|
At the time, I was fascinated by the 9900 chip (the CPU, not the 9918 graphics chip), having all 'registers' in memory. At an interrupt, you could swap context completely simply update the "Workspace Pointer" to another complete register set.
The bad thing was of course that register operations were no faster than memory operations, so all things considered, the chip was terribly slow. But interrupt handling was terribly fast!
Sometimes, when I dig up memories of those quite different solutions that were ditched, I can't help but wondering: What if someone took up the ideas today and modernized them? A 9900 model 2022 would of course be a 32 or 64 bit chip, and I assume it would be built with a caching mechanism for the register blocks. Maybe, for some application areas were super fast interrupt handling was essential, and "unlimited" number of threads was a valuable property, that idea of just setting a pointer to the appropriate register block might have some merit!
(I have worked on a machine with something resembling it: Interrupts were prioritized into 16 levels; a high level could suspend a lower one. Each level had its own physical register set. The first instruction of the interrupt handler started executing 900 ns after the arrival of the interrupt signal, which was rather impressive in the mid 1970s, when these small 16 bit minis (PDP-11 class) where developed and sold.)
|
|
|
|
|
Yes, I remember it well. I learned assembly for the TMS9900, and I recall that the architecture of the TI-99/4A was such that the video RAM was separate from the CPU RAM. In order to write to video RAM, the CPU had to set certain memory locations and then perform several NOP instructions to allow the video processor time to read the values that were just written.
This led to extremely poor graphics speed, and in my view was just a dumb design.
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
I don't expect anyone picking up a 30-40 year old architecture to feel obliged to carry forward any or every specific detail in the old design. Like the example you give: That seems to be something we can leave behind.
The important thing is if there are forgotten solutions that might be more viable today than they were when they were left. Even though 99/4A had a poor video RAM interface, still the idea of a register block pointer to memory may be far more realistic today than forty years ago. As far as I remember, at the time of TI-99 cache memory was completely unknown in the microprocessor world (and for some years to come). I am not stating as a fact that cache makes the register block pointer idea a good one, but it might change the situation so that in some context where fast interrupt handling and thousands of threads is essential and far more important than beating x64 on number crunching.
Whenever I tell my 'hope' that Intel would reconsider the ideas behind the 432, to make a truly object oriented processor, someone instantly jumps up: Forget it! The 432 was limited to 8 Ki objects - that is completely useless today! ... So why not make a 432 model 2022 with 4 Gi objects, then? But, I hear you cry, the 432 objects could be of max 4 Ki bytes! (if my memory is correct - don't quote me on that limit) ... So why not let the 432 model 2022 have object up to 4 Gi bytes?
The same thing here: Even though there may be bad partial solution that you don't want to pick up, don't let that keep you from reconsidering other partial solution that might be much better.
|
|
|
|
|
I also had a fascination with the TI 9900 chip. Loved that register concept. I don't collect vintage computers, but I have my TI99/4A in it's original box, with some of the original paperwork, and many cassettes of old basic and assembler programs. I haven't turned it on for 20 years. I think I tossed my TI cassette player, but I still have it's box. So, no collecting vintage hardware, but one piece that I can't seem to let go.
|
|
|
|
|
Yes, that's the one. The selection of suitable graphics chips that are still available in some way unfortunately is not very large.
I have lived with several Zen masters - all of them were cats.
His last invention was an evil Lasagna. It didn't kill anyone, and it actually tasted pretty good.
|
|
|
|
|
Much more entertainment writing the games (even noddy ones) than playing them IMHO.
|
|
|
|
|
refarding that failures list, for me meeting high expectations is not same failure.
never released to market is likly the story of 100 to 1 products never released. Pippin, 64dd, sega cd, 32x.
oh we only sold 9 million, boo, we wanted a billion: Sega Saturn, Dreamcast, VirtualBoy, WiiU,
I saw enough in the world, but still marketcap expectations: ngage, ps vita.
yeah, the paragraph is enough to explain its dislike: ps classic.
for any others, if the companies broke even, well guess what, stop being greedy.
|
|
|
|
|
I put together a GitHub - Kris-Sekula/mgh80: Simple 5 IC Z80 singe-board computer[^] kit, just for the fun of it. I played with it for a while then I got another simple Z80 board, added a keypad, 6 - 7-segment display and created a monitor program for it in assembler. Had a lo9t of fun with it but haven't had a chance to go back and do any more with it.
The most expensive tool is a cheap tool. Gareth Branwyn
JaxCoder.com
|
|
|
|
|
Looks interesting! Must've been like programming a TRS-80!
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
Probably, never had one. Had a Commodore64, that was a lot of fun.
The most expensive tool is a cheap tool. Gareth Branwyn
JaxCoder.com
|
|
|
|
|
I think that's more computing power than Honey the Codewitch uses on a daily basis. 
|
|
|
|
|
That could be true: Not long ago, I was working on an IoT-project (similar to Honey) using an ARM M0 that could be clocked as low as 32 KiHz. Total power consumption is very dependent of clock frequency, and IoT is very dependent on not draining the batteries too fast. This M0 at 32 KiHz had extremely low power requirements. (A common in-house joke was that it could run on the leakage current from the battery  )
I would not be the least surprised if an 1.5 MHz 6509 could do a lot more that that ARM IoT chip. At the same time, I suspect that the factor in power requirement was a lot higher that the factor in processing power.
|
|
|
|
|
You mean like that '80-es ZX81 sitting next to a '90-es Sailor radiotelephone and a '60-es EICO oscilloscope? Yes, I like old stuff.
Mircea
|
|
|
|
|
Beautiful!
"In testa che avete, Signor di Ceprano?"
-- Rigoletto
|
|
|
|
|
Have ZX80 and ZX81 and all sorts of old calculators.
"A little time, a little trouble, your better day"
Badfinger
|
|
|
|
|
I remember this freak. Very cool at the time.
"A little time, a little trouble, your better day"
Badfinger
|
|
|
|
|
A full working ZX Spectrum 48k is in my desiderata. I miss it.
"In testa che avete, Signor di Ceprano?"
-- Rigoletto
|
|
|
|
|
I find it fun to take trips down memory lane, but have neither the space nor the budget for anything other than modern equipment.
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
I have several ZX Spectrums 48K and two Sinclair QLs. I also collect vintage Slovenian (Yugoslav) computers. I write software for them. I know about Vectrex, because I was evaluating the possibility to emulate it on an old Slovenian CP/M machine called Iskra Delta Partner. This machine has a vintage graphical chip that can draw vectors fairly fast so I was hoping to utilize it. However, besides 1K of Vectrex RAM you also need to emulate the 8KB of ROM and the 32 KB of cartridge ROM. And that together is 41Kb. Add the emulator and memory for remembering emulated vectors and ... I'm still thinking about it. But I realistically only have 112KB of RAM available, if I ditch CP/M and take control of the whole banked system.
Anyone here writing software for retro 8-bit machines?
|
|
|
|
|
Did the Iskra Delta Partner have a vector display?
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
Sort of.
It does not have a vector display, but it is driven by (the Thomson EF9367) chip that can only do Bresenham lines and 5x8 chars and there's no access to the raster.
So for graphics you are limited to line drawing (or the 5x8 font which is also drawn by lines and can't be clipped).
The Iskra Delta Partner GDP (graphical model) is quite special because it also has the SCN2674 chip for text only mode and it "glues" the video outputs by both chips into one image and text area is larger then graphics area. Quite exotic.
|
|
|
|
|
Um, the 6502 was a MOS Technology processor. Motorola did make a 6802 micro back in the day.
|
|
|
|
|
Richard Andrew x64 wrote:
Does anyone else like to collect vintage computing devices?
I kept my Commodore 64, my Commodore Pet 2001 with the chiclet keyboard, my Franklin Ace 2000 (Apple IIe clone), and an assortment of green screen and color Wyse terminals for many years. I think it was after my 2nd child was born, and we needed the space, that I finally sold them off. I wouldn't mind having my Pet back. I think that one was my favorite.
Money makes the world go round ... but documentation moves the money.
|
|
|
|
|
But they probably expected you to add more ram in the cartridge if you needed it.
|
|
|
|
|
Wordle 379 4/6
🟨⬜🟨⬜⬜
⬜⬜🟨🟨⬜
⬜🟩🟩⬜⬜
🟩🟩🟩🟩🟩
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
