Difference between Amiga and Atari 520?

[Speelgoedmannetje]

shoggoth wrote:

The copper co-processor makes it possible to use multithreading natively. No extra software needed. No cpu cycles spilled waiting. [/quote]

You'll have to explain that further. I'm fairly familiar with the Copper, but I can't see how this relates to multitasking. It's a coprocessor. It's not like it makes context-switching in the CPU any faster.

-- Peter[/quote]It has the 'wait' instruction.

[shoggoth]

Speelgoedmannetje wrote:

shoggoth wrote:

The copper co-processor makes it possible to use multithreading natively. No extra software needed. No cpu cycles spilled waiting.

You'll have to explain that further. I'm fairly familiar with the Copper, but I can't see how this relates to multitasking. It's a coprocessor. It's not like it makes context-switching in the CPU any faster.

[/quote]It has the 'wait' instruction.[/quote]

And exactly how does that help when switching contexts in a multitasking OS?

-- Peter

[Speelgoedmannetje]

One can program a very quick scheduler when one can use already a wait commando (without making the cpu wait).

[stopthegop]

Windows users think that just having two programs loaded in memory constitutes multitasking.  I've always said that unless you've used an Amiga, you've probably never seen, much less experienced, true pre-emtive multitasking.  That still holds true today.  Just recently I formatted 10 SFS disk partitions on three separate scsi drives on my A4000...at the same time.  Try that on a PC!  

[itix]

One can program a very quick scheduler when one can use already a wait commando (without making the cpu wait).

You can't because copper cannot control CPU.

[Speelgoedmannetje]

itix wrote:

One can program a very quick scheduler when one can use already a wait commando (without making the cpu wait).

You can't because copper cannot control CPU.

Since when?
It can generate interrupts.

[shoggoth]

You can't because copper cannot control CPU.[/quote]Since when?
It can generate interrupts.[/quote]

Dude, lots of stuff can generate interrupts. Still has nothing to do with multitasking. You're confusing apples with oranges.

-- Peter

[Speelgoedmannetje]

"Dude" I know. But it's one way to control the CPU.
Well then, do your own research about this subject then. I got an old Dutch manual wich states that the copper can be used for multitasking.

[shoggoth]

Speelgoedmannetje wrote:
"Dude" I know.
Well then, do your own research about this subject then. I got an old Dutch manual wich states that the copper can be used for multitasking.

The Copper is a nice gadget, but either you've got it wrong or we're discussing two completely different topics. According to that philosophy, triggering a few routs based on external timer interrupts would also be multitasking, which it certantly is not.

Sorry for going off topic, btw.

-- Peter

[Speelgoedmannetje]

Interrupt handling on itself indeed isn't multitasking.
But 68k on itself isn't multitasking. Yet the Amiga can do multitasking. On a low level. At 7 mHz. Go figure.

[spirantho]

True, but there are other machines that'd be even better, hardware-wise. From what I recall the WAIT instruction waits for a vertical position of the electron beam, which is obviously nowhere near enough resolution to help with kernel level multitasking.

For instance, the scan rate of a PAL signal is 50Hz, yes? This means that given a vertical resolution of about 1,000 rows (including the blanking signal), that's an ability to raise an interrupt 50,000 times per second, i.e. 50KHz. Now we're using CPUs with a resolution of 50MHz, not KHz. If the multitasking resolution was limited to a signal from the copper, the CPU would be sitting around waiting most of the time we did a context switch.

As I see it, the copper has nothing to do with multi-tasking in the OS. It IS very useful in generating display lists because it takes the load off the CPU, but that's because it's a processor itself.

And if you want proof that AmigaOS doesn't use the copper, ask any Draco  owner if his machine a) has a copper and b) runs AmigaOS.

[whabang]

InTheSand wrote:
And don't forget it took Microsoft until Windows XP to have a consumer OS that finally dumped the underlying DOS-based stuff!!

I hate to be a knitpick, but that's just not true! The first true 32-bit version of Windows was Windows NT 3.1. IIRC, it was released in 1993.

[TjLaZer]

I would say the biggest difference is the OS.  AmigaDOS 3.x is like eons ahead of TOS 4.x, even with OS enhancements like MagiC and MiNT, etc.  But never the less, the Atari computers are fun to play around with.  I remember the great Atari vs Commodore wars, ah the memories!!!

[shoggoth]

Speelgoedmannetje wrote:
Interrupt handling on itself indeed isn't multitasking.
But 68k on itself isn't multitasking. Yet the Amiga can do multitasking.

No, the 68k on itself isn't multitasking. It's a CPU, for god sake.

So you mean multitasking on 7Mhz is some kind of "proof" for your so called "hardware multitasking" theory? There is no such thing. There are hardware solutions which enables you to do several things simultaneously in a smooth manner. That's not necessarily the same as executing multiple tasks concurrently. It's unrelated.

On a low level. At 7 mHz. Go figure.

You're mixing apples and oranges. Are you familiar with low level 68k programming, btw?

-- Peter

[shoggoth]

TjLaZer wrote:
I would say the biggest difference is the OS.  AmigaDOS 3.x is like eons ahead of TOS 4.x, even with OS enhancements like MagiC and MiNT, etc.  But never the less, the Atari computers are fun to play around with.

TOS4.x was basically just a more colourful 030-compatible version of TOS2.x. It really didn't add anything at all, except some new calls to handle new hardware. The MultiTOS that was shipped with TOS4 machines sucked. It was (sort of) Posix compliant, though.

Later incarnations of MiNT (known as FreeMiNT) has come a long way, but that's some 15 years too late

Atari never saw the point of having good OS functionality, not even when they released the Falcon. At the time of the release of the ST, I can understand this. The Amiga wasn't known for it's pioneering OS, but rather how many colours that could be displayed on screen. This changed during the years, and the Amiga was ready for it, the ST range wasn't.

I remember the great Atari vs Commodore wars, ah the memories!!!

Yeah. Some psycho professor somewhere must have written a book about it. There must be some medical explanation for this

-- Peter

[Giana]

Two fantastic machines !!!
I still have both.

Atari is more simply made, poor architecture, no custom chips so overall graphics and sounds capabilities are not as good as the Amiga(s). Simpler to code too (on first approach ...). And true first demos, first stunning gfx effects were coded on Atari first coz the challlenge was big : how to push the limit of this little machine ...

On the other hand, during the early good old days, games development on Amiga were often badly done : simply a kind of cross compiling with no true use of the custom chips, so games were slightly identical. But this didn't last too long, and "specific" games appeared : I remember a great change from the first release of Shadow of the Beast. I was 13 and and it was like a nuke fell down on our heads.

Well, this is my point of view ...
Giana