Atari v Commodore

[Rybags]

I was basing it on a 24 line display.

[Fröhn]

"Rerun on Atari - to be more fair, with a 25 line high text display."

 

You can also skip character lines on +4, but it isn't a 5 minute code.

[Rybags]

That's the thing with +4 and 64.

 

To achieve such things that the Atari can do without assistance costs you time and programming.

 

As such, blanking part of the display is a cost vs return proposition, where on the A8 it costs nothing.

[Crazyace]

The big win on the a8 is the use of the lower res graphics modes, with their lower bus use. That's what allows Rescue to run as well as it does.

[Fröhn]

Atari needs display list assistance

 

On +4 it's fairly easy: two small raster IRQs with 2x LDA STA. And the more char lines you skip, the less "free cycles" advantage for A8 (due to less double-badlines). But if you are happy with 40x25, you don't need anything: No CPU interaction and no display lists.

[Rybags]

Still, it's easier here.

 

The Plus/4 has an advantage for blank screen running depackers, which is about all you use a blank screen for on A8.

[Crazyace]

I find the plus4 interesting because that's how I'd expect an 'antic/gtia+' to operate if Atari had redesigned the 800XL to compete against the TI/Vic20 ( and C64 ) more.

Maybe no direct colour from memory - but 16 pallette registers and a foreground/background colour byte would be pretty easy to implement. ( As well as player improvements, and maybe even better GTIA style true colour modes )

[Crazyace]

I wonder what speed was possible for the 6502 in 1982. The BBC ran memory at 4MHz and the 6502 at 2MHz ( with no wait states ) - An 800XL with 3.84MHz memory and a 3.84MHz 6502 would have ruled ( especially if it have additional graphics modes )

[Rybags]

I can't recall any original 8-bit flavour 6502 of that era at > 2 MHz, even arcade machines.

 

By the mid 1980s it had fallen out of favour since the 16 bit CPUs were affordable and much quicker.

 

It's probably only in embedded applications that later 6502 descendants were ran at a higher speed (of course we had the Apple II gs computer though).

Edited January 15, 2009 by Rybags

[Crazyace]

If the original chip was available at 1.79MHz in 1978 , I can't believe that a 65c02 varient wouldn't run at 2.38MHz in 1982 if it had been requested. ( After all this is wishful thinking anyway )

[atariksi]

I find the plus4 interesting because that's how I'd expect an 'antic/gtia+' to operate if Atari had redesigned the 800XL to compete against the TI/Vic20 ( and C64 ) more.

Maybe no direct colour from memory - but 16 pallette registers and a foreground/background colour byte would be pretty easy to implement. ( As well as player improvements, and maybe even better GTIA style true colour modes )

 

Player movement vertically would have been easy to add to ANTIC if they just allowed user to set the 8-bit initial offset for each player in its 256 byte RAM. So P0 data = PMBASE OR P0Ofs, P1 data = PMBASE OR P1Ofs, P2 data = PMBASE OR P2Ofs, P3 data = PMBASE OR P3Ofs, and P4 data = PMBASE OR P4Ofs. Currently, LSB bits of PMBase are always starting at 0 ofs at beginning of frame. And even this OR is just putting address lines together.

 

Color registers are OK since you get the ORed colors, but each missile having its own color register would have made it better.

[Rybags]

Here we go... Apple II accelerators pushed the boundaries.

 

http://en.wikipedia.org/wiki/Apple_II_accelerators

 

And the Apple IIc Plus ran a 65C02 at 4 MHz.

[atariksi]

...

A good example of this is the way the 5200 was put together. Not made compatible with anything, Analog Joysticks that were hard to center, different cartridge port, and the hardware was mapped in different locations. Plus it cost way more than the 2600 or any game console of that time. If they simply used the same memory map, cartridge port, and joysticks, it would have been cheaper and would have a huge game library ready right away. Its cartridge port could address 32k directly. only big advantage.

 

I cannot say Tramiel made all bad decisions, The XEGS should have happened much sooner to compete with Nintendo. But in 1984/1985, everyone thought the Game Console was dead so Atari was gearing toward computers. It wasn't the Game Console was dead, the 2600 was dead after releasing a poor port of Pac-Man, ET, and a flood of poor games coming from all those different companies. People were looking for better game technology and the C64, A8, and eventually the NES all had them. All it took for one to become popular was a "Killer Application" and that ended up being Super Mario Bros.

 

They did screw up one part of XE/XEGS machines-- the joystick port is tighter so some DB9 connectors don't fit although they fit in 400/800/XL.

 

On the Atari 5200, perhaps they wanted them to have some incompatibilties with their 8-bit computers and I think they were trying to out-do some other analog sticks that had appeared on the market. A case where it's better to do user your own technology than copy others. Analog sticks are definitely inferior to digital joysticks given they were basically porting similar games from 8-bit line.

[Crazyace]

There was nothing amazingly wrong with the PM graphics that I would have changed , apart from adding more of them ( even the Amiga didn't add vertical positioning ) - especially if this was 1982!

As Rybags found a 1982 AppleII card running at 3.58MHz I could even dream of a turbo 800

 

3.58MHz 6502 and ram .... 80 colum graphics mono - everything else double , + c64/+4 style colour attribs. + 8 16pixel wide players ( no missiles, just 8 players )

 

That would have been a cool machine!

[atariksi]

They say Commodore made mistakes with the Amiga. Maybe... Commodore and Atari were home computer manufacturers that failed to adapt to the end of the home computer era.

 

From the marketing at the time, the ST and Amiga didn't play up "home computing" so much even though 99% of both machines wound up being glorified game consoles. Both seemed to be trying to be the "affordable Macintosh" though neither company officially sanctioned the inevitable emulation products. The mainstream PC OS was still DOS and mainstream PCs were still running stodgy menu driven business software and really grotty looking games with that beeper for sound when both machines were introduced. These things came with mice and desktop GUIs. People who might want a Mac but couldn't afford one and perhaps found them boring when not using a WYSIWYG word processor were a conceivable market. Perhaps the real mistake was catering to geeks a bit too much and not so much to the artier types that went for Macs.

 

Come '88 we start seeing PCs with the Adlib board and VGA. Those machines were top end and rather expensive and PCs of that magnitude didn't start getting affordable until 90 and that is when they became mainstream for gaming. I saw Doom and Descent on Radio Shack branded PCs in '92. Games like Myst and 7th Guest weren't far behind. The writing really was on the wall at that point.

 

The time to adapt the Amiga into a PC chipset would have been when the Adlib and VGA were getting common but the mentality just didn't exist. It also wouldn't have made much sense with anything less than a 286 and a 386 would have been better. The window to do that was short and the mentality simply didn't exist to exploit it. Even that would have only bought so much time before it had to compete with the likes of Creative and NVidia. And as just another multimedia chipset it wouldn't have seemed so special running Windows.

 

Motorola did have a bigger market share so it would be hard at that time to decide to put Amiga chipset in a PC. Commodore had PCs that they sold w/o Amiga chipset. Adlib boards had no DACs (for sampled audio) nor were they a standard part of a PC package-- just good for music; Amiga/ST had the edge in audio even with the Adlib boards in PCs. Sound Blaster 8-bit which came later on was more competitive to the Amiga/ST audio although that also lacked multiple channels. I remember the Thinkpad 701C w/butterfly keyboard that had the Sound Blaster integrated in its motherboard.

[Fröhn]

Player movement vertically would have been easy to add to ANTIC if they just allowed user to set the 8-bit initial offset for each player in its 256 byte RAM.

It's not as easy. It would require that every Player/Missile had it's own memory pointer, it's own raster comparator, it's own height counter and everything made available through registers. Quite a lot of extra transistors without the space for them.

 

Color registers are OK since you get the ORed colors, but each missile having its own color register would have made it better.

Again quite a lot of extra transistors.

[frogstar_robot]

Motorola did have a bigger market share so it would be hard at that time to decide to put Amiga chipset in a PC. Commodore had PCs that they sold w/o Amiga chipset. Adlib boards had no DACs (for sampled audio) nor were they a standard part of a PC package-- just good for music; Amiga/ST had the edge in audio even with the Adlib boards in PCs. Sound Blaster 8-bit which came later on was more competitive to the Amiga/ST audio although that also lacked multiple channels. I remember the Thinkpad 701C w/butterfly keyboard that had the Sound Blaster integrated in its motherboard.

 

The point that I was making was that PCs were starting to catch up at that point. They hadn't caught all the way up in '88 but they pretty much had by '92.

[Crazyace]

I read some comment that the Amiga could have been a 128k game console in 1982/1983 - cant remember where though at the moment.

That would have been way better than the 7800

[atariksi]

Player movement vertically would have been easy to add to ANTIC if they just allowed user to set the 8-bit initial offset for each player in its 256 byte RAM.

It's not as easy. It would require that every Player/Missile had it's own memory pointer, it's own raster comparator, it's own height counter and everything made available through registers. Quite a lot of extra transistors without the space for them.

 

Color registers are OK since you get the ORed colors, but each missile having its own color register would have made it better.

Again quite a lot of extra transistors.

 

Since when did having an initial OFFSET to address DMA data require raster comparators, height counters, memory pointers, etc. There's already one offset in the ANTIC but can't be initialized to a user defined value. Currently, player 0 DMA data is at PMBase+1024+(8-bit counter). The 8-bit counter begins at 0 but if user had access to it, it could initialize it to an offset which would cause sprite to move vertically. And there is space within the 54272...54287 register space as the following registers are unused: 54278, 54280. And following three registers are read-only so we can use them for writing offsets: 54283 (VCount), 54284 (LPenh), 54285 (Lpenv). There, 5 registers for 5 player offsets. So not "quite a lot" of transistors.

 

In fact, using modulo arithmetic of the 8-bit data, they only had to have a STROBE to reset the offset like register 54282 (WSYNC). Then one could just reset the counter at a particular scanline depending on the Y-position.

 

As far as GTIA missile color registers, again you are being vague by stating "quite a lot of extra transistors". There's already a way to set the color of the missiles to either 53266..54269 or 53273 for each missile. So instead of that choice you would have a choice of 53266..53269 or 53256..53259. Currently only 3 values of 53256..53259 are used so they can be re-used to avoid extending the address space.

[atariksi]

Motorola did have a bigger market share so it would be hard at that time to decide to put Amiga chipset in a PC. Commodore had PCs that they sold w/o Amiga chipset. Adlib boards had no DACs (for sampled audio) nor were they a standard part of a PC package-- just good for music; Amiga/ST had the edge in audio even with the Adlib boards in PCs. Sound Blaster 8-bit which came later on was more competitive to the Amiga/ST audio although that also lacked multiple channels. I remember the Thinkpad 701C w/butterfly keyboard that had the Sound Blaster integrated in its motherboard.

 

The point that I was making was that PCs were starting to catch up at that point. They hadn't caught all the way up in '88 but they pretty much had by '92.

 

They hadn't caught *ALL* the way up with the latest 2.8Ghz Dell system I bought a couple of years ago.

[frogstar_robot]

They hadn't caught *ALL* the way up with the latest 2.8Ghz Dell system I bought a couple of years ago.

 

In terms of graphic and sound horsepower they pretty much had. No doubt you can find little corner casey things that Amigas do better than today's PCs and if you want to argue that Amiga is more elegant from an engineering POV then no argument from me. But the fact remains that today's PCs can accomplish anything a classic Amiga* can even if it has to use sheer brute force to emulate the Amiga itself.

 

*The funny little closed PCs that may more may not come out sometime "soon" are modern machines that emulate the old stuff if they bothers to run it at all. Braaaaaaaaaiiinnnns!

[Fröhn]

Since when did having an initial OFFSET to address DMA data require raster comparators, height counters, memory pointers, etc. There's already one offset in the ANTIC but can't be initialized to a user defined value.

There is no offset. They just used the rasterline counter ($D40B) as lowmost bits for PM data fetching.

 

Also your offset version doesn't save much compared to the comparator version: You have to introduce offset registers for every player + adding logic + register logic.

 

In fact, using modulo arithmetic of the 8-bit data, they only had to have a STROBE to reset the offset like register 54282 (WSYNC). Then one could just reset the counter at a particular scanline depending on the Y-position.

That would reset the entire rasterline counter and mess up the display timing. Also, it would reset the Y-position of all PMs and not single PMs.

 

As far as GTIA missile color registers, again you are being vague by stating "quite a lot of extra transistors". There's already a way to set the color of the missiles to either 53266..54269 or 53273 for each missile. So instead of that choice you would have a choice of 53266..53269 or 53256..53259. Currently only 3 values of 53256..53259 are used so they can be re-used to avoid extending the address space.

GTIA is another matter and has nothing to do with PM data fetching (or Y-positions for that matter) of the ANTIC. Also that choice of color registers doesn't eat much space. The color registers exist anyway, you just select the source of the color bits.

[atariksi]

They hadn't caught *ALL* the way up with the latest 2.8Ghz Dell system I bought a couple of years ago.

 

In terms of graphic and sound horsepower they pretty much had. No doubt you can find little corner casey things that Amigas do better than today's PCs and if you want to argue that Amiga is more elegant from an engineering POV then no argument from me. But the fact remains that today's PCs can accomplish anything a classic Amiga* can even if it has to use sheer brute force to emulate the Amiga itself.

 

*The funny little closed PCs that may more may not come out sometime "soon" are modern machines that emulate the old stuff if they bothers to run it at all. Braaaaaaaaaiiinnnns!

 

It's hard to say exactly what is "today's PCs" since hardware varies with various systems. If you go by standard hardware (present in >95% of PCs), they can't fully emulate the Amiga nor Atari nor some other machines brute force or otherwise. And if you are going to target some specialized hardware on a PCs then you mine as well add some specialized hardware to the Amiga and Atari as well.

 

As far as emulating older PCs on newer PCs, they don't really run that well either unless you have a PC that maintains the legacy I/O ports and an OS that lets you access the I/O ports.

[atariksi]

Since when did having an initial OFFSET to address DMA data require raster comparators, height counters, memory pointers, etc. There's already one offset in the ANTIC but can't be initialized to a user defined value.

There is no offset. They just used the rasterline counter ($D40B) as lowmost bits for PM data fetching.

 

Also your offset version doesn't save much compared to the comparator version: You have to introduce offset registers for every player + adding logic + register logic.

 

...

 

You do need 5 8-bit counters that count just like the VCount, POTs, VPenV, ... but they can be reset like POTs to start recounting. Not as complex as you make it out to be with memory pointers, raster comparators, etc. Yeah, you do need some logic to select the LSB 8-bits of the PMBase pointer.

 

>>In fact, using modulo arithmetic of the 8-bit data, they only had to have a STROBE to reset the offset like register 54282 (WSYNC). Then one could just reset the counter at a particular scanline depending on the Y-position.

>That would reset the entire rasterline counter and mess up the display timing. Also, it would reset the Y-position of all PMs and not single PMs.

 

No, the LSB 8-bits of PMBase pointer would get the correct counter. If you used only one counter, then all sprites would move up vertically together. There's no display timing to mess up.

 

>>As far as GTIA missile color registers, again you are being vague by stating "quite a lot of extra transistors". There's already a way to set the color of the missiles to either 53266..54269 or 53273 for each missile. So instead of that choice you would have a choice of 53266..53269 or 53256..53259. Currently only 3 values of 53256..53259 are used so they can be re-used to avoid extending the address space.

>GTIA is another matter and has nothing to do with PM data fetching (or Y-positions for that matter) of the ANTIC. Also that choice of color registers doesn't eat much space. The color registers exist anyway, you just select the source of the color bits.

 

And we weren't talking about Y position with GTIA-- only colors for missiles. You mixing up the two. There's a VDELAY which affects Y-positioning for double line resolution but that's not the point here.

[JamesD]

I wonder what speed was possible for the 6502 in 1982. The BBC ran memory at 4MHz and the 6502 at 2MHz ( with no wait states ) - An 800XL with 3.84MHz memory and a 3.84MHz 6502 would have ruled ( especially if it have additional graphics modes )

I seem to remember an option of the OSI computers for an "ION Implanted 6502" or something like that. I *think* the option was up to 4MHz on some of their machines but it was pretty pricey. The OSI Challengers were out in 1978. It's been a long time so I could be wrong.

 

A 2MHz C64 would have ruled... but then the 1MHz did anyway. At least sales wise.