thorsinclair wrote:Could it be an expansion card with a FPGA with a more horsepower CPU core, more expansion RAM up to 2 or 4 MB, Quimi was also named, connectors for modern monitors, slightly better graphics in terms of resolution and colors, SD card interface and/or Qubide. What do you think?
All that, and battery backed realtime clock, more ROM were also named recently. Sometimes people asked for ethernet, keyboard interface, sound and decent speed serial also.
The Q68 contains about all the above, just more RAM, and it's physical size is exactly the QL expansion area. For a long time I considered making the Q68 an expansion card.
But once I have an expansion card covering all that, I have a completely new computer. The question is: Which parts of the QL would still be in use then? For me, the original case and keyboard would be the only things left.
In the end I decided to make the Q68 a standalone system instead of expansion card, mainly for two reasons:
1) The board side occupied by the QL expansion connector became free for other connectors. This way, there was room for two normal SDHC card connectors, some LEDs, and sound output. Personally I do not like handling the extremey small microSD cards, especially when using the cards for data exchange.
2) The Q68 is small and has very low power consumption. As a standalone system, that makes it easily portable, simply using a 5V rechargeable battery pack.
I'd find it interesting to know how many prefer the Q68 as "standalone" and how many as "QL expansion"
There might even be a compromise: Putting the standalone Q68 inside a QL and connecting it internally by I²C or via the SPI (e.g. by the ROM socket board from QL-SD which already has an extra SPI). That's fast enough to transfer keyboard data.
thorsinclair wrote:Especially when it comes to a FPGA CPU core I guess you've made some reasonable experiences with the Q68. There is a lot of talks around in the net about CPU cores but I think the implementation is much more difficult.
CPU cores were difficult to debug for QL code usage, but Richard and me did that already. Hopefully to a point where no serious misbehaviour of the CPU is left. Ironically, a different FPGA board already used the results before the Q68 could be made available. Nevertheless it looks like the Q68 still has major points. So time and health allowing, I try to find a way to make the Q68 available for others. This year I received great support from Wolfgang Lenerz, the maintainer of SMSQ/E. I'd like to finish something, before I report more details.