Things the Amiga didn't get right from Day 1

[Gorf]

Originally Posted by Bruce Abbott

But hey, why not do it? Then Commodore could tell all those A1000 owners that their machines are redundant now so they will have buy the 'next generation' if they want to run the latest software. Then do it again next year. Massive profits guaranteed!

No need to be sarcastic here:
That was due to the rapid technical improvements very normal in the 80s and 90s
Yes: you had to have the latest hardware to run the latest software. That was true for PCs, Macs, SUNs ….

Not updating the features of their computer platform fast enough was one of the biggest mistakes of commodore.

And actually: commodore offered A1000 owners to give it back in exchange for a A2000 at a reduced price and many took that offer - even though the A2000 did not upgrade the chipset (in a meaningful way).

[NorthWay]

Didn't the ST use 150ns ram? (Overclocked too AFAIK.)

[drHirudo]

Originally Posted by Gorf

Yes: you had to have the latest hardware to run the latest software. That was true for PCs, Macs, SUNs ….

Commodore didn't update the C64 for more than a decade, yet it was still selling with a big margin, since it was costing almost nothing to produce them with their own CPU and chipset at MOS technology. They were seeing the Amiga as 16-bit C64. i.e. the Amiga3000UX was the C128 of the Amiga's - running alien and totally unsupported by the chipset operating system. In the late 1980-ies Commodore was manufacturer of wide variety of systems - 8-birs, 16-bit, IBM-PC Compatibles, Monitors, Peripherals and many more.

Would have they survived if they focused only on the Amiga? Not sure.

Atari were mostly focused on their ST-line and later the Falcon (a monster 68K machine with DSP etc..), along with some Atari Lynx and other side projects, yet they didn't survive.

[Gorf]

Atari also sold PCs for a while and Jack made some bad decisions… the worst was probably to buy "Federated Group" a electronics reseller in the USA. They faked the numbers and Atari did not only pay way too much, but was confronted with a huge loss in the first year …

They also kept the Atari 8-bit line too long.

And like Commodore: the did not update the ST line fast enough .. the TT faced the same problems as the A3000 for almost the same reasons … only the (monochrome) resolution was great: 1280x960

And like the Amiga 1200/4000 the Falcon was simply too late … and it’s design is equally flawed as the A1200:
while the A1200 has no FastRAM and is only half as fast as it could be, the Falcon only has a 16bit wide RAM interface on a 32bit CPU.

If both companies had merged and unified their platform around 1990 (the Amiga can emulate the ST 100%) there might have been a chance to survive.

[pandy71]

Originally Posted by Bruce Abbott

Some DRAMs were capable of 'page' mode in 1984, but a lot weren't - so using page mode would have limited what DRAM chips they could use.

Some DRAM was capable page mode since at least 1982 (like Motorola MCM 6664A) but still some first 41256 was not equipped with FPM...

Originally Posted by Bruce Abbott

The A1000 only shipped with 256k of 150ns DRAMs, and they were not page mode capable.

Lot of people forget that A1000 don't use 41256 chips but 4464 and most of them was not FPM

Originally Posted by Bruce Abbott

The Spectrum still needed clock stretching to slow the CPU down in video memory.

4116 and 4132/4164 for sure was incapable of any modern DRAM features and ULA hold CPU when both trying to access video RAM.
https://sinclair.wiki.zxnet.co.uk/wiki/Contended_memory

[pandy71]

Originally Posted by Gorf

No need to be sarcastic here:
That was due to the rapid technical improvements very normal in the 80s and 90s
Yes: you had to have the latest hardware to run the latest software. That was true for PCs, Macs, SUNs ….

Not updating the features of their computer platform fast enough was one of the biggest mistakes of commodore.

Commodore firstly need some return from investment in Amiga technology, A1000 was expensive so CBM focused not on adding features but rather on fixing most important issues and cost reduction - as such A500 was born - without A500 probably Amiga will die sooner than later due overall low sale.
I don't blame CBM for this strategy.

[pandy71]

Originally Posted by Gorf

And like Commodore: the did not update the ST line fast enough .. the TT faced the same problems as the A3000 for almost the same reasons … only the (monochrome) resolution was great: 1280x960

CBM offered A2024 with same or higher resolution.

[Gorf]

Originally Posted by pandy71

CBM offered A2024 with same or higher resolution.

They did - but it was (like the fickerfixer) a kludge.
The A2024 and the "Hedley-device" used expensive dual-ported RAM to store 4 normal screens, Denise would display one after the other ...

Sadly Commodore did not even manage to merge the Hedley-device and Amber for the A3000 ... this would have been still be an awkward solution, but a step in the right direction.

[Gorf]

Originally Posted by pandy71

Commodore firstly need some return from investment in Amiga technology, A1000 was expensive so CBM focused not on adding features but rather on fixing most important issues and cost reduction - as such A500 was born - without A500 probably Amiga will die sooner than later due overall low sale.
I don't blame CBM for this strategy.

These both things are not mutually exclusive.
It was just the thinking of Commodore, the had fallen out of time:
release a model and never touch it again, but go on to the next incompatible thing....

That somehow worked for the VIC20 and the C64 ... despite of the flaws .. but it did not work für Commodores business line of computers (which should have been a warning) and not very well for the C16 and Plus/4.

And of course it did not really work out for the Amiga and when Commodore realized that operating systems and chipsets are never "finished" or "done", but need constant development end evolution, it was to late: all the geniuses where gone...

[pandy71]

Originally Posted by Gorf

They did - but it was (like the fickerfixer) a kludge.
The A2024 and the "Hedley-device" used expensive dual-ported RAM to store 4 normal screens, Denise would display one after the other ...

yes - kludge or some workaround but possible thanks to flexibility of the architecture, expensive - perhaps for average consumer but not for those who rely on high resolution - probably still cheaper than anything comparable on market...

Originally Posted by Gorf

Sadly Commodore did not even manage to merge the Hedley-device and Amber for the A3000 ... this would have been still be an awkward solution, but a step in the right direction.

Worse, they can't manage super hires to be properly processed by Amber... but this is quite clear why not - double the memory and pushing clock beyond CSG capability...

[Bruce Abbott]

Originally Posted by Gorf

No need to be sarcastic here:
That was due to the rapid technical improvements very normal in the 80s and 90s
Yes: you had to have the latest hardware to run the latest software. That was true for PCs, Macs, SUNs ….

On the contrary, sarcasm is needed because too many people accept this situation.

Not updating the features of their computer platform fast enough was one of the biggest mistakes of commodore.

Other 'alternative' computer manufacturers brought out new models more often, but it didn't help them. Commodore's biggest mistake was thinking they could just make non-PC compatible hardware and the software (which is what people buy computers for) would follow. But it took years for developers to get to grips with the power of the Amiga's custom chips, by which time it had no chance against the enormous PC market.

In fact the lack of software titles was the Amiga's Achilles heel from day one. They should have been able to answer the question 'But is it IBM compatible?' with 'It doesn't need to be. We have every app and game you might want and more, so forget about PC compatibility!".

And actually: commodore offered A1000 owners to give it back in exchange for a A2000 at a reduced price and many took that offer - even though the A2000 did not upgrade the chipset (in a meaningful way).

There's a lot more to the A2000 than the chipset.

[pandy71]

Originally Posted by Gorf

I despise spreadsheets,

Made spreadsheet for you - LibreOffice (can be portable so no install required) with attached add-on (install add-on to get RGB color preview) and spreadsheet - test by yourself how 4 bit quantization range affect RGB output - for example there is not possible to create grayscale with such limited quantization range.

Add-on made by :
https://forum.openoffice.org/en/foru...470676#p470676
seem there is newer version
https://forum.openoffice.org/en/foru...484842#p484842

[Bruce Abbott]

Originally Posted by Gorf

Sadly Commodore did not even manage to merge the Hedley-device and Amber for the A3000 ... this would have been still be an awkward solution, but a step in the right direction.

The A3000 overall was a step in the wrong direction. By 1990 Commodore should have realized they would never compete head-to-head against PCs. But there was still a large niche to exploit at the low end, with a little help from the cheap PC-inspired hardware coming out (IDE hard drives etc.). That is why the A1200 would have been a better direction to go in. Unfortunately they discovered that about 2 years too late.

[Gorf]

Originally Posted by Bruce Abbott

The A3000 overall was a step in the wrong direction. By 1990 Commodore should have realized they would never compete head-to-head against PCs. But there was still a large niche to exploit at the low end, with a little help from the cheap PC-inspired hardware coming out (IDE hard drives etc.). That is why the A1200 would have been a better direction to go in. Unfortunately they discovered that about 2 years too late.

It’s not so much about PCs (office) but about the academic fields, about engineering, about rendering, simulations and so on - low end simply was an economical death trap.
Not only for Commodore, but for all other low end systems as well.

And especially in gaming the hardware sets the limit: and while the Amiga was certainly great the first 5 years in this particular field it was attacked from PCs and game consoles alike and simply got steamrolled due to the lack of innovation and vision by Commodore…

99% of gamers just would not care about the system: it only mattered to them if it can run Doom …

But the Amiga spirit is so much more - and especially the part that makes it unique and so much fun is the combination of the operating system and the hardware capabilities. True multimedia before anyone knew that term.

And multimedia always included animations… the first demos in the Amiga emphasized this … and animations, video, abundance of colours, renderings and so on need more power - always.
That is why the Amiga as a system would have needed more development and more powerful hardware to survive. It would have needed models that satisfy the needs of digital artists - as well in the graphics department as well as in the sound department.

Apple got lucky the Mac was adopted by the DTP crowd… but it is a shame the Amiga did not became the natural platform for image manipulation and processing, for rendering and even for CAD and of course for digital audio workstations.
These are the fields it initially showed great prospect and started a lot of developments … but these soon moved on, because Commodore did not provide suitable powerful hardware.
The demand was there, the market was there - CBM wasn’t.

[Gorf]

Originally Posted by pandy71

Made spreadsheet for you - LibreOffice (can be portable so no install required) with attached add-on (install add-on to get RGB color preview) and spreadsheet - test by yourself how 4 bit quantization range affect RGB output - for example there is not possible to create grayscale with such limited quantization range.

Add-on made by :
https://forum.openoffice.org/en/foru...470676#p470676
seem there is newer version
https://forum.openoffice.org/en/foru...484842#p484842

Thank you very much! I will have a close look at this tomorrow.
Just for now without looking at it:
YUV at 4 Bit per channel would give you 16 shades of gray - the same as the OCS/ECS provides now…

Edit:
Just because I don’t like spreadsheets, does not mean in don’t have all commonly used programs of this kind installed - I do. ;-)

[Gorf]

Originally Posted by pandy71

Worse, they can't manage super hires to be properly processed by Amber... but this is quite clear why not - double the memory and pushing clock beyond CSG capability...

This very much illustrates the lack of communication and leadership within CBM.
While Hedley and Haynie were working an a workstation-like Amiga the chip-design team was … well not entirely sure what they were trying to archive with SHRES …
To call this mode "half baked" would be very polite.

At leased ECS is flexible enough to be programmed to have some 800*600 mode (interlaced) - at least Amber should have been capable of deinterlace such a mode. I did some calculations a while ago and came to the conclusion the amount of dual-ported RAM would have been sufficient, but Amber has a hard 640 pixel limit for one row…

[chb]

Originally Posted by Gorf

well not entirely sure what they were trying to archive with SHRES …
To call this mode "half baked" would be very polite.

IMHO with SHRES they just found some application for the new 35 ns modes while keeping a TV-compatible signal. Maybe it even would have been more complicated to leave it out. It is moderately useful e.g. for video titling, because the "oversampling" provides some anti-alias.

But yes, it shows clearly lack of focus to release a new chipset in the A3000 whose main features (new resolutions) are made obsolete by the expensive kludge that amber was. I suspect that ECS predates the A3000 by several years and was considered to be underpowered in 1990 for a high-end machine, plus they chose not to implement the VRAM features (probably rightfully so, that would have been another kludge).

[Overdoc]

I think the main problem was that Commodore decided to make the Amiga both a gaming and professional computer.

They made probably at least 80% of sales from gamers, who bought Amiga 500s and C64s before.
But they did not bring any real new hardware on the market for those gamers until they went bancrupt.
Basically, gamers would still have to play Amiga games in 1994 on more or less the same hardware like on a 1987 Amiga 500.

AGA was not much difference (at least in games), sound was still the same, speed was not much improved if you did not have some expensive add on turbo cards, and there was no good 3D grafic chip (were there any 3D games at all which used the Akiko chip of the CD32?!?)

But I think the main reason why Commodore could not develop a completely new gaming computer was that they held on to the Amiga system for much too long, tried to be backwards compatible, etc.

In my opinion, they shoud have splitted it up:
Continue developing the Amiga as a professional computer system, while developing a completetly new gaming computer system, which would have had full power 3D grafic support, which clearly was the way where gaming went in the early ninetees.

Maybe it would not even have been a computer, but a gaming console?
(what the CD32 could have been, if it had been a real new 32 bit 3D gaming system, and not just more or less an A1200 without keyboard, but with CD drive...)

Commodor split these market segments up before the Amiga. There was the PET line of computers for professionals, and there was VC20 and C64 for home users.
So why did they not return to this later?

[pandy71]

Originally Posted by Gorf

Thank you very much! I will have a close look at this tomorrow.
Just for now without looking at it:
YUV at 4 Bit per channel would give you 16 shades of gray - the same as the OCS/ECS provides now…

Check calculations - 4 bit quantization range prevent chroma to be truly bipolar such gray is always with some tint (IMHO green tint is most neutral).

Of course we can limit chroma to 15 unique levels then problem is solved OR something i really miss from 1-st day - 15 bits per pixel - but i understand limit - 48 pin packaging for Denise. If Amiga would offer 15 bit per pixel...
Sadly Amiga designers don't decided to do some pin multiplexing...

[pandy71]

Originally Posted by chb

IMHO with SHRES they just found some application for the new 35 ns modes while keeping a TV-compatible signal. Maybe it even would have been more complicated to leave it out. It is moderately useful e.g. for video titling, because the "oversampling" provides some anti-alias.

35ns resolution for scrolling (smooth ticker bar) and overlay vs background positioning was really appreciated by video titling software - cheap Amiga with cheap software can compete directly with machine character generators used in video broadcast at fraction of cost and offering with ECS more genlock features.