QL replica

[Derek_Stewart]

Hi Peter,

SMSQ/E does not have the use of Slave Blocks, pity it could be run on a QL from rom, replacing Minerva or QDOS without a Super/Gold Card.

[Peter]

SMSQ/E does not have the use of Slave Blocks, pity it could be run on a QL from rom, replacing Minerva or QDOS without a Super/Gold Card.

SMSQ/E for the QL would indeed make some sense, but the SLAVEing problem with (otherwise) fast mass storage is the same under SMSQ/E.

[Wasa]

This an interesting thread, though I'm a little lost, as I can read again, but skim reading. I had also written something up to post about a modern representation of what could have been done with the QL myself, but the login authorisation too so long other things came up and I got sick. I want to look at it to see if it's worth posting. I've determined not to do it myself.


viewtopic.php?t=4820


On this thread. I may have a few suggestions. I found it googling Smallest Sinclair QL. It occurs to me that the QZero could be a basis for such a machine, with the modifications I mention in my thread. The Commodore 64 HP serial bus was ahead of it's time, a space that USB and Thunderbolt now occupy. But, such a scheme was viable on more condensed logic than the previous generation Commodore attempt, and healthy profits for the company selling the peripherals. So, a QZero plus USB, HDMI over USB SD card and basic sata (or whatever replaced it) and few other interfaces, represents a complete system in the space a pi zero W would take up. In today's terms there are a lot of open hardware out there, and graphics gpu softcores, with their own drivers. But, all that is needed for a cheap machine is the USB soft cores drivers and a few more graphics modes (and that blitter).

Another idea, that might be of help developing drivers and apps, that I had years back, was to vsee things on another platform which offers services. Android uses linux and the hardware drivers which come with that. I understand the universal driver attempts on Linux failed, but android's core exists. Linus got his start from owning a QL, and a android Linux core, would be a natural relationship.

For software, a JavaScript engine, like Firefox OS and Chrome OS did, but local installed desktop applications instead, to maintain web compatibility and offer a way to write and get cross platform applications. When JavaScript is not good enough, native code is used (there is also a way to run Linux apps through a framework on JavaScript).

Using these approaches, after the original system is implemented existing drivers and apps, hardware soft cores, USB libraries and whatever services attached to the Linux kernel under android, can be used from elsewhere.


I had also been looking for a Sinclair QL on Pi Pico emulator. I have found various 8 bit Sinclair, Jupiter, Commodore, Atari systems, and the Nes. But, the QL is not far off of an 8 bit system, and does the Pico use internal 32 bit data which would make emulation easier. A high speed method of emulating systems, is to translate the code to local machine language. Except for lack of graphics output, the Pico offers a reasonable minimal target system. I do not like the whole stick a Pico on a base board to make an emulated system, when they could be using the Pico and Pi chips themselves as the basis of the base boards.

Manufacturing and legal protection, certification, IP, warranty, distribution are very complicated before you even do the business. There must be a better way to get this all done, to take it off of our hands.

[Wasa]

- QL K/B to PS/2 adapter: i was not aware of this, my opinion is that this interface could be a success. I used a Spem keyboard until it broke, now I am using happily in one of my QL the QIMSI to interface also mouse/keyboard

You are aware this was QL matix keyboard to PS/2, not vice versa?

- Q68 is very close to perfection, I don't want to reopen an old discussion with Peter about the speed of the Q68 at maximum graphics resolution, a solution should be found to reach perfection (cache?)

The Qzero has dual SDRAM onboard, allowing more than 100% speed increase in 1024x768 highcolor mode over the Q68. A new Q68 could re-use this hardware, either by Qzero as a module (faster development) or an all-in-one PCB (cheaper). I diverted to the QIMSI project for a while. But if interest in the Q68 remains, a follow-up is something I would seriously consider.

- Q60, lovable and perfect fast unit. Mine it is still operative and in good shape after more than 20 years of daily operation. I am using the CPLD that Peter gave to me without any major issue.

Thanks for the flowers. But for me, the black screen area is a major issue.

I have been thinking about this for over a decade. And a Q60 replacement remains tempting to overcome the video problem. A board with just 64 MB SDRAM (onboard, no module) and Q68 style peripherals, CP2200 ethernet, etc. would not even be terribly hard to design. But that would not be 100% backward compatible, and the nice old, resource-saving Linux distros would no longer run.
On the other hand, a fully backward compatible high-end solution with 512 MB SDRAM modules, fast ATA, ideal overclocking, etc. would be an insane amount of work. The fastest 68060 mainboard on the planet, 30 years after the CPU was released. Crazy.

Most likely, neither a Q68 nor Q60 replacement would go into the QL case, so we're getting off topic here I guess.

Peter. What about the following suggestion, as the basis of a Sinclair Next (QL+), and to please different people:

For those who don't not want to replace their main board:
The Qzero as the basis of a a daughter board which fits into the socket of the 68008 and at least one ULA? If the CPU can be blocked, then in one ULA?


That this work also form the basis for the new machine? Which can come in editions with minimum interfaces and maximum interfaces?

That a newer higher performance core based on forth architecture also be included as the basis of the new platform for: small business (Sir Clive's dream) and as a retro hobbyist games machine? These forth/misc cores are interesting, small and fast, and outnumber all other language based softcores for FPGA.

Small business does not need most of the power of a modern PC, they used to use Commodore 64's. A new QL could be a cheap alterative which is propped up by hobbyist sales.

I'm looking at a thread on the following alternative:

Alternatively, a Linux QL which emulates the existing operating systems and GUI's? That's a bigger project, requiring an ARM or Risc-V chip. A cheap phone chip and modified Linux for it, is all which is needed to do low end business.