Atari 64? Commodore 800?

[Allas]

Speccy has the better technical innovation software, they developed the isometric adventure games, software sprites, realistic mathematical events.

[cas]

Go for keyboard handling as per C64 as the A8,

i.e. control over the hardware to scan keys, rather than

the A8's last key pressed sort of thing.

 

But Console keys or Function Keys? The A8 win this.

 

Or do I get something wrong? I'm not so familiar with the C64, and how things work there,

 

 

Ok, yes, I got something wrong. The C64 can detect multiple keypresses at once, the A8 only one keypress plus Control and shift and console, but not, for example Key "A" and "D" pressed simultaniously. Nice to know.

 

Carsten

[carmel_andrews]

I wouldn't take anything from the c64, both machines have their good points and their bad points

 

Sid is ok, but there again so was amy/amie and atari coin op's 'quad pokey'

 

VIC 1/2 is so so, but there again these chips will only work with MOS based processors

 

If Atari wanted to have extra features on PAG (pokey, antic, gtia), Atari would and should have built those extra capabilities into their speciality chips, they didn't because their research at the time (and subsequentally) didn't warrant that sort of approach

 

To prove my point, how many software houses did 64 or 128k Atari 8bit games...very few if memory serves me right, you would have had the same problem if Atari came out with newer versions of the PAG chipset

 

That is why if Atari had gone ahead with the amiga (or their version of it) Atari were only planning on doing a 16bit (68000) based version, and not the original 650x version that Amiga had designed (along with a 68xxx version)

[TMR]

To prove my point, how many software houses did 64 or 128k Atari 8bit games...very few if memory serves me right, you would have had the same problem if Atari came out with newer versions of the PAG chipset

 

Erm, i think you're missing the point here... it's just a bit of fun, nobody's actually trying to build these hybrids or market them.

 

Not every upgrade path was taken and the Commodore 128, as another example, was equally under-used even though it has some unique features that could've been put to far more use. On the other hand, the 128K Amstrad CPC and Spectrum 128 hardware were both supported quite extensively (the CPC Plus hardware less so, sadly) so it's not always the case that an upgrade will remain idle.

[cas]

To prove my point, how many software houses did 64 or 128k Atari 8bit games...very few if memory serves me right, you would have had the same problem if Atari came out with newer versions of the PAG chipset

 

Erm, i think you're missing the point here... it's just a bit of fun, nobody's actually trying to build these hybrids or market them.

 

Not every upgrade path was taken and the Commodore 128, as another example, was equally under-used even though it has some unique features that could've been put to far more use. On the other hand, the 128K Amstrad CPC and Spectrum 128 hardware were both supported quite extensively (the CPC Plus hardware less so, sadly) so it's not always the case that an upgrade will remain idle.

 

It's a matter of support from the OS or Programming Languages. The A8 OS is very modular, and new extensions mostly work "out-of-the-box" when well designed, even with Applications never intened to be used with such new extensions (like SIO2PC, SIO2USB, etc.).

 

Unfortunatly, the A8 OS has no API for memory management, or accessing the external RAM other that through a ramdisk. Because of the differnt RAM Extension solutions, each programm must include the test for external RAM and all routines for all supported types of RAM extension. A simple but standardized driver extension like moving a page (256 byte) from normal ram to extension and back would probably enough. Sadly, such extension to the OS was never made.

[Rybags]

Too right about the OS.

 

The 130XE has almost a straight carry-over of the 1200XL OS.

 

The key IRQ routine still has the F-key logic, which stores to PORTB. So, adding the function keys to a 130XE would be a rather pointless mod since pressing them would mess about with extended RAM.

 

I still stand by my previous comment that they would have been way better off scrapping the pointless Self-Test and second character set. That would leave 3K of space. You could fit a half-decent resident DOS in that area.

 

Plus, a nice addition to the XL could have been GR.9-11 with text window (via DLI). They took the trouble to have fine-scrolling for the "E:" handler, but left other stuff out.

[Fröhn]

100% Atari 800.

Apple II computer to hold up my monitor.

Commodore 64 for a foot rest.

I used an apple IIc as a mousepad for years.

And Speccy as doorstopper ofcourse

LOL! And though 16 bit, I've used the Mac Plus' as a step stool from time to time.

But Speccy is the better doorstopper, it's rubber-keys prevent it to slide on the floor when the door is slammed too hard.

[emkay]

But Speccy is the better doorstopper, it's rubber-keys prevent it to slide on the floor when the door is slammed too hard.

 

 

Well.... You know what is Breadbin is for ?

[DamageX]

Fujitsu FM-77AV

320x200x4096 graphics mode

 

Sony HBF1-XV

BASIC

floppy drive

 

Sony HBF1-II

keyboard

sprites

 

PC-Engine

7MHz 8-bit CPU

 

or the Hitachi 64180 CPU from the JVC HC-95

 

I just remembered there was an 8-bit computer with a YM-2151, the Yamaha CX5M, so that is my choice for sound chip.

Edited June 22, 2006 by DamageX

[brpocock]

16-bit chip refers

The SID, 6502, etc can only access 8-bits at a time. True, it can scan from 65,536 locations, but only 8 bits at a time.

 

Hence, 8-bit.

 

Vice versa, the SID can be adressed with 8 bits, because the CPU can only set 8 bit registers. But inside the chip, 2 of them are combined... remembering "16 Bit channels".

That's the cause for my comparision.

 

Well... the SID's address bus is actually 5 bits. It's just used for memory-mapped I/O. Then you have the VIC-II (C=64/128), which has a EDIT:Typo six-bit memory-mapped I/O port, but a 14-bit outbound address fetch availability.

Edited June 21, 2006 by brpocock

[brpocock]

Technically, what is the better mode to design games, c64 or c128 ? (sorry for the out of topic)

 

Short version? 128 mode, if you're targeting the 128, since all of the hardware is available: virtual memory mapping for multiple zero pages and stacks, dual monitors, dual CPU's, and the rest.

 

I'm kinda amazed that more people don't use the Z80A CPU to boost the speed of some types of operations on the 128.

[telengard]

I'd take the SID from the C64 as well as the chroma/luma output. From the Atari I'd take an 800 keyboard and 4 joystick ports.

 

~telengard

[MacbthPSW]

I'm kinda amazed that more people don't use the Z80A CPU to boost the speed of some types of operations on the 128.

The C128's Z80 only runs at 2Mhz. A Z80 needs to be clocked around 4 Mhz before it has any clear speed advantage over a 1 Mhz 65XX processor.

[Heaven/TQA]

can the z80 and the 8510 run in parallel?

[Fröhn]

can the z80 and the 8510 run in parallel?

Since there is only 1 bus, no.

[Heaven/TQA]

so Z80 only for CP/M or how can you use it?

[Fröhn]

so Z80 only for CP/M or how can you use it?

You can simply switch the CPU with a register. However, since the Z80 is so slow nobody ever uses it except for CP/M software.

[+warerat]

I just remembered there was an 8-bit computer with a YM-2151, the Yamaha CX5M, so that is my choice for sound chip.

 

I'm in the process of interfacing a YM-2162 OP-N (Sega Genesis vintage) to the Atari 8-bit-- similar to the YM-2151 but it doesn't need the external DAC, the output is direct analog stereo. 6-channel FM stereo with one channel available for 8-bit PCM samples. Only thing is the sound clock won't be at the exact same rate of the donor machine it is coming off (7.16MHz vs the original 7.67MHz, I'm using the Atari's 14.31818MHz with a divide by two instead of the Sega's 53.6931MHz divided by seven). Don't know if the overhead makes it feasible to run this on a 1.79MHz machine, but being there are a lot of 1st gen dead Genesis machines around, I think it would be fun to get it running.

 

Really hard to find a datasheet for this particular one. To test it I plan to use some raw VGM files to send data to the registers to see if I can get some noise out of it...

[Rybags]

But how does it get it's samples? Does the processor load each one in? Does the chip have an onboard buffer? Or does it have to use DMA?

[DamageX]

Cool, I didn't know the 2612 had analog output. I'll have to see if I can find one.

 

There is no DMA, I think the CPU has to write PCM data one byte at a time (which is basically what the Z80 in the Genesis is for). On one of my PCs I have a programming manual for the 2612, I'll try to find it.

 

I'm not that familiar with .VGM files but .GYMs should be easy at least.

[ijor]

16-bit chip refers to the data bus not the address bus.

 

I know. But there are real 8 bit cpu's around. With 8bit Adress - and Data - Bus. Intel built CPU's with real 16 bit.... And Motorola built the 16/32 CPU's (as named within the ST label)...

 

The bus width (data or address) doesn't exactly qualify the chip. It is not the bus what decides the "category" of a given CPU. The 68008 has an 8-bit data bus, but nobody would call it an 8-bit CPU.

 

It is the "natural" word size that it usually the most important factor. But also are the generic power and capabilities. Again with the 68K as an example, I don't think it is the 16-bit data bus what makes it a 16-bit CPU. It is more about the ALU, which can only operate in 16-bit quantities (it needs multiple ALU cycles for operating with 32-bit quantities).

[+warerat]

 

I'm not that familiar with .VGM files but .GYMs should be easy at least.

 

Thanks for reminding me about the .GYMs. Looks the the GYM format is simpler than the VGM, and for the sake of seeing if the thing works, perfect for this. Looks like something an Atari could easily digest running in a VBI every 1/60th of a second for the timing...

[+warerat]

But how does it get it's samples? Does the processor load each one in? Does the chip have an onboard buffer? Or does it have to use DMA?

 

Most systems that used this chip had a dedicated Z80 CPU to talk to it, in the case of the Genesis the Z80 had 8K of SRAM to buffer data. Unfortunately, I don't have the actual programming info on how to use the PCM mode on this particular chip, but what I've read with similar designs is you sacrifice one of the six channels to turn on the DAC and the CPU dumps raw 8-bit data to a port. The chip itself is a 24-pin IC and has four registers and an IRQ line, but the interrupt isn't used in the designs I've seen.

 

Not a bad sounding chip. I feel less inclined tearing out POKEYs and SIDs from dead machines than I do mass-produced Yamahas..

[DamageX]

OK here is a Genesis programming manual which includes a fairly detailed description of the YM2612. It's an MS Word document and tends to crash wordpad for some reason but hopefully you can make use of it.

 

http://www.hyakushiki.net/junk/sega3.doc

[Fröhn]

While talking about samples: The SID was first supposed to be a wavetable based soundchip, but due to cost reduction they decided to do a few simple oscillators instead.