saybur / scuznet Goto Github PK

The Nuvolink emulation is unstable on some systems. There seems to be a general correlation between faster systems and more stability, but this is not 1:1. Known trouble spots:

• Emulation is stable on a IIci running system 7.5.5.
• Emulation is stable on a LC running 7.0.1 (at least during initial protocol development, there may have been regressions).
• Emulation is not stable on SE/30s by @superjer2000 and Chopsticks
• Attempting to mount volumes on a SE FDHD running 7.0.1 causes a divide by zero error.

There is a more extensive discussion about this on the 68kmla forum thread. It would be helpful if other users experiencing trouble commented on this issue.

Per report from superjer2000 cards >=32GB were not formatting correctly.

Both the Nuvolink and DaynaPort devices have support for changing the MAC address at runtime. This is not currently implemented.

Not really an issue... more an observation. In the INI, you have...

[hdd]
Controls the hard drive emulation. This section is optional and may be omitted. Up to four sections are supported: [hdd1] (aliased to [hdd]), [hdd2], [hdd3], and [hdd4], allowing up to four virtual hard drives.

Shouldn't the 'drives' really be hdd0, hdd1, hdd2, hdd3? IoW, shouldn't "hdd" alias to "hdd0"?

It would be more intuitive to let the board create an empty filesystem image if none exists on the card. This feature should be added.

The INQUIRY response may not be spec-compliant, making mounting on newer systems error-prone.

With the FAT filesystem support, it would be nice to support CD-ROM emulation.

The only way to update the firmware right now is via a PDI programmer. It would be better if a bootloader could be installed on a blank chip via DFU, with further firmware updates supplied via the memory card.

Hi,

I went through your repo and you stand in the hardware part that

The '574 is unnecessary and a result of a bad design choice on my part. It could be replaced with a '245.

I'm not sure to understand why. You would need a buffer to latch the data lines, so the 574 seems a good choice. What would the 245 improve ?

Thanks in advance :)

CSV BOM appears to have an incorrect MCU listed (ATXMEGA64A3U) while firmware is for ATXMEGA128A3U

When 'mode=fast' is set on larger images, the check for whether a file is contiguous takes so long the system does not see the virtual hard drive correctly.