Things the Amiga didn't get right from Day 1

[Gorf]

Originally Posted by Bruce Abbott

You're dreaming.

No, this definitely was the case from 1988-1992

Where's SUN now?

There: oracle.com
It survived until 2010 on ist own .. 16 years longer than CBM
And it did not go bankrupt but was bought for US$ 7.4 billion.

Where's Atari now?

I mentioned the Atari TT because it proves, that even with limited resources a company could build a affordable workstation with 64bit gfx-bus at this point in time.
The A3000 was fine, but to use it as a workstation one needed a additional expensive gfx-card, since even the included Amber was not enough for what people expected in 1990 from a workstation.

If you want to sell a workstation you better make sure it does what a workstation is expected to do..

Which would soon not be enough.

The A3000 lasted a scant 2 years before being dumped at fire-sale prices.

Sure:
as every other model after the A1000 the A3000 came one year to late ...
and as for every computer from any company back then:
after two years it was outdated.

CBM should have brought the A3000 with higher resolution in 89
the A3000+ with DSP in 90
the AA3000+ in 91
....

Amiga fans couldn't afford it,

you and me did ...

while professional users didn't want it.

I mentioned the reasons for that.

But that didn't stop them continuing the 'A3000 approach'. The A3000+ was going to have even more expensive hardware in it, but they couldn't get it to work. How much money would it take to go far enough?

They spent almost nothing on the A3000 in terms of development - what was it? Two or three people working on it?

They wasted a lot more money on the CDTV at the very same time.
Or on the A600 a little bit later (that turned out to be even more expensive to produce at the A500 instead of cheaper, while the margins where shrinking and shrinking for this kind of computers...)
And of course they wasted money on the PC department, which had much more engineers than the Amiga department, while they where just doing what every cheap clone-manufacturer did.

So how much more for the A3000 project?
I would say they should have at least tripled the effort.

[Bruce Abbott]

Originally Posted by Gorf

They spent almost nothing on the A3000 in therms of development - what was it? Two or three people working on it?

Engineers salaries are not the only R&D cost.

They wasted a lot more money on the CDTV at the very same time.

So you say. But the CDTV was supposed to be a 'high-end' consumer product, just like the A3000 was a 'high-end' computer. Same attitude, same result.

Or on the A600 a little bit later (that turned out to be even more expensive to produce at the A500 instead of cheaper, while the margins where shrinking and shrinking for this kind of computers...)

I keep hearing this story, but it's misleading. The A600 was originally designed to be nothing more than a 'cut-down' A500. If that was more expensive then you would have a point. But they sensibly decided to make it more than that. There were several reasons for the A600's manufacturing cost being higher than expected, but it was still good value for money.

The R&D effort put into the A600 wasn't wasted, it helped with the design of the A1200. The problem was that people were expecting something like the A1200 from the beginning, so when the A600 arrived instead a lot of them were disappointed. Nevertheless it sold well and they are now highly sought after.

And of course they wasted money on the PC department, which had much more engineers than the Amiga department, while they where just doing what every cheap clone-manufacturer did.

Evidence that it was 'wasted'?

So how much more for the A3000 project?
I would say they should have at least tripled the effort.

I say they should have tripled the effort on the A1200 instead. 'high-end' Amigas never sold well, but an accelerated AGA (or better) 'low-end' machine replacing the A500 in 1990 (when the average 'low-end' PC was a crappy 386-SX) would have been a killer.

[Bruce Abbott]

Originally Posted by Gorf

you and me did ...

A few of us were rich enough to buy 'high-end' Amigas. Sadly that didn't last long for me. I lost my high-paying job in NZ Telecoms when the government sold it to a private company. Since then I have not been able to justify spending over $7000 on a computer system. I did waste thousands adding an 060 and RTG to the A3000 though, eventually selling the lot for a miserable $1000 (would get a lot more for it now as a retro computer, but I needed the money and Amigas were worthless in 2001).

In 1991 the A3000 cost the same as a 'high-end' PC, which is how I justified the outlay. But had Commodore spent even more R&D on it the price would have been even higher. I couldn't afford a workstation, and neither did I want or need one. Commodore had no experience in that market and weren't suited for it. Their only hope was to OEM a 'cheaper' machine to a company that did, but where's the money in that? If only they had concentrated on the market they knew...

[pandy71]

Originally Posted by Gorf

Again someone claiming "instead" ... why?
These things are not mutually exclusive.

Well, only HAM could be beneficial from YUV, not EHB - and things at some point are mutually exclusive - as pointed before you have 12 bit - you can use them to produce 4096 colors in RGB and those coloras are visible on your display or you can use 12 bit to produce for example 2000 visible colors, additionally this will cost you more as you need video OPAMP's capable to perform subtraction...

Buy LMH1251 and do own Vidiot so you can verify your idea today...
Or buy some YPbPr to HDMI converter - 20$ and you will know how YUV will work... (personally i would use Blue line as Green so 5 bit HAM mode can be used)

[Gorf]

Originally Posted by Bruce Abbott

Evidence that it was 'wasted'?

The sales of Commodore PCs were close to zero in the US, the last country with any significant sales was Germany and even there these were dwindling so fast that CBM stopped all PC-compatible business in 1993 to finally concentrate on the Amiga.
(of course again way too late)
Despite having more engineers working in the PC department the whole range of PCs was hopelessly outdated ...

So yes: it was obviously totally wasted.

I say they should have tripled the effort on the A1200 instead. 'high-end' Amigas never sold well, but an accelerated AGA (or better) 'low-end' machine replacing the A500 in 1990 (when the average 'low-end' PC was a crappy 386-SX) would have been a killer.

Wedge design was futile at this point. A mid/low-cost A1000 NG would have been the right answer.

But I suggest we leave it there, since our argument just runs in circles:

Your point is CBM should have concentrated more on the low cost sector and my point is, that is what they did and what was their demise ...



Lets agree to disagree

[Gorf]

Originally Posted by Bruce Abbott

So you say. But the CDTV was supposed to be a 'high-end' consumer product, just like the A3000 was a 'high-end' computer. Same attitude, same result.

And repeating the same mistake they made with the overpriced A2000:
just repacking the old technology.

Many users and computer-fans were initially amazed by the new CDTV. CD-ROM was hip of course and finally "multimedia" was recognized in the computer- and electronics-world...

But as soon as the first "tear down" articles appeared in magazines, everybody went from "whooo" to "meehhh":
the same (now 5 year old) chipset - even the cpu still at 7Mhz ... even reusing Kickstart 1.3

While CD-ROM was hip and not a bad move ... the CDTV was soon seen as just a A500 with a CD-drive.

You can try to sell an low-end product as "high-end" ... but usually this will fail and so did the CDTV.

If you want to sell "high-end" better make sure you can keep the promise.

[Gorf]

Originally Posted by pandy71

Well, only HAM could be beneficial from YUV, not EHB

no, also EHB would benefit, as I pointed out earlier, since it would give you different colours instead of just darker shades of the same colour.

- and things at some point are mutually exclusive - as pointed before you have 12 bit - you can use them to produce 4096 colors in RGB and those coloras are visible on your display or you can use 12 bit to produce for example 2000 visible colors, additionally this will cost you more as you need video OPAMP's capable to perform subtraction...

Well: again, the visibility depends entirely on the gamut you choose - you could fit all 4096 YUV values in a larger RGB "cube", if the gamut for your YUV space is much smaller than the gamut for your RGB space ...

In practice you probably would go for a compromise: you might lose some "extreme" colours, but most would translate just fine into RGB space.
Among them virtually all colours developers and users alike do actually choose in their palettes..

(as mentioned before, it's not like Amiga's RGB space does not waste a good part of its possible values as it is now:
It is a common complaint about EHB that the dark colours are only ever useful for shadows ...)

[pandy71]

Originally Posted by Gorf

no, also EHB would benefit, as I pointed out earlier, since it would give you different colours instead of just darker shades of the same colour.

You have single bit and this single bit could modify Y so you will get same effect as currently in RGB, for U and V this will not work so simple...

Originally Posted by Gorf

Well: again, the visibility depends entirely on the gamut you choose - you could fit all 4096 YUV values in a larger RGB "cube", if the gamut for your YUV space is much smaller than the gamut for your RGB space ...

In practice you probably would go for a compromise: you might loose some "extreme" colours, but most would translate just fine into RGB space.
Among them virtually all colours developers and users alike do actually choose in their palettes..

(as mentioned before, it's not like Amiga's RGB space does not waste a good part of its possible values as it is now:
It is a common complaint about EHB that the dark colours are only ever useful for shadows ...)

So you need to perform some limited YUV so it fit in RGB... perhaps it will work but not sure if this can be easily done...
Well EHB is other way to use 6-th bitplane - of course it would be better to have 64 color registers or even more so they can be switched with single Copper MOVE. You imagine even programmable 6-th bitplane i.e. some ALU in RGB signal path but still - half level is not so bad at all, with proper color allocation you may have almost all 64 colors unique.

[sandruzzo]

Even though we don't have YUV on hardware, we still have the copper that allow us to change palette every line. Do you remember how incredible was dynamic Hires

[Gorf]

Originally Posted by pandy71

You have single bit and this single bit could modify Y so you will get same effect as currently in RGB,

exactly

for U and V this will not work so simple...

It will give you a different colour.

Therefore I suggested to split the EHB mode into two modes:
one just affecting Y and leaving U and V unaltered (darker shades)
one just affecting U and V and leafing Y unaltered (different colours)

So you need to perform some limited YUV so it fit in RGB... perhaps it will work but not sure if this can be easily done...
Well EHB is other way to use 6-th bitplane - of course it would be better to have 64 color registers or even more so they can be switched with single Copper MOVE.

sure this would have been nice ... but here I fully accept the argument of restrictions of silicon space.
As far as I understand the colour registers take quite some space.

[Bruce Abbott]

Originally Posted by Gorf

The sales of Commodore PCs were close to zero in the US, the last country with any significant sales was Germany and even there these were dwindling so fast that CBM stopped all PC-compatible business in 1993

This is deceptive. The entire Commodore PC range was designed and manufactured by Commodore Büromaschinen GmbH in Germany (though we had many of them down here in New Zealand too).

Sales figures from those times are hard to find, but here are some quotes:-

TechMonitor 09 Dec 1987

COMMODORE “HAS TOP BUSINESS PERSONAL, HOME COMPUTER IN WEST GERMANY”

More figures, this time from the West German Chip personal computer magazine, which suggests that the Commodore PC-10 MS-DOS micro is the top-selling business personal computer in the West German market – and the number two spot also goes to Commodore with the PC20. The Apple Macintosh comes third, but Commodore pops up again in fourth place with the PC40 AT-alike.

New York Times: January 30, 1991

Commodore Net Soars on Europe Sales

Commodore International Ltd. today reported more than a threefold gain in earnings for the quarter ended Dec. 31, primarily because of strong European sales of personal computers...

Commodore said its growth in the quarter was dominated by its Amiga personal computers, whose sales have surged on the strength of their powerful graphics and video capabilities. Sales of the company's I.B.M.-compatible personal computers also grew...

Irving Gould, Commodore's chairman and chief executive, said in a statement. "We are pleased with the strong sales performance of our European operations, which accounted for 85 percent of total sales for the quarter. Several European countries experienced sales growth of over 50 percent."

Some things you might not know about these German machines. They had:-

- 'autoconfig' to prevent I/O conflicts, many years before 'Plug&Pray' became a standard for PCs.

- on-board graphics, parallel and serial ports etc. when most clones needed several ISA cards to do the job.

- a 9 pin bus mouse port that was compatible with the Amiga, when most clones were using crappy serial port mice.

They were not the latest technology, but solidly built and reliable - which is what most businesses wanted.

Originally Posted by Gorf

to finally concentrate on the Amiga.
(of course again way too late)
Despite having more engineers working in the PC department the whole range of PCs was hopelessly outdated ...

Outdated perhaps, but not hopelesly. I was selling the later model 'slimline' Commodore 386SX-16 alongside the A1200, and the PC was a much easier sell. Wish I had kept one of them for posterity because the motherboard had familiar names on it. That particular model is now very rare.

[Bruce Abbott]

Originally Posted by Gorf

And repeating the same mistake they made with the overpriced A2000:
just repacking the old technology.

Many users and computer-fans were initially amazed by the new CDTV. CD-ROM was hip of course and finally "multimedia" was recognized in the computer- and electronics-world...

The CDTV was designed to be a consumer product like a VCR etc., and was not initially sold in computer stores. It was 'high-end' for the market it was designed for.

But as soon as the first "tear down" articles appeared in magazines, everybody went from "whooo" to "meehhh":
the same (now 5 year old) chipset - even the cpu still at 7Mhz ... even reusing Kickstart 1.3

Eh? Surely everybody knew what was in it? Well I did of course, because I was a registered CDTV developer. Having an 'old-technology' Amiga in it was a good thing. It meant we could get right to work on producing CDTV titles before even before getting a unit to test them on. Familiar hardware and OS, low cost development platform (any Amiga) etc. This was rather important when it cost NZ$1000 to have a 'gold disc' master made.

While CD-ROM was hip and not a bad move ... the CDTV was soon seen as just a A500 with a CD-drive.

Yes, CDROM multmedia was seen as 'hip' - by technologists. But that didn't translate into sales because consumers didn't really want it. Unfortunately I too was sucked in by the hype and lost some money on that venture (was still worth it for the experience though).

You can try to sell an low-end product as "high-end" ... but usually this will fail and so did the CDTV.

CDTV failed because there was no market for the product. Philips CDi had much higher specs ('on paper' at least) and a similar price - and also bombed, for the same reason.

If you want to sell "high-end" better make sure you can keep the promise.

And now we know why Commodore didn't want to sell the CDTV to Amiga fans... never satisfied - always complaining about their machine not being powerful enough, or cheap enough, or not having some pet feature.

I bet you thought the CDTV should have had all the guts of an A3000 in it, right? Plus YUV of course. And would have sold for 8 times the price. Just imagine how that would have gone down in the consumer audio/video market!

[pandy71]

Originally Posted by Gorf

exactly

It will give you a different colour.

Therefore I suggested to split the EHB mode into two modes:
one just affecting Y and leaving U and V unaltered (darker shades)
one just affecting U and V and leafing Y unaltered (different colours)

Different but any use of such color? YUV work differently than RGB so changing UV at the same time has no sense - only Y change is useful as such EHB functionality will be exactly the same as currently in RGB and same complaints ( i bet other HW platforms from same age will praise to have such mode as EHB - HW shadows is nice feature as shadow is natural thing).

Originally Posted by Gorf

sure this would have been nice ... but here I fully accept the argument of restrictions of silicon space.
As far as I understand the colour registers take quite some space.

Probably and if they are designed as Dual Port RAM then they can be even more complex than regular SRAM cell.

[pandy71]

Humble reminder - CBM failed not due of the Amiga but PC market collapse - CBM had warehouses full of average/outdated CBM PC's when price of the PC suddenly lowered.
CBM can't compete with other PC makers and simply died.
If CBM focus all resources on Amiga then we can't be sure how long they will axist on market but at some point moving Amiga technology toward PC HW was unavoidable anyway but perhaps Amiga OS could be something different.

[grond]

Originally Posted by Gorf

CBM should have brought the A3000 with higher resolution in 89

Most of all, ECS should have had 32bit DMA for all custom chips, not just for bitplane DMA, and the low-cost Amigas with ECS should have had a CPU with 32bit bus (i.e. the 68EC020) turning them into complete 32bit systems. Oh, and ECS should already have been AGA+chunky. Then 1990 would have been okay.

the A3000+ with DSP in 90

I don't understand this obsession with DSPs. Yes, in 1990 DSPs seemed a very powerful tool that could do things that normal processors could not but it would have been a very specialised expensive processor that would have added a lot of complexity to the system. Furthermore, the DSP would probably have been a resource with no reasonable resource sharing between tasks. I think a DSP should rather have been an optional add-on located on a Z3-card with good support from the OS.

They spent almost nothing on the A3000 in therms of development - what was it? Two or three people working on it?

The major factor in the A3000 development certainly was the development of ECS which probably was a more complex development than the step from ECS to AGA.

They wasted a lot more money on the CDTV at the very same time.

But the CDTV at least was a visionary product and has a prominent place in consumer electronics history. It wasn't a product where everybody but Commodore knew that it would flop before it even was released.

[grond]

Originally Posted by Bruce Abbott

In 1991 the A3000 cost the same as a 'high-end' PC, which is how I justified the outlay. But had Commodore spent even more R&D on it the price would have been even higher.

That's not how a company calculates. Commodore derived the chipset for their low-end computers from the same single investment into the ECS chipset meaning that the A3000 sales were the cream on top of the entire sales volume. Furthermore, the high-end computers generate PR and reputation for the entire line of computers like Amigans still like to point out how some TV series were made using Amigas. Of course, high-end Amigas were used with extra hardware, not lowly A500s...

[Gorf]

Originally Posted by Bruce Abbott

New York Times: January 30, 1991

Yes: 1990 was the last big deal.
Commodore won the tender to "computerize" the offices of the former East-German railway "Reichsbahn".

So Ali was correct about this in January of 1991.

Also were 1990 and 1991 exceptional good years for Commodore (and Atari) due to the fall of the iron curtain - "cheap" (for west European standards) homecomputers were the only thing many people in Eastern Europe could afford in the first couple of years ...

But this was also very misleading for Commodores executive staff. They did not realize that this was just a lucky incident that would not last ...

[Gorf]

Originally Posted by grond

I don't understand this obsession with DSPs. Yes, in 1990 DSPs seemed a very powerful tool that could do things that normal processors could not but it would have been a very specialised expensive processor that would have added a lot of complexity to the system. Furthermore, the DSP would probably have been a resource with no reasonable resource sharing between tasks.

According to Dave Haynie a AOS Library should provide the functions and would have been a good match, since the DSP provided some multitasking internally.
I did not have a closer look at the newly rebuild A3000+, but as far as I understand the DSP it is mostly working now...

But the CDTV at least was a visionary product and has a prominent place in consumer electronics history. It wasn't a product where everybody but Commodore knew that it would flop before it even was released.

But why not a least a 14MHz CPU ... and some real upgrades in the chipset would have also been a good idea.

As I said: the idea was not that wrong, but is was implemented cheaply ...

It was kind of O.K. as a stopgap to the CD32, which came too late of course ...

[grond]

Originally Posted by Gorf

Therefore I suggested to split the EHB mode into two modes:
one just affecting Y and leaving U and V unaltered (darker shades)
one just affecting U and V and leafing Y unaltered (different colours)

I don't think anyone would have seen a 2nd set of 32 colours with UV components changed together according to some predefined ratio useful. I guess a 16 colour / 4 shades per colour EHB-type mode would have been more practical to use than that.

[Gorf]

Originally Posted by pandy71

Different but any use of such color? YUV work differently than RGB so changing UV at the same time has no sense - only Y change is useful as such EHB functionality will be exactly the same as currently in RGB and same complaints ( i bet other HW platforms from same age will praise to have such mode as EHB - HW shadows is nice feature as shadow is natural thing).

Please read again, what I wrote about this mode.

Of course changing UV and keeping Y would not give you "half bright", but different colours with the same brightness!

So you could choose between an "half bright" mode , as we know it, or a "different tone" mode.