@fishy_fis

Quote:

I wonder where mankind would be today if religion hadn't held back science for hundreds of years.

Cam you give an example?

Quote:

I sometimes really hope there is a God though, or other deities, just so some people can get the eternal slap they so richly deserve

The problem with that I think is that we'd all end up in the same line and being judged ourselves.

Anyway you said something about organised religion. I actually prefer that one. What I hate is disorganised religion making a mess of the place!

@cdimauro

Quote:

You can take a look at the Bible, for example, from the very beginning, carefully reading the stories, and trying to imagine what they describe. I warmly recommend it to quickly become an atheist.

A few years back I read Genesis on holiday. It wasn't read quickly nor did I become an atheist after it. I don't think I was swayed either way from when I started it. Perhaps I read the wrong version.

I'd quit my job and finally have more time to mess with my miggies

_________________
____________________________
c64-2sids, A1000, A1200T-060@50(finally working!),A4000-CSMKIII
! My Master Miggies- Amiga 1000 & AmigaOne X1000 !
mancave-ramblings
X1000 I BELIEVE

@mcbone

Forget the money; it would be far more beneficial to go back with the knowledge we have today. For example, if Jay Miner had known in the early eighties that the Motorola 68000 does not scale, would he still have selected it?

Here's another one: if the Amiga design team had known that id Software would have Wolfenstein 3D running on a 286 with VGA in 1992, would they have changed the design of the original Amiga graphics architecture?

Now for some real heresy: what if the Amiga had used a 286 and had hardware memory protection from the start?

Last edited by bison on 26-Aug-2015 at 06:25 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

@mcbone

I think with 8 billion pounds and compound interest you could probably own the world by now.

_________________
X5000 / microA1(OS4.1 FE U2) / CodeBench / Imagine / Blender
Lightwave 2019 / Microsoft Visual C++

@bison

Quote:

For example, if Jay Miner had known in the early eighties that the Motorola 68000 does not scale, would he still have selected it?

68000 was sensible choice. Even knowing what we know today, I would choose it - transition to another architecture could follow later (1994/1995).

Quote:

Here's another one: if the Amiga design team had known that id Software would have Wolfenstein 3D running on a 286 with VGA in 1992, would they have changed the design of the original Amiga graphics architecture?

Original Amiga architecture was great for 1985/1987, but Commodore didn´t make substantial improvements until its own demise. AGA was introduced 7 years after Amiga release and I hardly can call it revolutionary design like OCS was in its time.

There are 5 things I would do with enough money in the right time (for 1990 A3000 release):
1. Improvements on the OS side - OS2.0 with basic RTG, that would allow greater choice of GFX hardware (not tied to classic Amiga chipset).
2. New GFX chipset without burden of past.
3. Cheaper OCS compatible reimplementation - all Amiga chips in one enabling full compatibility with older generation in next generation machines and reducing price of budget models. 256 colour mode (from 4096) would be nice, if possible.
4. "A1000+" with above mentioned architecture (and 68020 or 68EC020) for 1500 USD - A3000 was too isolated in its price segment, A1000+ could be used as reference hardware for games developers.
5. "A500+" with OCS reimplementation and HDD controller.

With these steps, Commodore would be in much better position to continue as relevant power in its target markets (video applications and games).

@bison

At the time the '286 was current, my company was building 68020 based systems.
Given a choice between those two processors, I'd still choose the 68020.
Intel architecture did not really compete until the '386 was introduced.
And the 68040 and 68060 kept Motorola competitive until the Pentium was introduced.

@iggy

Quote:

And the 68040 and 68060 kept Motorola competitive until the Pentium was introduced.

68040 didn´t clock above 40 MHz and 68060 was introduced year after Pentium.

@Hypex

Its 7.15 a.m, so my brain isn't awake yet, but some simple examples are Copernicus, Galelaio, Cecco d'ascolli, Marco Antonio de Dminos, burning of the library of Alexander and Roger Bacon.
Heck, there was even a succession of popes from 12th (if memory serves correctly) century onwards that banned the study of anything that could go against what the church taught.
People were imprisoned, tortured and murdered for studying and teaching for hundreds of years just for thinking freely.

If you want more specific or articulated responses you'll have to wait for coffee to be involved :)

@bison
Quote:

... if Jay Miner had known in the early eighties that the Motorola 68000 does not scale, would he still have selected it?

I'm not disagreeing, I'm just trying to understand.

What makes the 68000 not a scaleable architecture compared to (say) the 8086 or the 286? Or in other words, could we have gone from the 68000 to a hypothetical 68999 with multiple cores in a CPU, vector units, etc, whilst still keeping the 68000 ISA, much like the modern x86 PC can still run DOS? Ignoring, for the moment, 64bit etc. Or is there something fundamentally non scaleable in the ISA itself?

FWIW I think Jay and co made the right decision. What's interesting is that the MIPS proc was released in 1985, though presumably it was crazily expensive.

_________________
Blog

8 billion would not be enough. What killed Commodore '94 was not lack of ideas, it was not lack of a market and it wasnt a lack of good engineers. It was poor management. I dont care if you have 60 billion, without good management, management that doesnt understand its industry and without UNIFYING management all you are doing is wasting your time. Thats why Sculley failed at Apple. You had a guy that was in over his head and who understood the soda industry but not the computer industry. Steve Jobs kinda failed at NeXT because what he didnt take into account was that the enterprise and education markets are not like the consumer markets. Enterprise and Education dont upgrade systems once a year, they go 5 or 6 years without upgrading and longer with servers, the one thing Steve had on his side was that even though he wasnt a programmer, he wasnt an engineer but he understood the concepts. You could have a conversation with him and he knew what you were talking about. From my understanding, and it may be heresay, but Sculley and Gould knew nothing about the industry and knew nothing of operating a company in the industry. The computer industry is a very unique industry and if you dont understand it, if you dont understand at least the basics of your product you will fail. With 8 billion euro's or dollars or whatever and you retain the same piss poor management and incompetent management my wager would not be if you were going to fail but when you were going to fail.

_________________
http://www.pc-opensystems.com

@Aslak3

The 68000 was chosen for no reason other than being a good balance of power and price at the time.
A year or 2 earlier or later and something else would have been used.
It could have been pushed further, but PowerPC, being a joint venture, was theoretically better financial prospect. The joint venture side more or less fell through though much to Motorola's chagrin, but they were pretty much stuck with it by this point due to the amount invested.
Bit of a shame really as there was no real technical advantage to the architecture.
Over time this changed, but only because it was further developed whereas 680x0 wasn't, apart from veering of into Coldfire, but development there wasn't centred around enhancing the CPU.

@fishy_fis

Quote:

The other side of your post is somewhat confused though (not me, your thought process).

I am trying to say two things without going on and on about it, as I'm know to do.

1. There is not a single Amiga experience since the demise of Commodore that is worth preserving. Or that is not worth trading for any other experience, including complete platform death, i.e. It's a lot easier to get through the grieving process when there isn't a ghost lingering around and haunting you.

2. 'Tis better to have loved and lost than never to have loved at all - Alfred Lord Tennyson
Not always. And it depends on how many years one has to go on without it since first having it.
Some would say it's a matter of perspective, glass half full and all. That might work if one is secluded in the wilderness somewhere, eking out an existence amidst a log cabin setting. But when you are reminded every day of the thing you have lost and know you will never have it again, how does that glass look now?

Last edited by agami on 27-Aug-2015 at 06:33 AM.
Last edited by agami on 27-Aug-2015 at 01:51 AM.

_________________
All the way, with 68k

@Aslak3

Quote:

What makes the 68000 not a scaleable architecture compared to (say) the 8086 or the 286?

I don't really know, but there are probably people on this forum who do. It could have been some kind of technological barrier (complex instruction set), or it might have been economic -- Motorola may have decided at some point that it wasn't worth sinking more money into.

The ironic thing is that the 286, which Bill Gates referred to as a "brain dead" processor, had an upgrade path to the 386, which was really quite good (it's little-endian, which sucks, but other than that...). And the 286 wasn't as bad as Bill said it was -- it had trouble with MS-DOS, specifically that issue of switching from real mode to protected mode, but it ran Xenix fine, as far as I know.

I can't help thinking that if Amiga had had hardware memory protection and chunky mode graphics from the start, it might have made it further along the road.

_________________
"Unix is supposed to fix that." -- Jay Miner

@Hypex

Quote:

Hypex wrote: @cdimauro Quote: You can take a look at the Bible, for example, from the very beginning, carefully reading the stories, and trying to imagine what they describe. I warmly recommend it to quickly become an atheist. A few years back I read Genesis on holiday. It wasn't read quickly nor did I become an atheist after it. I don't think I was swayed either way from when I started it. Perhaps I read the wrong version.

I quote myself:

"Maybe it's better to stop replying, or create a new OT thread and move there, if someone is interested on continuing the discussion."

@iggy

Quote:

iggy wrote: @bison At the time the '286 was current, my company was building 68020 based systems. Given a choice between those two processors, I'd still choose the 68020. Intel architecture did not really compete until the '386 was introduced. And the 68040 and 68060 kept Motorola competitive until the Pentium was introduced.

The 80286 had very good performance too, and it was very competitive from this point of view, even against the 68000.

@Aslak3

Quote:

Aslak3 wrote: @bison Quote: ... if Jay Miner had known in the early eighties that the Motorola 68000 does not scale, would he still have selected it? I'm not disagreeing, I'm just trying to understand. What makes the 68000 not a scaleable architecture compared to (say) the 8086 or the 286? Or in other words, could we have gone from the 68000 to a hypothetical 68999 with multiple cores in a CPU, vector units, etc, whilst still keeping the 68000 ISA, much like the modern x86 PC can still run DOS? Ignoring, for the moment, 64bit etc. Or is there something fundamentally non scaleable in the ISA itself?

The only problem with the 68000 was due to the variable-length instruction set, and some exceptions on the opcode table. However x86 had similar problems, and with their prefixes it's even worse.

But 68020 increased a lot the complexity of decoding, especially due to the double memory indirect modes, which also extended the maximum instruction length to 22 bytes (x86 has always a 15 bytes limit, which is much more "friendly" to handle).

Executing instructions with the double indirect modes presents its problems, too.

Anyway, I have nothing to measure the scalability of each processor family, but looking at the 68060 and compared to the Pentium I feel that Motorola had its hard time to manage such complex ISA.
Quote:

FWIW I think Jay and co made the right decision. What's interesting is that the MIPS proc was released in 1985, though presumably it was crazily expensive.

Absolutely.

@fishy_fis

Quote:

fishy_fis wrote: @Aslak3 The 68000 was chosen for no reason other than being a good balance of power and price at the time. A year or 2 earlier or later and something else would have been used. It could have been pushed further, but PowerPC, being a joint venture, was theoretically better financial prospect. The joint venture side more or less fell through though much to Motorola's chagrin, but they were pretty much stuck with it by this point due to the amount invested. Bit of a shame really as there was no real technical advantage to the architecture.

It was easier and cheaper, at the time, to get better performance for a RISC processor, compared to a CISC (especially complex ones like a 68020+ or an 80386+).

However CISCs quickly filled the gap in a few years, and at the end of '90s they reached similar performances.

Due to that, some "noble" RISC processors died.

@bison

Quote:

bison wrote: @Aslak3 Quote: What makes the 68000 not a scaleable architecture compared to (say) the 8086 or the 286? The ironic thing is that the 286, which Bill Gates referred to as a "brain dead" processor, had an upgrade path to the 386, which was really quite good (it's little-endian, which sucks, but other than that...).

The 80286 was one of the first processors to pass some military security specs.
Quote:

And the 286 wasn't as bad as Bill said it was -- it had trouble with MS-DOS, specifically that issue of switching from real mode to protected mode, but it ran Xenix fine, as far as I know.

The 286 had problems switching from protected to real mode. However there was a LOADALL undocumented instruction which allowed to load all registers, so it was possible to switch back to real mode using it.
Quote:

I can't help thinking that if Amiga had had hardware memory protection and chunky mode graphics from the start, it might have made it further along the road.

Chunky/packed pixels mode would have been added after the first chipset, but definitely NOT after so many years, and certainly not the with the badly AGA patch.

What also crippled the future of the platform was the bad design of the o.s.. I can understand that a 68000 has no memory protection at all, but an o.s. with a good would have enabled it transparently when running on an capable processor...

So, both hardware and software had their big issues, unfortunately.

@bison

Quote:

For example, if Jay Miner had known in the early eighties that the Motorola 68000 does not scale, would he still have selected it?

I think so. Big endian is friendly to computer and human alike. The 68K was easy to program to assembler once you learn it. And the main reasons are that the address space space is linear, you could access any location easily. Large register count, 8 address and 8 data. And let's not forget that externally the 68K was 16-bit like the average 80's CPU but internally it was 32-bit. It was 32-bit design from the ground up. It was simply superior. Perhaps the best CPU on the market. But the Amiga got the old 68000 model when the 68020 was available.

Compared to the next big thing, x86 or 80x86 at the time, was terrible. Now if we compare like for like we stick to CPUs below the 80386. It was based on the 16-bit 8086. It had segmented address space, so you had to fiddle with 16-bit offsets and shifting them left to break the 16-bit limit. Less register count meaning you needed to stack a lot, good for your C compiler. The Amiga came out at a time when you learned about the memory map, CPU and programming it directly. 80x86 wouldn't have been as friendly. Even if it would have continued the "LO BYTE, HI BYTE" legacy of the 6502.

Quote:

Wolfenstein 3D running on a 286 with VGA in 1992, would they have changed the design of the original Amiga graphics architecture?

I don't think so. There was already talk I read of Jay Miner discussing chunky but planar was better for memory conservation. And Commodore like to cut costs. And besides, I see no reason why the pixel width (set at 1) couldn't have been increased to 2,4, and 8 bits for one plane to do chunky later. So increased pixel width as opposed to depth. Same colours, same space required, different packing model. But, looking back to the C16/+4, in multicolour mode they were chunky.

Another thing is that these games didn't use features like dual playfield or blittering. So unsuitable that way. But yes I think a chunky mode would have helped. And it would have been superior with a copper!

Quote:

Now for some real heresy: what if the Amiga had used a 286 and had hardware memory protection from the start?

No point. It was still 16-bit. You'd need the 80386 to get a 32-bit register length. And even if was cheap, it was still x86! IMHO, the best choice was made.