembler and Monitor * Word processing, SpreadSheet and Database * Several say 16-25 simple games * File management software ( A Disk Operating System ) * CP/M * Paint program * Maybe a GUI ala GEOS * Networking stuff * C Compiler * BASIC compiler Another question would be what sort of I/O do you want? Do you want a keyboard like the commodore 64 with the symbols draw on them? You probably want 4 joystick ports, Atari style but you could also add support for other types of joysticks. What sort of drive should the system be able to use? A FDD a flash disk, or does it need to be connected to a commodore one? I could go on but I have somewhere else to be. Richard James :) -- sig fail on line -1
From: BruceMcF on 12 Apr 2008 21:32 On Apr 12, 8:12 pm, Pheuque <Ratte...(a)gmail.com> wrote: > Additionally, it would need at least WiFi as a form of communication. > Video output is pointless. in a years time NTSC and PAL will both be > dead. Note that this is not true. It is simple to reduce the world in our mind to the US, Japan and Western Europe, and sometimes Australasia, but that's just a mental model, and not one that matches reality. NTSC and PAL will continue for quite a number of decades to come.
From: BruceMcF on 12 Apr 2008 22:55 On Apr 11, 11:37 am, Harry Potter <maspethro...(a)aol.com> wrote: > I would like somebody to do what Apple did with the IIgs and make a > C64/128-compatible 16-bit computer. It would have the following > features: OK, point by point. > * A version of the 65816 processor Fine, for the reasons mentioned ... get a dirt cheap processor, and the FPGA can just be mapping the processor environment. > * 16-bit graphics/sound Question on graphics is software base. The C64 display modes map inside QVGA, so does the more common SNES mode (256x224, 256 colors, 15-bit color). So keep the "new" display mode simple to map the SNES dipslay into and easy to mix with the C64 display: A pair of flat QVGA framebuffer with eight rows at a time mapped into the memory map is 320x8x1byte pixels, or 10 pages ... there are 30 of these slices in the framebuffer, so the memory map register only need a byte ... five bits for $00 to $1D for the framebuffer slices, $1E for only bring the color map into memory, and $1F for "out of memory map". A sixth bit to select which QVGA framebuffer is accessed. A seventh bit to select which QVGA framebuffer is active. An eighth bit to toggle between C64 bitmapped only display and mixed (greenscreen) C64 and QVGA framebuffer. The color map is 2 pages, colors 0-127 in the first page and 128-255 in the second ... you always want it around when working with the framebuffer, and sometimes want it around when not, so its the first 2 pages of the shadow RAM, and the framebuffer slice the next ten. So I reckon QVGA is the go. And I have had QQVGA in my Zire 31 for a while now, and its just too low res. I've seen a SPI 12-bit DAC, so sound other than something like a SID chip can easily be offloaded. > * hardware-based 1571 burst mode No, leave external floppies in the dirt and move straight to SD flash Ram. > * compatibility with CMD drives and CBM devices No, leave external floppies in the dirt and move straight to SD flash Ram. > * modular 8-bit and 16-bit BASICs Allow cartridges modules to be autoloaded, permitting someone to steal and use Microsoft's 8-bit BASIC if they wish. I don't understand the point of a 16-bit BASIC. lcc > * in-ROM compression and text editing Built in support for accessing a SD file system, including distinguishing which directory is the root executable library for which module, and compression and text editing can be provided in the flash if its been done for that module. > * RGB monitor Leave CRT's behind. > * math coprocessor With a fast SPI bus, this is already available. > * character and bitmapped graphics modes and sprites Programming! The base module works like the C64 display map, except with a toggle one of the colors in bitmappped graphics mode is replaced by the equivalent pixel (high res) or pixels (four color) from the framebuffer. > What do you think? Good idea, or not? On second thought, you want something that fits in the hand, with text entry via a "Quickwriting" system on the joypad and the full range of SNES buttons, mapped in the C64 mode into suitable keypresses ... .... or was this more of a desktop unit?
From: Jim Brain on 13 Apr 2008 00:07
Harry Potter wrote: > I would like somebody to do what Apple did with the IIgs and make a > C64/128-compatible 16-bit computer. It would have the following > features: > > * A version of the 65816 processor > * 16-bit graphics/sound > * hardware-based 1571 burst mode > * compatibility with CMD drives and CBM devices > * modular 8-bit and 16-bit BASICs > * in-ROM compression and text editing > * RGB monitor > * math coprocessor > * character and bitmapped graphics modes and sprites > > What do you think? Good idea, or not? I wasn't sure where to hang this idea, given the splintering of the question into the various factions, so: Although I'm not sure I grok all of the above points, I think there is value in developing something that might give users many of these items at a reasonable cost. As others have noted, support is the key, but I think everyone has been so focused on building a new PC, they have ignored other ideas. Consider: o A new machine that does not emulate at least a 64 in all ways will not be accepted as a replacement. o Emulating a complete 64 is a daunting job. o Such an all-or-nothing proposition scares enthusiasts away. Some would be OK tackling replacement CPU, but feel way out of their element doing video or sound. o Few want to recreate something you can buy for $10.00 on eBay. That's not where the potential excitement lies. Thus, I propose an alternate idea: Develop a very compact 40 pin daughterboard that contains an FPGA, softswitches for passthroughs to a 40 pin socket, and a serial interface to load the FPGA. Develop a second version that includes a small uController that can be accessed via the address/data lines on the IC and can be connected to the serial interfaces of the other boards to load them with FPGA data. (I'm assuming one could use a 40 pin board in the 28 pin SID socket. If not, a 28 pin variant is needed) A regular user can buy and install a few of these boards under existing ICs like SID, VIC, CPU, CIA, etc. Nothing special occurs. However, interested developers can than target these boards to develop new VIC cores, new SID stuff, etc. No need to emulate the original, it's sitting on the passthrough. New functionality can be loaded into the FPGA like today one can write new data to flash ICs through their normal address/data lines. Think about it. You pop in a disk, type load, and the 64 loads in a new SID core and the game to use it. Or, you load in a new CPU, and play around with it. With some work, such boards could be sold for other ICs as well. Let the Atari folks design new ANTICs, etc. The boards could be used (with a suitable core) to replace a faulty IC in a normal 64. No worries about lack of CIAs or CPUs. Yes, there are limits to such a design, and it's not as cheap as a single board design, but I think it is more attainable. It may allow a developer who otherwise would never attempt C64 core to pour their heart into a new IC design. Then, as people gravitate towards certain parts, someone can then take the best and build a new design around them, offering a complete replacement to the community. I can't speak for anyone else, of course, but I know this is the single main reason I am not thrilled to start an FPGA design. I have no interest in a new VIC or SID, but I'd love to try out some new CPU designs. Jim |