Thursday 23 April 2015

Introducing the MEGA65 Retro Computer

Hello all,

For a few months now I have been working behind the scenes with the good folks at m-e-g-a.org, exploring our mutual desire to create a physical 8-bit computer in the spirit of the C65, but that is open-source and open-hardware so far as is possible, so that the community can sustain, improve and explore it.

Basically, we agreed that we wanted to do this, and that the C65GS was the logical basis for this, and thus the MEGA65 project was born, to take the C65GS core, to work together to improve it, and plan towards creating a physical form that is strongly reminiscent of the C65 prototypes.

Our initial announcement is online at http://www.m-e-g-a.org/mega65-introduction/.

This is also the point at which we are ready to offer pre-prepared FPGA bitstreams.

These bitstreams are still experimental, and updated bitstreams may break things that were working in previous bitstreams as we progressively spiral in on the final product.  This process might take another year or two depending on what support we can raise, being done voluntarily in our discretionary time as it is.

What we have decided to do to help support the project is to ask for a donation of your choosing to be given access to our FPGA bitstream build server that will contain the latest bitstreams.  Our desire is not to exclude anyone -- which is why we are not dictating a donation that is in any way representative of the costs of undertaking this project.  Also, everyone will still be free to compile the bitstreams themselves, but we hope that you will appreciate both the convenience and opportunity to support the project.

Finally, we now have a Twitter account (@Mega65Retro), and Facebook page (https://www.facebook.com/MEGA65RetroComputer) -- so feel free to spread the word.

58 comments:

  1. Great news !

    I'll make a donation for sure.

    ReplyDelete
  2. Same here! Any intention of putting it all in a C65 like case with real physical disk drive as media input/output?

    ReplyDelete
    Replies
    1. Hello,

      This is exactly the plan. It may take us some time to get there, but we want to offer a complete brand-new 8-bit computer. While we plan to include a disk drive, we don't really expect people to use it much in practice.

      Paul.

      Delete
  3. I have donated to http://mega65.com. I do not want to build the machine at home I want to get it from http://mega65.com itself. I want that c65 keyboard case, i want that physical real disk drive...I want to get it from http;//mega65.com. Would you provide in future the cost and will I be able to BUY mega65 from http://mega65.com?

    ReplyDelete
    Replies
    1. Yes, our plan is to design every part of the product -- which will still be open-source -- and offer complete units for sale in some way. We won't know what the price has to be until that time.

      Paul.

      Delete
  4. That site got me stoked.
    I really hope that the c65 style casing becomes a reality - injection molded full size with keyboard, keycaps and all, and not just as a 3d-printed project (yes, i know tooling is expensive).

    Regarding the motherboard, It may be possible to implement the system on a sodimm artix7 based module as shown in the link below. That will make it easier to upgrade the fpga like you would do it with a cpu module.in the future, instead of replacing the entire motherboard of the mega65.

    http://www.enclustra.com/en/products/fpga-modules/mars-ax3/

    ReplyDelete
    Replies
    1. We will not be doing 3D printing for the case or keyboard, because we recognise that the result would be inferior. We have some creative ideas for creating the various components that we will need, and have some students working on some parts of that as well.

      For the motherboard, we will likely stick with the Nexys4 boards for several reasons. First, they are not really any more expensive than those SODIMM form-factor ones you have indicated (we need the 100T part). Second, the Nexys4 board is likely to be available for a really long time, allowing us all stability. Third, we are pretty confident that we only need glue logic to extend the Nexys4 board, meaning that the resulting PCBs could be manufactured by anyone at home -- no surface mount, only 2-layer etc. I don't think that this would be the case with the other. Also, the Nexys4 comes with some fun extra sensors on board (e.g., accelerometer -- think pinball games that have a real tilt), microphone and audio out, VGA out, that all make the design process much easier. This is important for a volunteer open-source project. Of course, the community can adapt it to other FPGA boards if they wish.

      Paul.

      Delete
  5. Can I ask a favor? When you make the Commodore 65 case and all of that can you make the most beautiful and elegant LED for the power and disk drive and hard drive and LAN? Can you give the user the choice of the color for their LEDs and make it flashy and bright?

    One other thing, can the user use physical real HD IDE? Can you also add an LED that flashes number of flashes when the C65 crashes, much the same way the Amiga classic flashes number of flashes if there is some issue with the motherboard?

    Also can we have a special BLUE SCREEN OF DEATH for the C65 much like the way Amiga guru's that is fun and pleasurable much like the GURU in Amiga.

    ReplyDelete
    Replies
    1. Hello,

      Thanks for your suggestions regarding the power LED. We will keep this in mind. It will depend on how many spare IO lines we have.

      The FPGA board has a microSD card slot. Given that you can buy 256GB microSD cards, I don't really see any sound argument for the complexity and cost of adding an IDE interface.

      Regarding the BSOD for the MEGA65, we are already thinking about this, and how to make it unique.

      Paul.

      Delete
  6. One final thing...can we have an expansion trap door at the bottom to expand their C65 with more RAM and other expansions in the future?

    ReplyDelete
    Replies
    1. Hello,

      It is extremely unlikely that there will be a trap-door slot because of the way the design works. The Nexys4DDR board has a 128MB DDR2 RAM, that while we don't support it just yet, it planned to be supported later on. It seems to me that an 8-bit computer at 48MHz shouldn't really have need for >128MB. Such an upgrade would require an upgraded FPGA board, and I think that is the best way to achieve this.

      Paul.

      Delete
  7. Cool. I think 128 MB is all we ever need honestly. :) What about the previous request?

    ReplyDelete
  8. I just reread your previous reply and I thank you sooooooooooooooooooooooo much. Sweeeeeeeeeeeeeeeeeeeetttttt.

    ReplyDelete
  9. While the nexys board has a microSD card slot, it would a good idea to place a sd2iec drive unit in the floppy bay, and use the former eject button as a disk change button, when you need to change between side a and b.

    That way you don't have to fiddle with the more delicate microsd card, which just remains inside the machine as a permanent storage you can copy stuff into from the sd2iec drive.

    http://www.retro-kit.co.uk/page.cfm/content/SD2IEC-1541-Emulation-hardware-for-the-C64/

    ReplyDelete
    Replies
    1. Hello,

      I understand where you are coming from, but I don't think it is necessary, because you will have 100mbit ethernet as a convenient way to get data on and off the unit. Also, the internal microSD controller is available to the Hypervisor, so that double-tapping RESTORE will trap to the hypervisor, and allow, among other things, selecting which disk image to mount. Also, we want to put a real 3.5" floppy drive in the drive bay. If we have enough IO pins left, we may allow for an externally facing SD card slot, but I am not yet sure if we will have enough IO pins available.

      Paul.

      Delete
    2. I'm always thinling if some generic SPI bus can be implemented, many devices later can be attached, even eg codec ICs capable of decoding MP3 for a few dollars (this is even done for more "smaller" computers like Z80 based MSX systems) and such. The needed bandwidth is moderate, and I guess even a simple SPI bus clocked at some MHz can be enough. And it does not require too much I/O pins (well, if MOSI and MISO is common for all devices, for select lines, there should be, but an external demultiplexer as a single 74'138 (or maybe '154 for 16 devices and still needs only 4 lines then towards the FPGA) or something similar that can help if it's really a problem). I am not sure if it's good idea though. The other possibility to have some "true" external (parallel) bus, where you can connect additional devices in the "more traditional way".

      Delete
    3. This is a good idea -- it will come down to just how many IO lines we have left. There is going to be a trade-off between IO lines free and maximum speed at which the ports can be read/written. Priority will be given to make sure the disk drive port is good for 1MHz, but other ports might be 100KHz to minimise pin count, so that we can add extras like this.

      Delete
  10. Doesn't the LED feature require only 1 pin IO?

    ReplyDelete
    Replies
    1. 3 pins if you want an RGB LED that can pulse all sorts of colours. However, a simple single-colour LED that only needs to turn on or off at most 100 times per second can be implemented using a shared pin.

      Delete
    2. Exactly! I am not asking for multi-color LED :) Can you attach your own LED on the boards themselves with their different color to represent one for disk drive, one for power, one for LAN and one for error? And the rest are only turning on/off function.

      1) Disk Drive led starts blinking like the commodore 1541 disk drive if there is file i/o error such as write protected, read error,save error, file not found....etc. The user see the disk drive LED blinking like Commodore 1541 when I/O error occurs there.

      2) Error LED will flash number of flashes when there is system error or hardware. It will keep flashing those number or flashes that represent the error code until the user resets the system. If operating system or error or sum such it will enter the famous blue screen of death or whatever screen you do in the future and keep flashing that code. If it is a hardware error it too will flash hardware error code. Again could use sharing pin for that :)

      3) Since this is not Amiga and it is Commodore we want to follow the Commodore line and the power LED is always ON. So it is simply. This LED is guaranteed to be shared no matter what :).

      4) Finally, the LAN LED will blink at network activity like any modern PC would and usually have green colored LED.

      Again, can those 4 steps be all shared? Since we are not using the main board's LED to change colors..we will have physically the colored LED and via core do these things and then in the end it is matter of turning on and off LED at certain situation.

      But the at original post I was requesting if we the buyer could choose the LED of our choosing from the list of LED gallery when we buy our Mega65 from the store? It is like customizing our own car, I want this door, this wheel, this engine..etc. But instead we get the main computer and then have pictures of different LEDs, style, brightness, etc and say ok I want figure 1a go for power, figure 3c gof or LAN etc in the shopping cart and then click buy it now and poof.

      Delete
    3. We will probably take Henry Ford's approach and offer the LED in any colour you like, so long as it is red for the power one, and green for the drive one. The whole design is open, so people can replace the red LED with a plaid one, or whatever colour they prefer. But a stock MEGA65 will look externally like an 8-bit computer from the 1980s.

      There is also a practical reason for this -- if we offer lots of options, they all take time to create, and we are using our spare time just to create this computer in the first place. So we have to be a bit ruthless for the time being. Again, the design is open, so anyone can change bits that they want to.

      Delete
  11. Since the Nexys4 boards are quite expensive (listed for $320), wouldn't this mean that a complete unit would cost at least around $500? I'm not much into the world of FPGAs, but aren't there any cheaper ones with about the same functionality?

    ReplyDelete
    Replies
    1. Hello,
      Unfortunately the only FPGA board with a big and fast enough FPGA for this project are quite expensive -- the Nexys4 is actually at the cheap end for the functionality and connectors we need. Historically, these boards get cheaper over time, and this is what we are expecting with the Nexys4. In the meantime, you can get them for US$160 if you have an academic email address.

      Paul.

      Delete
  12. I've just sent my donation !

    Waiting for my download email :)

    Thanks for everything.

    ReplyDelete
  13. I am 'very' interested in this project. With regards to peoples requests, they seem to be aimed much to wanting this to be a new Amiga, with requests like 'trap door' and 'guru errors' As much as I love the Amiga, I'm pretty sure the aim of Mega65 is more about an open sourced C64/C65 compatible computer for 2016 right? Support for .D64 (and possible .PRG and .TAP too?) excites me a lot, I have one question that concerns me.
    Making new casing for such a computer costs a lot of money, like thousands to produce the moulds and that's before the price for the plastic to create the casing. How do you plan to raise the money for this? via a Kickstarter? Finally my last question (for now ;)) Will you also be selling this computer caseless too with options to buy separate add-ons, ie; Internal Floppy, PSU, extra memory and such. Sorry to bombard you with questions but I think many other people may wish to know this too. I think that's all I need to know right now except maybe the 'blue led power light' is a great idea ;)
    Btw we have posted News about this project on both our websites, http://awesome.commodore.me and http://www.vintageisthenewold.com, my name is Christina (Kitty) I'm the Owner and Admin of these two sites. Please feel free to contact us about any latest updates and we will be happy to promote this project and spread the word :) Contact Email: awesome@commodore.me, here's hoping I will have a Mega65 on my desk next year :)

    ReplyDelete
    Replies
    1. typo: should be molds* ;)

      Delete
    2. Personally I think that C65GS is already much more than "only" a C64/C65 computer. I really like Amiga too, but it's not so much the direct relative of the good old C64, C65 is much more. So some can continue to even develop C65 further, indeed, somewhat Amiga like, but 65xx CPU, basically VIC-like video hardware etc, so still seems to be the imagined new generation C64 even after C65. Like in nature, dolphins look like basically a fish (but they are not) because it's the logical (optimal, near optimal) 'construction' for those animals, it's quite natural, that a "very modern" C64 will have features than other more (than C64) modern architectures, like Amiga, etc. I don't see problem here. Just for having a C64-like computer I wouldn't buy, I want the new features to feel that Commodore-line is not dead, and new computers released in the spirit of the old one though (it's not that to build a PC into a C64-shaped box of course ...). Well, I am not sure I could express myself well enough. As usual.

      Delete
  14. This comment has been removed by the author.

    ReplyDelete
  15. Hi Paul! I am planning to buy the Nexys 4 DDR board in the near future, but I still have got some questions.

    1.) Will (surely) there be any possibility to buy an "empty" system with all the hardware goodies (JUST without the FPGA board) from you later (and not so much more expensive of course than with all together), which I could mount my already owned FPGA board then into by myself?

    2.) The full price of the FPGA board is a little bit too expensive to me. The educational half-price would be rather reasonable though, but I have got no access to such a thing at all. Is there (or will/would there be at any time later at least) any other way to buy it slightly cheaper? (E.g. somehow via ordering from you?)

    3.) Will there be any possibility to add some real SID chips to the system, instead of the emulated ones? (E.g. like with Turbo Chameleon, one could use the original C64 motherboard.) Perhaps externally via any add-on accessory like a docking station etc.

    4.) What about mouse handling? (Micromys emulation planned?)

    Hi,

    Robert

    ReplyDelete
  16. Do not get me wrong I do not want to copy Amiga with crashing message. I want it to copy A CONSOLE WAY OF CRASHING. I want crashing system to be fun and used an Amiga as an example.

    Paul if I buy the board in advance will you provide a kit to setup up the way you guys are setting it up so a person don't have to buy double? The kit will include the case, led, disk drive and manual to setup physically for dummy without the need of soldering hacking and download latest core to meet these changes?

    ReplyDelete
    Replies
    1. Hello,

      Don't worry, we have plans to make the crash message fun. This will not be a problem.

      As for buying a board now, and getting the other parts later, this will definitely be an option. In fact, it will likely end up being the main option, because it will mean that we won't have to have capital sunk into stocking FPGAs people can buy cheaper elsewhere.

      Delete
  17. YES!!! YESS!! YESS! THANK YOU!! Please provide a kit for dummy and have adapters for dummy when it comes to hooking devices, LED, disk drive, etc with easy to read manual with lots of pictures :D :D This way, even if we buy the KIT it will be like buying the complete system from you guys and I am sure everything else including LED control and so on are done via that 4 MB core file itself and nothing to do physically hardware wise!

    THANKS PAUL! Now I am going to buy the actual board without any guilt. One final question, you said we could use the actual C64 case (empty) and an adapter to be hooked with the c65gs, I forgot the name of the adapter. I have one question about it...is there a picture or tutorial showing how to hook the adapter into the fpga board and what about the power supply from the adapter is that enough to run the entire system? Thanks.

    ReplyDelete
    Replies
    1. You need a 1.5A capable USB power supply. The Nexys4 board comes with a micro USB cable.

      The adapter you need is a Keyrah v2 from, IIRC, Individual Computers.

      Connection is just a matter of connecting the USB cable from the Keyrah to the FPGA board after installing the Keyrah in the C64 case.

      Apologies I cannot provide more detailed information right now. We will aim to develop some detailed documentation from July.

      Paul.

      Delete
  18. I am _so_ psyched about this announcement, and the enthusiasm shared by others here makes it really hard for me to not smile with delight! ^_^

    Paul - may I ask how far along you guys are with the case design? Will the floppy drive be positioned towards the user (like the original C65-design), at either side of the computer (like the desktop Amigas) or at the rear of the unit? Oh, and do you plan to maintain full 1581 compatibility? (if you are, you may want to check out the upcoming Sakura floppy drive for classic Amigas - as far as I know, Amiga floppy controllers are fully 1581-compatible, meaning the Sakura manufacturer may be a potential supplier for drive mechanics - http://retroami.com.pl/index.php?id_product=153&controller=product).

    ReplyDelete
    Replies
    1. :)

      We are intending that the case will look as similar to the original C65 prototypes as possible -- including the front-facing drive position.

      We haven't got too far along on the floppy drive design as yet, so we will certainly bear this in mind.

      Paul.

      Delete
    2. I'm just wondering if a real C65 is comfortable to use, ie the floppy drive may disturb some's hand typing on the keyboard? Anyway, that is the C65 design and nice to have some kind of seems-to-be original stuff at least :)

      Delete
    3. My C65 was very comfortable to use, once I stopped the left cursor key getting jammed on the casing. The drive didn't get in the way, and instead was a convenient place to keep disks not in use etc. It takes a little while to get used to the ports being on the left, that's all.

      Paul.

      Delete
    4. Why not think about a compact version of the computer without drive?
      I've made a quick mockup http://s30.postimg.org/r48m0ulw1/c65c.jpg
      An optional drive could be still chain-connected on right side. I got this conclusion because the machine has the built-in SD card reader and/or many people also have a real 15x1 drive (or SD2IEC device) so I think the 3.5" drive is not really necessary. It will also help to get price lower.
      Anyway, if the floppy reader can be connected as a "sidecar" we can almost reach the original design of the machine.
      Just my 2 cents :)

      Delete
  19. Just found this:
    http://scpu.amidog.se/doku.php?id=scpu:doom

    Gameplay:
    https://www.youtube.com/watch?v=G0ct1PjBvMs

    Initially i thought this was some kind of joke, but apparently someone got doom to run on a emulated supercpu system. This may be the most demanding piece of software ever to run on a c64 compatible machine to date.

    I don't know if it is of any use here, but if this can be adapted to run on the mega64, it could be a good tool to test both the new cpu and the enhanced graphics modes.

    ReplyDelete
    Replies
    1. Yes, this is some interesting work. The guy made a MIPS to 65816 translator, and compiled DOOM for MIPS, and then ran it through his translator.

      Something similar should be possible for the MEGA65, although the instruction sets are quite different. That said, my plan is to make a native C compiler at some point.

      Paul.

      Delete
    2. Actually C65 (and/or C65GS) can be useful if Blitter or something like that (I thought about the possibilities on the C64DTV where you can find some faster CPU, Blitter and extra video modes, maybe C65GS would be even better) can speed the process up a bit. Of course the additional screen modes would help. A native port (well "port", maybe more a full new game which looks like Doom on the PC ...) would be better but there is even some doom for ZX Spectrum: a fully rewritten engine optimized for a fast 65xx CPU can have the result of something interesting in my opinion. Oooops: https://www.youtube.com/watch?v=X3Oqz5WjDPI

      Delete
    3. Hello,

      Actually, the MEGA65 won't need a blitter to help, as the CPU is probably fast enough for this sort of thing. Certainly if a SCPU can do it, then I don't see a problem with the MEGA65 doing it, even with the lack of 16-bit registers.

      Delete
    4. Yes, but I mean about a "real" clean implementation of a game like that, which could be much more playable, I think. And if you want to really exploit the possibilities of C65GS with eg more colours and higher resolution the CPU (regardless C65GS CPU is fast) can be a bottleneck. https://www.youtube.com/watch?v=UiECpT7HdK4 This is a C64DTV demo, you can see some raycaster stuff near at the middle of the show. Well anti-aliasing for example can be brilliant but I guess that may be too much ... ?

      Delete
    5. Maybe the main bottleneck would be the addressing modes still limited to 64K. The SuperCPU can use full 24-bit addressing, thus accessing up to 16 MB native memory directly and quickly. In the C65GS it is more difficult to do. For some Blitter-like purposes, the built-in DMA (that is like a REU after all) could do a nice job probably, at least if it goes also on the 48 MHz clock (instead of just 1 MHz or so). Or at least as fast as possible.

      Delete
    6. Sorry to mention DTV all the time, but this is what I know better. It (DTV) can address 64K directly still, and memory is "paged" (4*16K of the CPU address space can be mapped to any of the RAM or flash ROM segment). DTV's raw CPU power is far beyond of C65GS still (basically clocked ~1MHz but an aligned 4 byte memory area can hold code and can be executed within a single clock cycle if skip cycles + burst mode is selected AND if the opcodes does not need any other memory R/W than the opcode fetch itself only - but even this way, I guess a 48MHz 65xx like CPU is faster than this), as I posted the video (about the DTV demo), nice things can be done with it. Also, afaik C65GS supports even 32 bit addressing with a nice trick (just I can't find now where it was something about ... some well organized document can be really useful at this point!) with the ZP, that is probably faster way and "more linear" than the 65816 can do (where you need to use long addressing or reload DP all the time). I don't know about C65 (and C65GS) DMA/blitter too much, hopefully it can do tricks like DTV, eg with blitter, logic operations between two sources, modulo, fractional steps, and things like this. It can be used than for even texturing. Well, IMHO. :)

      Delete
    7. Eh, sorry, my bad English, I wanted to say: DTV's CPU is weaker than C65GS's, "beyond" was a mistake to be used ...

      Delete
    8. I have already done some speed tests to the DTV and SuperCPU emulations in VICE (using my own algorithms), and it seems to me that both of them are very "sensitive" to memory: meaning that they are both super fast actually without any memory accessing being done (just the CPU in itself), however, the more memory operations in the code there are, the slower they will be in average. After all, for my own purposes at least, the DTV in fast mode seemed hardly any twice as fast as the stock C64, and the SuperCPU seemed also only 8 or 9 times faster than the normal 1 MHz operating mode. (While Turbo Chameleon made over 14 times and the C65 in MESS made also over 4 times, wow.) Thus, the C65GS has got a very-very good chance as opposed to them after all, I think.

      Delete
    9. Dunno, but with DTV you need to optimize to have opcodes on the 4 byte memory boundaries if possible in a sequence that ops having no extra memory requirement would fill 4 byte. And average ("not DTV optimized") code on the DTV burst+skip cycle mode would be somewhat faster but not as fast as it would be, I guess. For SuperCPU, I have not so much idea, 65816 has more or less the same timing as 6502 (well, extra cycles if 16 bit mem access is used for example) but the higher clock counts a lot :) Also, since SuperCPU (afaik ...) uses SIMM modules (?) and somewhat complex method to access the built-in RAM of the C64 (also slow down to be able to do it with some kind of cache) it's hard to tell. But I only remember some words about SuperCPU read somewhere years ago, so I can be wrong. Yes, you're right C65GS would be much faster anyway. In my previous "longer" post I used word "beyond" though, when I wanted to use "below".

      Delete
    10. Hello,

      Olessak: Can you make your benchmark programs and results available somewhere? I would like to start collating this information and providing MEGA65 results for each so that we can compare.

      You are all right in that DTV and SuperCPU have significant memory access sensitivity issues. The MEGA65 was designed to avoid these.

      As for accessing large memory spaces, the MEGA65 has two simple but effective innovations:

      1. 32-bit ZP indirect mode, so that pointers can be to anywhere.
      2. 32-bit modes for JMP/JSR/RTS to allow jumps and jsr to go anywhere in memory.

      Together, these mean that you can easily read and write any byte of memory, and also jump to code anywhere. All the bankswitching to implement this is done behind the scenes for you. These modes also support virtual memory (1GB total), so you don't even have to worry about how much memory your MEGA65 has. The hypervisor will handle the page faults for you.

      Paul.

      Delete
    11. Yes, I definitely would like to make it available, yet it is not ready to go at the moment (not the results, but the code). The problem is that it is not a small, self-standing Benchmark program in itself, but the embedded part of a much larger and monolithic other code, which I could not simply cut it out of. The larger code segment is the main booting part of that pre-alpha stage IF engine mentioned in my other comment (that's an over 100K long and close to 4000 lines ASM source together with so many other things...). I am planning to make it public (and open-source) in the next one or two months definitely. If you could not wait up till then, then I might also send it in a private message to any e-mail address, of course, at any time. (I am just right now writing some kind of user manual for that, too, however, it is not yet so ready, either.)

      Although I can describe the algorithm.

      Basically it is simple: I have set up a loop core that exactly takes 1000 cycles on a C64 and repeats itself endlessly (while counting the iterations). I let it run for 0.1 sec (that is 5 x PAL or 6 x NTSC video frames), and after that I look how many iterations have been done. If I get a hypotetically exact 100 times, it induces an exact 1 MHz clock (divided by 100). I can measure it at two places of decimals accuracy in this way. For all VICE emulations, I have already got their exact measurements: e.g. 0.98 and 1.03 for the PAL and NTSC C64 etc. These are the reference. So it is a "virtual" MHz value. I call it only "virtual" here because e.g. for the C65 I already get 4.14 (instead of that 3.54 real MHz) probably since the 4510 instructions take less cycles than the 6510 ones. If you divide it back by a C64 reference: 4.14 / 1.03 = 4.02, then you may say, that a C65 is approximately four times faster than a C64 in average all in all (as for the NTSC ones).

      By the way, the MESS emulation might be false at this point, e.g. when you switch it to the 64 mode with GO64, then the cursor flashes way faster than on a C64, however, as much I saw it on the videos uploaded to YouTube, it does not so on a real hardware. (I have measured 1.18 MHz in slow mode.) That is MESS v0.111 (from 2006), as this version is said to be the last one containing a working C65 emulation.

      Delete
    12. Hello,
      Yes, the cursor being fast in C64 mode is a clear bug in MESS. The cursor blinks the normal rate on a real C65, and also on a MEGA65.

      Let me know if you manage to extract your test routine.

      Paul.

      Delete
    13. I would not like as much to extract it, as rather to further develop that whole thing into a simple software diagnostic tool, for detecting, testing and displaying the most important system components (CPU, RAM, DMA, periferals etc.) on all types of CBM machines. It should be soon done, and when ready, I shall also post a download link to here then (surely in the next few months). I would really like it to contain a dedicated part for C65GS detection and use. One of my next things to do is to write that code part, too, and I am momentarily looking for your notes and infos about it, though they seem to be so uncomplete and scattered at many places now... Is there not a user manual or anything like with all programming infos and how-to-do's?

      Delete
    14. Yes, sorry about that. While the exact specifications are still evolving as the design is completed. IOMAP.TXT and the .A65 files in the git repo are probably the most useful materials.

      Delete
  20. In my opinion and this is opinion here only...that there should be one SINGLE native game made only for the C65 that is not a port or copy of other games on the market to showcase the hardware of this C65. It should be distributed with the purchase of the C65gs.

    This way, people will be enticed to make more native c65 games only available for the c65. Maybe in the future will have gb65.com instead of just gb64.com.

    ReplyDelete
    Replies
    1. This would of course be welcome, and anyone who wanted to write such a game would have our encouragement -- however, for us it is a lower priority than getting the machine working in the first place.

      Paul.

      Delete
    2. That would be then rather like Metal Dust for SuperCPU. Very cool and nice - just nobody plays to that. I think it should be a better approach to aim the most possible platforms instead (e.g. starting with some C64 features only, and then adding as much extra goodness as possible, as optional extensions, to that, while maintaining a backward compatibility).

      I have already started the development of a classical MUD (Multi User Dungeon) like Interactive Fiction engine (in the form of a text adventure, however real-time with a multi-player mode) aiming ALL 8-bit CBM platforms at once (yes, starting from the PET...), of course including the C65 and the C65GS, too. It will be unfortunately also a several years slow job to do, as being a sparetime hobby project as well as so complex and difficult that I am still at the beginning with it...

      Delete
    3. E.g. Doom itself stays also yet so interesting now at most because of being subsequently made on all possible platforms.

      Delete