@OneTimer1

Quote:

OneTimer1 wrote: Quote: cdimauro wrote: The engineers systematically put all the blame on the (unsuccessful) management, but they are by no Absolute BS Engineers like Jay Miner developed new chipsets before he left the company. The chipset would have used (dual ported) VRam, that costs more than the average A500 buyer would have paid, that's one of the reasons it was never put into production by the management.

Well, THIS is absolute BS: there's no proof of that!
Quote:

There where a lot of improvements made by the engineers of C=, that where not brought into production because the management was never interested, they could have had AGA in the A3000+ with a fast SCSI interface, but the management screwed it.

Making it more expensive...
Quote:

The management had stupid ideas how to sell the Amiga, ideas like a trade in of A500 computers for A500+ or CDTV, during a time when most people feared the A500+ would only be an incompatible upgrade for their existing A500.

The same engineers that developed the ECS (and then AGA. And then... Akiko!)?
Quote:

C= / Amiga Engineers where proud for having one of the best 68k Linux systems,

LOL: best? LESS expensive, yes, but... best in what?
Quote:

one of the first 68030 cards in the world

In 1990? Seriously?
Quote:

and the A2410 Tiga card

Which is pure crap: it's clear that you don't know how it works! Ask Thomas Richter about it, and you'll see...
Quote:

that was never really advertised by the C= managers.

By the selling department, then.
Quote:

The Amiga developers at C= where an tiny group of developers, hindered by the obsolete chip production technologies at MOS, technologies that have been good enough for a 2MH 6502 but couldn't even be used for most of the 8/14 MHz chips used in the Amiga.

RI-LOL: another great BS.

Have you ever took a look at what which logos are printed on the Amiga chips?
Quote:

C=/MOS could have continued the production of the 6502 as an embedded chip, if they would have had the opportunity of updating the old production process from NMOS to CMOS but inseatd they didn't even get the money for proper handling of the waste they produced. https://en.wikipedia.org/wiki/MOS_Technology#GMT_Microelectronics

And? See above: go take a look at the chips and then tell me what you've found. You could be surprised...
Quote:

Instead of focusing on the Amiga, some where forced to develop the C65 that was produced in a small batch of 200 peaces, but dumped by all C= sales departments.

Another pure crap of... the great engineers: adding 256 colours in PLANAR graphics to the C64 which was using the right, good graphic format (PACKED).

Only Commodore engineers made it possible...

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: we can see with regard to the successor of the original chipset, the so-called ECS. ECS wasn't a successor it was just another revision of the OCS similar to how OCS was just a minor revision to the EHB less one used in early A1000 prototypes.

I beg to differ: it introduced several news features. You can't compare it with the small addition of EHB to the first OCS revision.
Quote:

AGA was the "lets see how far we can push it without really going into it" revision.

AGA required TONS of (new) transistors: far away from a revision...
Quote:

AAA would have been far too little far too late (even if C= had survived and fast tracked it to a XMAS 94 release).

Like all Amiga chipsets after OCS.
Quote:

SuperAGA/Apollo is the alternate universe late 90s chipset that would have killed C= or whomever hold the IP in that time&reality for sure.

Unfortunately it suffers of other bad design decisions.
Quote:

-> constant and real updating the HW would have made the Amiga stay relevant till the late 90s at best at which point a break from the old designs would have been overdue. Either to 3rd party chips or if the had a 10x bigger cashflow to inhouse legacy free chips.

What about Apple then?
Quote:

Both would have required constant OS upgrades forcing developers to go through proper APIs instead of hitting bare metal. Neither happened in the 80s and the efforts done in the 90s were insufficient.

I agree on this.

@Kremlar

Quote:

Kremlar wrote: @OneTImer1 Quote: Engineers like Jay Miner developed new chipsets before he left the company. The chipset would have used (dual ported) VRam, that costs more than the average A500 buyer would have paid, that's one of the reasons it was never put into production by the management. @kamelito Quote: When the original team left the next chipset was finished they didn’t use it. Not true, at least according to Jay Miner. Sometime in the late 80s I think I found a BBS run by Jay Miner. And like the fanboy kid I was, I peppered him with questions including rumors of a new chipset they had designed before leaving Commodore. His reply was "Nope, never was such a thing". I had the foresight to print out that chat, and I ran across it years ago and scanned it to PDF. I posted it here years ago but can post it again if you are interested. Back then, from my point of view, Commodore fans like myself were hanging on the idea of a rumored groundbreaking new graphics chipset to put Amiga back in front. Not much came with the 3000, and only a minor AGA upgrade with the 1200/4000. Not enough to put Amiga back on the podium. This is an interesting topic and one I've thought about in the past. Commodore engineers are treated like celebrities, while Commodore's downfall is blamed on management. I suspect there was a lot of blame to go around. What resources were utilized when creating the original Amiga computer & chipset? How many people, how many hours, how much $$? Were those resources not duplicated under Commodore's management? I'm not sure. With all the time/effort put into other oddball things, I find it hard to believe Commodore wouldn't have wanted an updated graphics chipset. Perhaps the engineers on staff just weren't up for the task? Perhaps engineers underestimated the need for a major graphics upgrade?

*
Quote:

In my eyes, the best thing done under Commodore management and engineering was the continued development of AmigaOS. The hardware upgrades were marginal, but the OS jumped forward tremendously.

Hum. Yes, many features introduced, but also some big crap (TAGS & BOOPSI).

And nothing to address the security features & stability of the platform.

@cdimauro
Quote:

kamelito wrote: When the original team left the next chipset was finished they didn’t use it. AFAIR there's not proof of that. Different opinions are floating around this topic.

Per Jay Minder, it's simply not true.

@Kremlar

There are interviews from Jay Miner and Ranger prototypes (looks like one is displayed at latest AmiWest show) yet C= engineers like Dave Haynie know little about it. Jay Miner was payed by C= as a consultant after he left. It feels like C= buried Ranger and hushed up Jay.

Jay Miner Quote:

Commodore now has a high resolution chip set of Amiga chips that I worked on when we were with Amiga in Los Gatos. These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously. They increase the display address range to 2 Megabytes. These chips are completed and tested and only require a computer and memory to hold them together. I'd like everyone to know that Amiga in Los Gatos designed these chips! These improved Amiga chips can use a new type of ram called video ram. This new type of ram, video ram, is a giant step in computer improvement because it frees up the bottleneck into memory caused by competition between the computer itself and the memory fetchers required for the high resolution display. Imagine having an additional gigantic parallel output port thousands of bit wide, just for video. You wouldn't have to access it very often to dump a lot of memory data to a video picture.

http://www.bambi-amiga.co.uk/amigahistory/jayminerinterview.html
http://www.bambi-amiga.co.uk/amigahistory/jayinterview2.jpg

Last edited by matthey on 28-Oct-2023 at 12:20 AM.

@Kremlar

Quote:

Kremlar wrote: @cdimauro Quote: kamelito wrote: When the original team left the next chipset was finished they didn’t use it. AFAIR there's not proof of that. Different opinions are floating around this topic. Per Jay Minder, it's simply not true.

Quote:

matthey wrote: @Kremlar There are interviews from Jay Miner and Ranger prototypes (looks like one is displayed at latest AmiWest show) yet C= engineers like Dave Haynie know little about it. Jay Miner was payed by C= as a consultant after he left. It feels like C= buried Ranger and hushed up Jay. Jay Miner Quote: Commodore now has a high resolution chip set of Amiga chips that I worked on when we were with Amiga in Los Gatos. These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously. They increase the display address range to 2 Megabytes. These chips are completed and tested and only require a computer and memory to hold them together. I'd like everyone to know that Amiga in Los Gatos designed these chips! These improved Amiga chips can use a new type of ram called video ram. This new type of ram, video ram, is a giant step in computer improvement because it frees up the bottleneck into memory caused by competition between the computer itself and the memory fetchers required for the high resolution display. Imagine having an additional gigantic parallel output port thousands of bit wide, just for video. You wouldn't have to access it very often to dump a lot of memory data to a video picture. http://www.bambi-amiga.co.uk/amigahistory/jayminerinterview.html http://www.bambi-amiga.co.uk/amigahistory/jayinterview2.jpg

OK, then there was a prototype at least and even more: chips were completed and tested.

It would be interesting to find something which proves it. Something should be left somewhere: the prototype, the chipset schema, some documentation.

However what scares me is this:

"These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously."

it looks something new and different from the original chipset. So, not an evolution of it.

cdimauro Quote:

OK, then there was a prototype at least and even more: chips were completed and tested. It would be interesting to find something which proves it. Something should be left somewhere: the prototype, the chipset schema, some documentation. However what scares me is this: "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously." it looks something new and different from the original chipset. So, not an evolution of it.

You may be taking what Jay Miner is saying too literally. The chips being completed and tested from his perspective may not mean actual chips exist.

1. specification
2. design
3. prototype circuits (breadboard/wirewrap)
4. testing and verification
5. chip layout
6. chip production

Jay may have considered the chip to be done when he reached stage 4 or 5 above because that is where he handed off to others. It's possible there were prototype chips produced but that likely would have been the job of MOS/CSG and required C= approval which was the problem. The Ranger case design prototype and circuits look like they exist as can be seen at the most recent AmiWest in the following video.

Amiwest 2023 Commodore Amiga Show - Walk Around & Chat - My Show Experience - Retro Computer Miggie
https://youtu.be/4b1TerCwVTo?t=1418

Your interpretation of, "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously" may also be too literal. It probably just means that the Ranger chipset display is compatible with the OCS display modes which can be displayed simultaneously. The BoingBall demo where the screen is pulled down to display the Workbench screen behind it was popular then and displayed two screen modes simultaneously.

@cdimauro

Quote:

I beg to differ: it introduced several news features. You can't compare it with the small addition of EHB to the first OCS revision.

ECS Denise can be dropped into any OCS Amiga, (1MB) ECS Agnus in A500/2000 with minor to no modifications. The rest remained the same.
"New features" like SuperHighRes which just used the same bandwidth in a different way (and hardly anybody used) or the bigger blits that also were barely used?

Quote:

AGA required TONS of (new) transistors: far away from a revision..

It did again not upgrade outdated parts (1(2 speed HD floppies and UARTs without FIFO buffers). It's "just" using the extra bandwidth for more colors in the same flawed bitmap modes.

Quote:

What about Apple then?

Yeah, what about Apple?
The always had just enough HW to function and at some point either late in 68k or early PPC they switched to off the shelf GPUs and eventually GFX cards.
Early iPhones also used someone else's chips and only when they sold more in a year than C= sold Amiga over the whole 9 years (or C64s for that matter) they switched to do it in house based on licensed IP.
When that stuff matured and sold in 9 figures they started putting it into Macs.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@cdimauro

Quote:

cdimauro wrote: OK, then there was a prototype at least and even more: chips were completed and tested. It would be interesting to find something which proves it. Something should be left somewhere: the prototype, the chipset schema, some documentation.

Yes, it would be great to find some of that stuff. Meanwhile we'll just have to survive on reading between the lines and perpetuating myths.

I was not impressed by pictures of the 'Ranger' prototype. In this 1988 interview Jay Miner confirmed that its cards would be 'not shaped like IBM cards', IOW even less compatible with PC stuff. He complains about Commodore in Westchester and Germany not producing the A500 and A2000 in 1986 like they promised, and not cost-reducing the A1000. But the A500 proved to be their most successful model, and the A2000 was a great workhorse for serious applications. The A1000 really needed to die at that point (and I say that as someone who bought one in 1987 when I could have had an A500 for less!). I think Commodore made the right decision choosing them over the Los Gatos design.

Quote:

However what scares me is this: "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously." it looks something new and different from the original chipset. So, not an evolution of it.

This sounds suspiciously like a flicker fixer, or perhaps a separate graphics card. Jay Miner says chips were 'completed and tested', but 'only require a computer and memory to hold them together'. IOW there wasn't an actual prototype using these 'finished' chips.

Another interesting comment Jay Miner made was that Commodore made improvements to his original chipset before putting it in the A1000. He said 'The Amiga originally only had 320 colors across the screen, even in 640 mode. They helped us put full color in the 640 mode.', and 'they improved the color by moving the NTSC converter off the chip'. IOW, the Amiga originally didn't have RGB, and the color resolution was stuck at 320 (horizontal) even in hires. Imagine that! Without Commodore's input the Amiga would have been totally unsuited to serious applications. I for one would not have bought one - not after getting the Amstrad CPC664 with its nice crisp RGB and 640x200 mode that was actually readable on a monitor.

@matthey
Quote:

matthey wrote: cdimauro Quote: OK, then there was a prototype at least and even more: chips were completed and tested. It would be interesting to find something which proves it. Something should be left somewhere: the prototype, the chipset schema, some documentation. However what scares me is this: "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously." it looks something new and different from the original chipset. So, not an evolution of it. You may be taking what Jay Miner is saying too literally.

He's THE technical guy, so I expect that he uses the right words.
Quote:

The chips being completed and tested from his perspective may not mean actual chips exist. 1. specification 2. design 3. prototype circuits (breadboard/wirewrap) 4. testing and verification 5. chip layout 6. chip production Jay may have considered the chip to be done when he reached stage 4 or 5 above because that is where he handed off to others. It's possible there were prototype chips produced but that likely would have been the job of MOS/CSG and required C= approval which was the problem.

Yes, that could be the case. Maybe it was stage 4, at this point in time.
Quote:

The Ranger case design prototype and circuits look like they exist as can be seen at the most recent AmiWest in the following video. Amiwest 2023 Commodore Amiga Show - Walk Around & Chat - My Show Experience - Retro Computer Miggie https://youtu.be/4b1TerCwVTo?t=1418

That looks like only the case design. The picture says also "mock-ups".

It's difficult to state that it was holding a motherboard with the prototype.
Quote:

Your interpretation of, "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously" may also be too literal. It probably just means that the Ranger chipset display is compatible with the OCS display modes which can be displayed simultaneously. The BoingBall demo where the screen is pulled down to display the Workbench screen behind it was popular then and displayed two screen modes simultaneously.

As I've said before, Jay was the technical guy and I expect that he knows which words to use and their meaning.

As Bruce also stated, this looks like and independent component, like an additional graphic card.


@bhabbott
Quote:

bhabbott wrote: @cdimauro Quote: cdimauro wrote: OK, then there was a prototype at least and even more: chips were completed and tested. It would be interesting to find something which proves it. Something should be left somewhere: the prototype, the chipset schema, some documentation. Yes, it would be great to find some of that stuff. Meanwhile we'll just have to survive on reading between the lines and perpetuating myths. I was not impressed by pictures of the 'Ranger' prototype. In this 1988 interview Jay Miner confirmed that its cards would be 'not shaped like IBM cards', IOW even less compatible with PC stuff.

That goes on the direction of an additional card. So, not like and evolved OCS chipset.
Quote:

He complains about Commodore in Westchester and Germany not producing the A500 and A2000 in 1986 like they promised, and not cost-reducing the A1000. But the A500 proved to be their most successful model, and the A2000 was a great workhorse for serious applications. The A1000 really needed to die at that point (and I say that as someone who bought one in 1987 when I could have had an A500 for less!). I think Commodore made the right decision choosing them over the Los Gatos design.

I think so. The Amiga 500 and 2000 lines of computers was the right decision, as we can see on how they sold.

Only one think: 2 years late from the Amiga 1000 introduction was too much. Here we missed Jack Tramiel vision and strategy, IMO.
Quote:

Quote: However what scares me is this: "These chips use video ram and can produce a very high resolution 1024 display along with the present Amiga display simultaneously." it looks something new and different from the original chipset. So, not an evolution of it. This sounds suspiciously like a flicker fixer, or perhaps a separate graphics card. Jay Miner says chips were 'completed and tested', but 'only require a computer and memory to hold them together'. IOW there wasn't an actual prototype using these 'finished' chips.

Indeed. Everything goes on the direction of something unfinished and, more important, independent from the original chipset.
Quote:

Another interesting comment Jay Miner made was that Commodore made improvements to his original chipset before putting it in the A1000. He said 'The Amiga originally only had 320 colors across the screen, even in 640 mode. They helped us put full color in the 640 mode.', and 'they improved the color by moving the NTSC converter off the chip'. IOW, the Amiga originally didn't have RGB, and the color resolution was stuck at 320 (horizontal) even in hires. Imagine that! Without Commodore's input the Amiga would have been totally unsuited to serious applications. I for one would not have bought one - not after getting the Amstrad CPC664 with its nice crisp RGB and 640x200 mode that was actually readable on a monitor.

If it was really the case, then I agree on this as well. Commodore had its expertise here and it was a Good Thing for the Amiga.

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: I beg to differ: it introduced several news features. You can't compare it with the small addition of EHB to the first OCS revision. ECS Denise can be dropped into any OCS Amiga, (1MB) ECS Agnus in A500/2000 with minor to no modifications. The rest remained the same.

Then how do you use SHRES mode without the ECS Agnus?

Agnus is the orchestrator and schedules the fetch of data by the several DMA channels.
When the system operates in low-res and high-res mode, the bitplane datas are fetched very differently. I expect the same when activating the SHRES mode, since now you've to fetch 2 bitplanes at the time, repeated by 4 times, inside the usual 8 memory slots window.
Quote:

"New features" like SuperHighRes which just used the same bandwidth in a different way (and hardly anybody used) or the bigger blits that also were barely used?

Well, that's exactly the point of my article.

However here the context is different: ECS has introduced SEVERAL new features, so it's a new chipset version.

That almost all of those features were useless I agree, but that's another story.
Quote:

Quote: AGA required TONS of (new) transistors: far away from a revision.. It did again not upgrade outdated parts (1(2 speed HD floppies and UARTs without FIFO buffers).

Those were very MINOR parts, however.
Quote:

It's "just" using the extra bandwidth for more colors in the same flawed bitmap modes.

I agree, but nevertheless: those were HUGE changes. Which required A LOT of transistors to be implemented.

So and definitely NOT a simple revision of the chipset.
Quote:

Quote: What about Apple then? Yeah, what about Apple? The always had just enough HW to function and at some point either late in 68k or early PPC they switched to off the shelf GPUs and eventually GFX cards.

Whatever, but Apple had ITS hardware AND software AND it survived. Commodore died.
Quote:

Early iPhones also used someone else's chips and only when they sold more in a year than C= sold Amiga over the whole 9 years (or C64s for that matter) they switched to do it in house based on licensed IP. When that stuff matured and sold in 9 figures they started putting it into Macs.

This is very recent history, however: FAR AWAY from Commodore's demises...

I like ECS.
ECS was released in 1990
and in 1990 still has batter graphics than affordable pc.
It was AGA that kill Commodore.

@ppcamiga1

If design a new gfx chip in 3 years, and invested in a software libraries to rival Microsoft.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@cdimauro

Quote:

cdimauro wrote: This is very recent history, however: FAR AWAY from Commodore's demises...

Apple did start supporting 3rd party GFX chips when C= was still around which made the quality of their own GFX HW a non issue by the time AAA-Amiga would have been released in a "perfect" timeline.

As for ECS being a revision/version, the point is very little new (that really matters) features were added while all the old stuff remained just as it was.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

Kronos wrote: But there is that story that HP wanted to sell/rebrand the A3000UX running AMIX as their entry level UNIX workstation.

Sun, not HP.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@ppcamiga1

Quote:

ppcamiga1 wrote: I like ECS. ECS was released in 1990 and in 1990 still has batter graphics than affordable pc. It was AGA that kill Commodore.

LOL I suggest you to change your avatar with a parrot, because this is actually what you're: a parrot always repeating the same crap.

Why don't tell me about your beloved PowerPCs which are just regular PCs with the CPU replaced with way underpowered and crap PowerPC ones?
And, of course, using all PC "crap": USBs, 8086 firmware (to be EMULATED!) for SVGA graphic cards, PC sound cards, PS/2 and/or USB PC keyboards and mice, and so on.

All of that looks like an Amiga right?

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: cdimauro wrote: This is very recent history, however: FAR AWAY from Commodore's demises... Apple did start supporting 3rd party GFX chips when C= was still around which made the quality of their own GFX HW a non issue by the time AAA-Amiga would have been released in a "perfect" timeline.

But you were talking about IPHONES on this specific context!
Quote:

As for ECS being a revision/version, the point is very little new (that really matters) features were added while all the old stuff remained just as it was.

Let's say this: ECS added very little USEFUL things.

But it was a new chipset.

@cdimauro

Quote:

cdimauro wrote: But you were talking about IPHONES on this specific context!

Nope, just pointing out that the iPhone was where Apple eventually went back to doing custom GFX HW.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: cdimauro wrote: But you were talking about IPHONES on this specific context! Nope, just pointing out that the iPhone was where Apple eventually went back to doing custom GFX HW.

Which is, still, not the case: https://en.wikipedia.org/wiki/Early_iPhone_systems-on-chip

"iPhone and iPod Touch models released between 2007 and 2009 used system on a chip (SoC) circuits designed by Samsung and manufactured to Apple's specifications.
[...]
Apple later switched to SoCs that are designed in-house, starting with the Apple A4."

But, regarding the GFX, not even its first in-house SoC had a custom GFX unit: https://en.wikipedia.org/wiki/Apple_A4

"The A4 also features a single-core PowerVR SGX535 graphics processing unit (GPU)."

Only in the recent years Apple designed a custom GFX unit.

So, again, not a good example

Anyway, what was important in THIS specific part of the discussion was the period where Apple competed with Commodore. So, from 1985 to 1994.

Facts: Apple had custom chips and o.s., but it survived. Commodore didn't. The curtain falls...

@cdimauro

Quote:

cdimauro wrote: Let's say this: ECS added very little USEFUL things.

More chipram and more screen modes were darn useful, on Linux, ECS is required for VGA screenmodes.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC