Commodore is back!

[Caronte3D]

I never had a Commodore 64, but I've had every Amiga, from the 500 to the 4000 one.
I wish they the best.
https://www.youtube.com/shorts/J-GZBvIneQU

[TI-994A]

With all due respect, I don't quite understand the point of this reboot.

Commodore made some pretty amazing computers back in the day, as did Texas Instruments, Radio Shack, Sinclair, and the likes. Those who were privileged to have lived through that wonderful age won't soon forget it, and the nostalgia will continue to burn in our hearts.

However, while I still treasure my original TI-99/4A home computer, I have no practical use for it. I have taken it out from time to time and fired it up just to run some lines of code and hear it speak a few words (mine could talk ). But fun as it is, the novelty simply wears off, and back into storage it goes. I certainly wouldn't buy a brand new version of it now even if it offered enhanced capabilities.

The truth is, the magical memories of those good old days, as we tinkered with our wondrous new toys, making them do things as we typed in instructions, were quite incomparable, and that is what most of us continue to long for. Buying a brand new relic might conjure some of those nice memories, but that's about it.

Unless this reboot is able to carve a considerable hobbyist niche, I don't see it going far.

Sorry if I burst any bubbles. Just sharing my two cents.

[spikey]

I can see a use. I see a problem periodically in programmers who've only learnt high level OOP languages like C# but nothing like C++, and in fact even C++ only students are not immune to the problem if they've only studied the language in isolation. They've no idea whatsoever how a microprocessor actually works. This can lead to a number of blinkered thinking problems and even seriously flawed code in the wrong circumstances. I've more than once had to explain how there aren't really any objects kicking around and, sometimes, it was hard work.

Even though I learnt Z80 and 6502 series MPU's at college and some 68000 series in my apprenticeship I still find the Intel instruction set mind-blowingly complex and impossible to really comprehend in the way that I used to be able to program other systems. I have tried to study up a couple of times, the learning curve is incredibly steep and the documentation set is truly enormous!

Simple instructions sets on slow processors have a learning use case to ease the path into real modern CPUs which will be necessary for hardware and kernel level implementers. I would hate to try and teach a new class the Intel instruction set as their first CPU, I'm sure it would be really hard work for both lecturer and student.

[miso]

There is a live community around the retro machines. New games/demos are coming every month to C64, Amiga, NES, SNES, Plus4, etc.
They love to work under the limitations of these computers for fun. When one says something can't be done on this machine, and the other: wait, hold my beer.

[Caronte3D]

...When one says something can't be done on this machine, and the other: wait, hold my beer.

This is the essence
Anyway, I think like TI-994A, I loved the Amiga, but I wouldn't use it today.

[Piero]

Cuttin' edge audio cassette "duplicates"!

[miso]

I don't know whos interested in this little movie from my little country.
The title is : Demoscene - The art of the algorythms.

https://www.youtube.com/watch?app=deskt ... RkZcTg1JWU

It's in english, but theres people talking from all over europe in it. Has subtitles for 15 languages including english french german italian russian polish etc.

[Piero]

this little movie from my little country.

https://www.youtube.com/watch?v=tqX6Edi2kTY
Caribbean counter-electronics

PS: Electronics example:
https://www.youtube.com/watch?v=5imD1rhxGRo

[BarryG]

With all due respect, I don't quite understand the point of this reboot.

I grew up with a VIC-20 and C64 and know them very well (and from a coding point-of-view, too). I agree this is pointless reboot, done for nostalgia only. It's not going to go anywhere. Hobbyists, collectors and enthusiasts all prefer to use/restore the original machines. All the software for it is not even commercially available anymore, so it would take a whole slew of new software to get it going, which I don't see happening. I see it as no different to releasing computers with DOS only, and thinking it's going to become relevent again. Um, nope. Snowball's chance in hell.

Buying a brand new relic might conjure some of those nice memories, but that's about it.

I bought the C64 Mini (and also "Maxi") for my memories (https://retrogames.biz/products/thec64/). But, to be honest, I mainly just use VICE on my PC when I need my nostalgia hit (https://vice-emu.sourceforge.io/). It's hard for me to see a real C64 becoming popular again when the "Maxi" does the job already (it has a real keyboard and the unit is the same size as the old real C64).

[Piero]

[Olli]

It's hard for me to see a real C64 becoming popular again when the "Maxi" does the job already [...]

(personnally, I am not an old user of the CPU 68000, or other... I really discover the asm through the 8088, and I keep a 8086 CPU chip for possible future pedagogy.

There is however a very very big force in restoring old languages, or old technologies.
Today a clock cycle is very cheap (electric power value in Watts).
First very big force : we actually can put 256 old CPU cores into a recent chipset, so that less than 1% or ressource (in WATTs) of a recent CPU would be used.
Second very big force : the cost of a RAM memory cell (8 bits, 16 bits, or other) is very very low.

We are talking about millions times less expensive.

Why has Intel won in the 80's ?
-> byte code 0x0000 = ADD [BX+SI],AL
Because Intel has understood the motherboard and the devices have the biggest costs : they focused their research into the CPU 16 bits in order to compress the software functions (machine operations).
As you can see

Code: Select all

0000  ADD [BX+SI],AL

, just a op code 16-bits zero, get the sum of two registers, use the result as a 20-bits pointor, and calculates the final sum between AL and the resulting pointor.

In the others CPUs, a register has a identifying code. On Intel, you also have this translation table (code=register), but you have also a set of sub-instructions, whom an electronic engineer knows the physical advance of the getting the value.

Code: Select all

0000  ADD [BX+SI],AL
; this means *mem\a + volatileValue
; knowing that *mem can be got, before the real addition.

What it happens is, if you have any non explicit instructions, then you cannot use the 8086 as a 68000 which will allow the user to treat all the registers, near in the same manner.
On a 8086, each register, will have a virtual function which won't never be able to be changed.
User accu = AX
User counter = CX
32 bits math result = DX: AX
20 bits source memory address DS:SI ( implicitly = (DS<<4)+SI)
20 bits destination memory address ES:DI
Direction flag = Source and Destination Index (SI and DI) automatical change sign


On a 8086 (or its sister the 8088), all the indexing operations are not executed when the user need their results, but, before, when the registers had been modificated, because, physically a sum, it is very quick. A sum, it is physically 6 transistors per bit. On the begin of the 8086, I imagine an integrated transistor swaps a logical status every 10 nanoseconds. What it lets you imagine all the pointers values sums (and fixed bitwise shifts) could be pre-calculated and ready before knowing which instruction will be executed every microsecond.

An other technical advanced point of the 8086, is the total (software as hardware) compatibility with future DMA technologies.

Code: Select all

REPZ MOVSW
;copyMemory()

So if the memory is quicker than the CPU, that is was the way for the 8088 (a poor 8086 whom the data bus has been reduced by 2), providing the addresses was the main task. So the 8086 cpu was a very good memory copier.

I think the engineers of the 68000 did not know everything about the 8086. If they knew, they would hit the 8086.

Instead of grow the bits range from 16 to 32, Intel kept the 16-bits ALU, dedicated the registers, to respond to three demands...
1) an internal algo executed quickly
2) a communication without time gap
3) (1) and (2) can be executed at the same time
....And reduced the data bus from 16 to 8 bits to respond to an established market.

An other thing : why twenty bits (20 bits bus address on 8086) instead of 24 bits (68000) ?

Answer -> to align 16 * 8086/8088 motherboards in a orthogonal network.
This gives an equivalent to a good 80486. For civil embedded electronic, this was a breakthrough.



Restoring an old technology is as taking again the qualities of the past with the qualities of the present.

Note that PureBasic has this characteristic...

[Skipper]

the joys of zero-page addressing - I had almost forgotten about it....

[SPH]

I programmed on the Amiga using the basic programming language "AMOS." It had such an impact on me that I still use goto and gosub with PB!!
And I wouldn't change it!