Interested a modern QL clone concept: Programming the QL's sound, interfaces and microdrives?

[Wasa]

Since writing this, I have decided not to attempt to arrange this myself, due to illness. But, have been interested in a another machine, which can run an emulator, which can be based on the same compatible circuits.

Maybe the following will inspire people. We know that Clive was talking about doing a wafer scale spiritual successor to the QL with Unix or Linux. Maybe we should look at doing the same, a reboot, a QL Next, which people could be attracted to as a hobby, but also a regular machine useful as a small business machine, which fits with Clive's vision. I'm going to post a link over to the QL Replica thread I have just found, where people are already discussing what to do:

viewtopic.php?p=57167


Hi. I'm a previous QL owner, and am investigating a QL clone concept, of what Sinclair could have done back then, and am doing some basic research to see how doable it is.

It is along the lines of the present machine, but with all the extra hardware logic stripped out and replaced by serial interfaces, which today would be represented by a subset of USB peripheral support. This includes the 8049. But it occurs to me, that programs might be dependent on directly coding to the 8049 microcontroller to control the sound and interfaces, and not just using registers and calls, which can get handled by a software routine? Making it not suitable to be replaced?

If anybody is interested. I'm interested in a new slanted keyboard with slanted track pad in place of the micro drives, where a 3 inch floppy option would have gone under the track pad to the side. Even a second one in place of the expansion card port, for software, instead of a hard drive (for business configuration) moving to a dual slim line drive under the pad eventually. This would be USB ports today. A basic unit might have had a high speed micro cassette, as used in answering machines, instead, for low cost gaming. Compiled basic to machine code. API's like white lightening had for software sprites etc, which the latter compiled basic versions had, are an idea. Unfortunately, I don't remember the name of these basics, but one way Blitz basic I think and one may have been by ocean. In the day, such a person could have been contacted to develop such, as the replacement basic for QL, making coding a lot easier for people who aren't Jeff Minter (nice guy, as was Clive).

Similar form factor and styling to the QL, maybe Wafer Scale QL replacement concept styling.

Better graphics (palletted 16 colour mode, with 80 column 16 colour text mode with programmable characters, maybe a simplified dma blitter (compositing the image at high speed rather than sprites) moving to 256 colour 256 and 512 resolution width modes and different resolutions.

VGA is more standard, and able to be fit on a NTSC monitor, and widescreen VGA will scale well within a FullHD screen. But derivatives resolutions going up to 1024*512 are more compatible with PAL widescreen. Today we talk about fullhd and 4k and 16 HDR bit professionally, but memory register hooks in there for all these then future derivative resolutions would be important, even if we are only looking at up to widescreen VGA to start.

A few other game friendly graphics configurations, full 16 colour programmable characters. Maybe a 68000. Better sound.

Emulated CPM and Spectrum, with code translation to 6800x machine code. (Not all games of the day would work, because of self modifying code).

Sort of alternative to the CPC, and cheap alternative to the Amiga and PC.

A TTL version as proof of concept (suitable as a kit), and as a design basis for the FPGA or CPLD's, which represent the ULA's back then. The extra circuitry for the dma transfer circuit that is used for blittering, is not much. Altogether the circuit might be less complex, except the disk drive controller.

[Wasa]

So, are there any of these software dependencies, or can the 8049 support be dropped from the memory map? I'm not a QL machine code programmer, but to get an emulator done, I need to know about system dependencies in code? Even if systems provide standard routines to control features, programmers will still bypass them and program the hardware directly, breaking portability.

I have decided to arrange a new separate machine with option for an QL emulator. Even if on a raspberry pi Pico. But, Peters QZero is more like the sort of FPGA solution I should look at.