EJRS

Also, the STE (Mega STE and I think the TT, NOT the Falcon, ST or Mega ST) had the Microwire interface, this enabled Bass, Treble and Balance control along with YM and DMA sound mixing volume.

See here inexpensive RAM upgrade if you look in the comments there are links to the project. it seems to be somewhere in between "plug and play" and "expert" which you may be able to handle.

I was surprised when I found out that the 2600 had a palette of 128 colours (and good colours), this was in 1977 and would only be surpassed by the Atari 8bit machines (256 colours, 1978) and then not again until the likes of ST, Amiga, Megadrive, VGA, etc. I say "good" colours as most other 8bit machines not only had less colours but also the colours they had were not nice. C64 was a pastel range which was fine for skin tones but useless for explosions, the Amstrad CPC could do explosions but anything else looked garish. Then think of the likes of CGA and EGA, why did they choose those colours! Is there a technical reason these colors were picked, I am thinking readability on poor CRTs or something or less circuitry if colours kept in a range? In fact, if you limit it to 8bit machines/consoles I think the only ones that matched the 2600 and Atari 8Bit was the SAM Coupé released in 1989 the TurboGrafx-16 (yes 8bit CPU and 512 colours) release 87-89. The Commodore 16 and Plus/4 came close with 121 colours. There may be others, but I am not sure and they were not popular machines, the Fujitsu FM77AV40 (1986) apparently had 262144 colours.

I know it is an old topic but I was searching around as I had the same question. So from what I can tell, and I could well be wrong. The 68000 has 24 address lines so can address 16Mb. From what I can tell it technically has 23 and one more line which seems to be more of a upper/lower memory select (a bank select kind of) but it is essentially another address line – not sure why they did this rather than just label it A0. The MMU on the STE only has 20 address lines (well seems to, the pins are labelled funny), so 4Mb. Interestingly I think the MMU on the standard ST only has 19 lines so can only support 2Mb. From what I can tell this is also the case for several of the custom chips in the Amiga so samples, screen, sprites, etc had be in the lower 2mb (I understand later chipsets “addressed” these limitations) although I think the MMU supported 4mb for the CPU. The STE DMA/Chips hardware could address all 4mb. The data bus is 16bit so only a byte at a time is moved, I am not sure if a move.b only does one read from the BUS or if it was forced to do two reads. The 68030 could address 16Gb so I am guessing the Falcon MMU only had 24 address lines, however, it could only have 14mb and the last 2mb were masked off, I suspect this was done DMA or DSP reasons. I am not sure what the TT was capable of, but I think some of the chips were constrained to the lower 2mb. That said it may be that you could write software for the ST were a move.b $000,£ff8240 had the same result as move.b $000,£ffff8240 (as the upper 8bits were masked off) so it could be for compatibility. This is speculation, I have not done the math/research on this. There were upgrades for both the ST and STE to go beyond their limits and I assume they tapped directly into the CPUs extra address lines and then some bodged external MMU logic. Any memory above 4mb (2Mb on the ST) was classed as TT RAM (Fast RAM) so only programs and their data could be stored here (as the other chips did not have the address lines). Now I am unsure if this was technically “fast RAM” as I understand the purpose of calling it this was because when the CPU was accessing this memory it did not have to share the BUS cycles with the other chips, I wonder if this may not be true in this situation. If anyone wants to point out anything I am happy to learn.

I think it came with a couple of GEM apps (ST Basic for example) which technically would take advantage of the Blitter just by virtue of being GEM based.