Was anyone else disappointed with the A1200?

[alexh]

Originally Posted by sandruzzo

with raspberry Even copper and blitter can be pushed further(more than 68060)?

I don't think so. They are Chip RAM bus bandwidth limited not CPU limited.

[hammer]

Originally Posted by Bruce Abbott

In some ways it was, in others not.

In 1993 printer manufacturers were still supporting 'standard' protocols. But in the rush to the bottom this would soon change. Pentium CPUs and Windows 95 allowed 'GDI' printers to take off. No reason they couldn't be used on the Amiga but one - secret proprietary protocols and drivers that only worked in Windows.

In terms of the competitive market, Amiga wasn't isolated. Do you understand the purpose of exclusivity for a particular platform?

Apple's MacOS and NextSTEP had PostScript. I had TurboPrint on the Amiga.
Apple Mac dominated DTP.

The Amiga had timed exclusive with the Batman game with the Commodore UK's Batman pack.

Exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

Same thing happened with MODEMs. The first 'Wintel' MODEMs were awful because they sucked up all the CPU time. But faster Pentium CPUs eventually fixed that - then external MODEMs disappeared.

I still have USB 56K external modems and they were used for sending faxes.

This issue wasn't didn't break Commodore's core revenue streams.

Exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

Same thing would have happened with PCMCIA network cards too, if 'someone' hadn't ignored the naysayers and said "hey, why can't this work on the Amiga?" and developed a driver for it.

Reminder, the A300 project is to replace C64's price segment. Blame Commodore Germany for A300's scope creep that led to A500's cancelation and revenue tanking in 1992.

A500's higher price segment can have PCMCIA.

Exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

On the software side, most business apps only worked on PCs and their proprietary file formats were not published. My accountants wanted me to use the same DOS accounting package as them (written in Microsoft QuickBASIC) so they could plug my files into their system. I refused, and wrote a file conversion program to give them the required data. Most people wouldn't do that, they would buckle under and buy a PC.

Amiga's business market target is small and it wouldn't sustain Commodore.

Commodore had two main revenue pillars i.e. C64C and A500. A600 was a sales flop and with A500's cancellation, it was a disaster for Commodore's 1992 revenue stream.

When the Amiga lost its "power without the price" gaming advantages, Commodore was pushed out of the mainstream market.

Apple's DTP dominance is a larger market when compared to Video Toaster's.

Amiga's Video Toaster market target is tiny and it wouldn't sustain Commodore.

My point, Amiga's primary entry point into the user's hands is with games and it's the same for gaming PC's assault against market segments.

Gaming PC's GPU (i.e. NVIDIA's GeForce) to mount a killing assault against SGI's graphics workstation business.

Gaming PC's GpGPU (i.e. NVIDIA's GeForce CUDA) to mount an assault against IBM's CELL. GeForce CUDA is still alive and warp/wave32 length is the de-facto standard that forced AMD to follow NAVI's wave32 length.

Exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

The Amiga suffered because people took the 'easy' route of just using the proprietary file formats of Microsoft Office etc., rather than saving files in standard formats. Even PC users suffered.

I don't recall this being a major issue when my high school and university had Apple Mac labs.

MS Works has compatibility with XLS and DOC file formats.

Exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

Office wasn't cheap but you needed it to read that letter or spreadsheet someone sent you.

Lower-cost MS Works has compatibility with XLS and DOC file formats.

The workaround is to run MacOS 68K's MS Office Mac on the Amiga and that's what I did with 1995-era Shapeshifter on my A3000. This workaround doesn't advance the Amiga's platform.

Originally Posted by Bruce Abbott

The Internet had open standards, but that wouldn't last either - how can you force people to use your apps in that environment? So Netscape introduced proprietary HTML extensions that only worked with their browser, including the performance sucking security disaster they called JavaScript. Naturally Microsoft had to do the same, and the rest is 'history'.

MacOS 68K runs Netscape. My A1200+Pistorm-Emu68+Shapeshifter doubles as my retro MacOS just like my school's and university's Mac labs. My university had PowerMac labs.

Again, this workaround doesn't advance the Amiga's platform.

My family was close to selecting the Apple Mac in the 1990s. The PC's 3D games library is the breaker. Again, exclusivity is a competitive advantage over the competition.

Originally Posted by Bruce Abbott

Luckily for us some PC game producers still supported the Amiga. Even if they were only EGA ports we still got to play the games so it was OK. But some didn't. John Carmack famously refused to let Doom be ported to the Amiga because a stock A500 couldn't run it. Of course by this time Commodore was gone, and with no new Amigas being produced it was dismissed by mainstream game producers, which is understandable. But some gave up long before that. This was also understandable when porting your PC game to the Amiga might only net 10% more sales, but it still isolated the Amiga platform.

My statement with Amiga is not being isolated refers to a competitive environment. When a platform has exclusivity, it's a competitive advantage over other competitors.

From Q4 1992, Commodore was pushed out from the bottom price segment SNES and falling PC prices at the mid-price segment.

Originally Posted by Bruce Abbott

And you had to use them if you wanted more than 16 colors. No tuning the number of bitplanes to the colors you needed.
No dual playfields, no split resolutions or changing color palettes on a line by line basis, no sprites and no vsync to prevent tearing. All rendering had to be done with the CPU so you needed a fast one. I saw the results on a 16MHz 386SX and it wasn't pretty. Chunky pixels didn't help much if at all.

Is this applicable for Q4 1992's A1200 context?

SNES's and PC's Mortal Kombat port killed the Amiga version.

Originally Posted by Bruce Abbott

And EGA only had 16 colors from a yucky 64 color palette. Some artists did amazing stuff with it, but those garish colors were so limiting. No wonder PC users went gaga over VGA!

Is this applicable for Q4 1992's A1200 context?

[hammer]

Originally Posted by Bruce Abbott

But with a 50MHz 030 in the A1200 Doom runs identically to a contemporary PC, so I guess there's not much incentive to 'FastDoom' it.

You missed my point on the 486SX-25 and 486SX-33-based PC's prices against the A1200 with 68030 @ 50Mhz accelerator.

Your A1200 with 68030 @ 50Mhz accelerator argument lost against Gateway 2000's 486SX-33/486SX-33 PC's 1993 prices.

The "power without the price" has departed from A1200 with 68030 @ 50Mhz accelerator's total price.

Originally Posted by Bruce Abbott

I tested DoomAttack on my 50MHz A1200 in Graffiti mode and it was slower. In 'low detail' (2x1 pixels) I got 13.9 fps with blitter-assisted c2p, and 12.2 fps with graffiti.

My point with Graffiti is chunky pixel data organization without investing in extra R&D time with Blitter-assisted C2P. i.e. it's for developers who refuse to add extra R&D time for Blitter-assisted C2P.

[TEG]

Can we draw a parallel with Commodore? Even Intel pay the price for not behind at the edge of the technology:

Intel discloses $7 billion operating loss for chip-making unit

During a presentation for investors, Chief Executive Pat Gelsinger said 2024 would be the year of worst operating losses for the company's chipmaking business and that it expects to break even on an operating basis by about 2027.

Gelsinger said the foundry business was weighed down by bad decisions, including one year ago against using extreme ultraviolet (EUV) machines from Dutch firm ASML. While those machines can cost more than $150 million, they are more cost-effective than earlier chip making tools.

Partially as a result of the missteps, Intel has outsourced about 30% of the total number of wafers to external contract manufacturers such as TSMC, Gelsinger said. It aims to bring that number down to roughly 20%.

Intel has now switched over to using EUV tools, which will cover more and more production needs as older machines are phased out.

"In the post EUV era, we see that we're very competitive now on price, performance (and) back to leadership," Gelsinger said. "And in the pre-EUV era we carried a lot of costs and (were) uncompetitive."

Intel plans to spend $100 billion on building or expanding chip factories in four U.S. states. Its business turnaround plan depends on persuading outside companies to use its manufacturing services.

https://www.reuters.com/technology/i...ss-2024-04-02/

At a point of time, Commodore chipset plant was used to equip a lot of computers.

[Bruce Abbott]

Originally Posted by TCD

Here's John's original answer:

https://groups.google.com/g/comp.sys.../c/MZb9cC0FMhw
He didn't refuse to let it be ported. He knew how it ran on NeXTSTEP and (partially wrongly) assumed it wouldn't be feasible on any Amiga. Team 17 later approached id about the Doom license for their engine that became Alien Breed 3D and it simply wouldn't have been profitable to get the license with the expected sold copies.

I know you'll continue to 'spin' this story so that it looks like there was a conspiracy to not let the mighty Amiga™ get a commercial Doom port, but the simple truth is that nobody saw money in it and thus it didn't happen.

I'm not spinning. I totally understand why Carmack wasn't interested in it. By the time he made that statement we already knew that his chunky pixel argument was bogus. But he was totally right about the 'majority of the amiga base', which consisted mostly of A500s. Chunky pixels wouldn't make any difference when you only have a 7MHz 68000 and 1MB RAM. A stock A1200 wouldn't be any good either, chunky or no.

Although some people were running Doom on a 386SX in a tiny window, it really needed at least a 40MHz 386DX and preferably a 486. The A4000 was the only stock Amiga that was suitable, but there weren't enough of them to justify Carmack's time - even if it did have chunky pixels!

[Bruce Abbott]

Originally Posted by Thorham

Isn't that a 'little' late?

A little late for what?

[Bruce Abbott]

Originally Posted by Turrican_3

With a company so reliant on the gaming market and with a single, overall modestly successful product (consoles sold 10x and were a WAY more healthy ecosystem due to immensely higher average software sales, too) the writing was on the wall unfortunately, and had been for a long, long time.

The writing was on the wall from 1981. Commodore knew this in 1985 when they introduced the Amiga. They went against the flow for almost a decade with an architecture that the industry wasn't interested in. With better management they might have been able to go on for another few years, but eventually it would come to an end just as it did for all the other 'home' computers of that era (not counting the Mac because it was aimed at a different market).

I'm pretty happy with how far they got. AGA was the culmination of their original vision. Next generation Amigas would get more and more like PCs until the only point of difference would be the OS, and even that would be found wanting compared to Windows 95.

So I think it was better for the Amiga line to stop while it still had an identity, in the same way that the ZX Spectrum, Amstrad CPC, C64 etc. did. That meant we could continue to enjoy the machines we loved and get to know them more intimately. Computers have always been a hobby for me. That scene died in the 90's as PCs became appliances whose internals were not interesting, but the Amiga is still around with an ecosystem that's arguably better than it was in 1994. I'm not at all disappointed by that.

[Bruce Abbott]

Originally Posted by hammer

You missed my point on the 486SX-25 and 486SX-33-based PC's prices against the A1200 with 68030 @ 50Mhz accelerator.

You missed my point, which is that the A1200 offered a much lower cost of entry with ability to upgrade as your budget allowed, whereas the 486 had to be purchased as a complete package.

Your A1200 with 68030 @ 50Mhz accelerator argument lost against Gateway 2000's 486SX-33/486SX-33 PC's 1993 prices.

The "power without the price" has departed from A1200 with 68030 @ 50Mhz accelerator's total price.

Wrong and wrong. many A1200 owners never bought an accelerator card because they had no need for one. They saved a bundle over buying a PC that would be thrown in the trash 2 years later. Others did buy an accelerator card because they wanted to do more with the machine they loved. The total cost was irrelevant.

My point with Graffiti is chunky pixel data organization without investing in extra R&D time with Blitter-assisted C2P. i.e. it's for developers who refuse to add extra R&D time for Blitter-assisted C2P.

Developers who refuse to make use of the hardware available to them are a curse. "You have to buy more hardware because I'm too lazy to write a few lines of code". That might fly in the PC world but not in our world. It's not the way of the Amiga or any other home computer.

[CCCP alert]

Originally Posted by hammer

PS1 is about AUD $700 in Australia.

https://pbs.twimg.com/media/EaNDNXHV...pg&name=medium


https://archive.org/details/pc-home-...ge/96/mode/2up
From Silica. June 1994
For £799, Ambra PC with 486SX-25, 240 MB HDD, 4 MB RAM, VLB SVGA GD5424 card, SVGA monitor, 3.5 FDD and 'etc'.

https://archive.org/details/pc-home-...ge/20/mode/2up
For £649, 486SX-33, 4MB RAM, SVGA monitor, 170 MB HDD, 256 KB VGA,

https://archive.org/details/pc-home-...ge/92/mode/2up
For £667, 486DX2-66, 128KB L2 cache, 270 MB HDD, 4 MB RAM, VLB 1 MB VRAM GD5428, KB, mouse, 1 printer, 2 serial

https://archive.org/details/pc-home-...ge/86/mode/2up
PC Home (UK), April 1995.
For £629, 486SX-25, 4 MB RAM, 340 MB HDD, SVGA monitor, 1.44MB FDD, VLB 1 MB SVGA card, KB, mouse.
For £649, 486SX-33, 4 MB RAM, 340 MB HDD, SVGA monitor, 1.44MB FDD, VLB 1 MB SVGA card, KB, mouse.
For £659, 486SX2-50, 4 MB RAM, 340 MB HDD, SVGA monitor, 1.44MB FDD, VLB 1 MB SVGA card, KB, mouse.
Sound Blaster Pro = £58
Sound Blaster Pro 16 = £69

https://archive.org/details/pc-home-...ge/82/mode/2up
PC Home (UK), April 1995.
For £505, 386DX-40, 4 MB RAM, 250 MB HDD, SVGA monitor, 512 KB VGA card, FDD, Keyb, mouse, 2 serial, 1 printer, game ports.

4MB RAM targets Doom.

There's no sugarcoating.

Try again.

You need P100 minimum to play PS1 games like Need for Speed (or Superstardust at 50fps rock solid) not a 386DX40 that couldn't even match Lotus II on an A1000 (min 486-33 for that). I sold PCs for 10 years, I know it all, no PC even at double the price of PS1 was worth owning if you wanted anything like PS1 Ridge Racer 4/Tomb Raider/Colin McRae 1 etc etc. AU pricing has sweet FA to do with anything, PS1 was peanuts in EU/JP/US territories and it quadrupled the console market size all alone in total units sold.Try again yourself.

[hammer]

Originally Posted by Bruce Abbott

You missed my point, which is that the A1200 offered a much lower cost of entry with ability to upgrade as your budget allowed, whereas the 486 had to be purchased as a complete package.

Continue with Commodore's failed approach then!

What you missed is that the 486SX's slightly higher price delivered the "new 32-bit 2.5D and 3D gaming" experiences.

What you missed is economies of scale with setting a higher spec AGA machine. David Pleasance wanted a higher-specs CD32 as a baseline and to deliver a gaming experience that SNES wouldn't be able to deliver until SuperFX2.

Higher priced 3DO has a 2 million install base which is far higher than AGA's 250,000 install base.

To increase the AGA install base, at least A1200/CD32 needs to be purchased along with a higher-performance 32-bit CPU accelerator card.

From Q4 1992 to Q4 1994 for stock A1200 gaming experience, it's mostly 16-bit A500 games that competed against SNES's strong 16-bit 2D games.

SNES CPU (with 128 KB SRAM) and GPU (with 64 KB VRAM) have discrete memory buses. SNES has tiled 2D graphics architecture and Mode 7 has chunky pixel support.

Originally Posted by Bruce Abbott

Wrong and wrong. many A1200 owners never bought an accelerator card because they had no need for one.

Continue with Commodore's failed approach then!

Originally Posted by Bruce Abbott

They saved a bundle over buying a PC that would be thrown in the trash 2 years later.

My 386DX-33 has a 486DLC clones upgrade path and overclock options e.g. 40 Mhz.

The 386DX baby AT motherboard was upgraded to a Socket 7 Pentium baby AT motherboard.

I could have carried over the ET4000AX ISA card, but the no-name OEM S3 Trio 64UV+ PCI card is cheap.

Keeping ET4000AX and Sound Blaster Pro sound ISA cards is like keeping a CPU-less CD32 AGA SBC (small board computer).

The existing ISA sound card was carried over into the Pentium era.

PGA 486's 1990s-era motherboard had a Pentium Overdrive Upgrade path. This upgrade path has the same 32-bit FSB bottleneck as 68060's 32-bit front-side bus.

Originally Posted by Bruce Abbott

Others did buy an accelerator card because they wanted to do more with the machine they loved. The total cost was irrelevant.

Continue with Commodore's failed approach then! Cost is relevant for the "power without the price" entry point.

What you missed is economies of scale with setting a higher spec AGA machine. David Pleasance wanted a higher-specs CD32 as a baseline and to deliver a gaming experience that SNES wouldn't be able to deliver until SuperFX.

Hint: SNES CPU and stock A1200 CPU (7 Mhz effective 68EC020) have similar MIPS.

Originally Posted by Bruce Abbott

Developers who refuse to make use of the hardware available to them are a curse. "You have to buy more hardware because I'm too lazy to write a few lines of code". That might fly in the PC world but not in our world. It's not the way of the Amiga or any other home computer.

Continue with Commodore's failed approach then! Stay small in a few 250,000 AGA units while 3DO has a 2 million install base.

IBM's VGA standard has out-of-the-box chunky pixels and 3rd party SVGA clones made it faster.

[hammer]

Originally Posted by CCCP alert

You need P100 minimum to play PS1 games like Need for Speed (or Superstardust at 50fps rock solid) not a 386DX40 that couldn't even match Lotus II on an A1000 (min 486-33 for that).

50 hz is a PAL refresh rate.

Lotus 2 on A500 runs at 25 fps which is half of 50 hz PAL. Try again yourself.

[ Show youtube player ]
Lotus 3 running on 386DX-33 with ET4000. Frame buffer's performance also matters.

VGA has at least a 60 hz refresh rate like NTSC, hence 30 fps is a valid frame rate target.

Originally Posted by CCCP alert

I sold PCs for 10 years,

And? I also sold PCs after my university.

Originally Posted by CCCP alert

I know it all, no PC even at double the price of PS1 was worth owning if you wanted anything like PS1 Ridge Racer 4/Tomb Raider/Colin McRae 1 etc etc.

For Western markets, PS1 was released Q4 1995 and it's well into the Pentium era.

You inserted "Oddly stuff like Final Writer 95 is much more usable on a £600 worth of A1200 than a £600 PC Office 95" argument.

The same 1995 era £600 worth of A1200 wouldn't run PS1 Ridge Racer 4/Tomb Raider/Colin McRae 1 etc etc.

Originally Posted by CCCP alert

AU pricing has sweet FA to do with anything, PS1 was peanuts in EU/JP/US territories

Euro currency didn't exist until the 1st of January 1999.

[hammer]

Originally Posted by TEG

Can we draw a parallel with Commodore? Even Intel pay the price for not behind at the edge of the technology:

At a point of time, Commodore chipset plant was used to equip a lot of computers.

https://www.theverge.com/2024/4/2/24...7-billion-loss

It also will receive up to $8.5 billion in funding from the U.S. government, as part of the new CHIPS Act

[hammer]

Originally Posted by Bruce Abbott

I'm not spinning. I totally understand why Carmack wasn't interested in it. By the time he made that statement we already knew that his chunky pixel argument was bogus. But he was totally right about the 'majority of the amiga base', which consisted mostly of A500s. Chunky pixels wouldn't make any difference when you only have a 7MHz 68000 and 1MB RAM. A stock A1200 wouldn't be any good either, chunky or no.

Although some people were running Doom on a 386SX in a tiny window, it really needed at least a 40MHz 386DX and preferably a 486. The A4000 was the only stock Amiga that was suitable, but there weren't enough of them to justify Carmack's time - even if it did have chunky pixels!

From 1992, "big box" Amigas had Enhanced Graphics System (EGS) Zorro II/III SVGA graphics cards, but they were expensive due to the tiny big box Amiga install base. Again, it's economies of scale issue.

A3000T/040 with EGS or A4000/040 with EGS could do it.

Commodore didn't exploit A500's single Zorro I edge connector and create an Amiga model with a single Zorro II slot with an asking price between A1500/A2000 and A500.

Commodore didn't exploit A1200's internal Zorro II/III-like 32-bit edge connector and create an Amiga model with a single III slot with an asking price between A4000/030 and A1200.

There was a potential for a modular gaming Amiga model. Unlike Apple's USD $999 Quadra 605 (with 68LC040), the Amiga didn't move towards the native 040 bus design.

Commodore purposely avoided mid-price segments for the Amiga i.e. refer to 1993 era Commodore 486DX and 486SX PCs.

Modern low-end gaming PC with AM4's A520/B550 and AM5's A620/650 has a single PEG (PCIe Graphics) slot.

[sokolovic]

Originally Posted by hammer

Continue with Commodore's failed approach then! Stay small in a few 250,000 AGA units.

Why dont you even lower your AGA units sales since you're totally inventing the number ?

[hammer]

Originally Posted by sandruzzo

with raspberry Even copper and blitter can be pushed further(more than 68060)?

With PiStorm32 with Emu68, A1200's 7.1 MB/s CPU to Chip RAM remains, but it's enough for Star Wars Dark Forces to run +50 fps 320x200 with 256 colors.

A1200 has no problems with playing pre-baked animation at 320x200 +50 fps.

[hammer]

Originally Posted by sokolovic

Why dont you even lower your AGA units sales since you're totally inventing the number ?

I was generous.

Disprove http://www.bambi-amiga.co.uk/amigahistory/sales.html 's AGA install base i.e. 180,300 AGA + more than 10,000 Vampires.


UK's CD32 numbers are missing.

Escom's AGA numbers are missing.

[hammer]

Originally Posted by TCD

Here's John's original answer:

https://groups.google.com/g/comp.sys.../c/MZb9cC0FMhw
He didn't refuse to let it be ported. He knew how it ran on NeXTSTEP and (partially wrongly) assumed it wouldn't be feasible on any Amiga. Team 17 later approached id about the Doom license for their engine that became Alien Breed 3D and it simply wouldn't have been profitable to get the license with the expected sold copies.

I know you'll continue to 'spin' this story so that it looks like there was a conspiracy to not let the mighty Amiga™ get a commercial Doom port, but the simple truth is that nobody saw money in it and thus it didn't happen.

John Carmack exaggerates when a lower-cost 68LC040 would have done the job i.e. Doom is a fixed point integer only.

The "full 68040" is not required for Doom.

[Bruce Abbott]

Originally Posted by hammer

John Carmack exaggerates when a lower-cost 68LC040 would have done the job i.e. Doom is a fixed point integer only.

The "full 68040" is not required for Doom.

John Carmack's only 68k experience was his 68040 equipped NextStation, which wasn't built to be a gaming machine. But none of this really matters. The important thing is that a stock A500 or A1200 wasn't going to cut it anyway, and even if it did there weren't enough of them out there (especially with all the piracy going on).

But in the end it didn't matter. In 1997 they released the source code and we had Doom on the Amiga. Since then other games have been ported too, which Amiga owners can now enjoy. I never played those games back in the day, so they are a new experience for me!

[sokolovic]

Originally Posted by hammer

I was generous.

Disprove http://www.bambi-amiga.co.uk/amigahistory/sales.html 's AGA install base i.e. 180,300 AGA + more than 10,000 Vampires.


UK's CD32 numbers are missing.

Escom's AGA numbers are missing.

No you're doing cherry picking and adding apple to oranges (Germany total sales plus A1200 worldwide sales on a short period. The numbers on this blog aren't representative as stated here. There a discussion about these figures here. "Worldwide" numbers presented in your blog are UK ones for instance.

There is much more data that have been extrated from the press since this blog post . Here you have a more correct and argumented estimation for the A1200 alone.

What we know for sure is that the problem for the 1200 was not selling them but providing the demand.

Even in the chat with 2 Commodore employees posted before on this thread IIRC it also was mentionned that the A4000 was selling above expectations. Sales were'nt a problem for the AGA range.

[Bruce Abbott]

Originally Posted by hammer

50 hz is a PAL refresh rate.

Actually PAL is 25 Hz, and standard cinema film is 24 fps. Doom is very smooth at 25 fps.

VGA has at least a 60 hz refresh rate like NTSC, hence 30 fps is a valid frame rate target.

VGA doesn't have frame sync, so you get tearing which looks horrible in any game with sideways scrolling.