Everything defines Amiga. Hardware, software, services like Aminet and the people. Just computers have moved on and Amiga market have shrinked so small that it can't keep up. But then again, thanks to Warp 3D Nova and ShaderJoy we can do fun things with GPU's. And there's some nice piece of software like SketchBlock, for example.

PS. And remember every computer has a chipset. (Except first IBM PC's had everything on addon cards, I think.)

Last edited by Tomppeli on 11-Nov-2021 at 12:21 AM.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

According to at least one boisterous participant, what defines an Amiga is the lack of native chunky pixels/gfx support.

As the decade rolled over from the '80s to the '90s, no one denies that Commodore was making a lot of mistakes. Many things were planned and scheduled to come out earlier, many corners were cut, technology people were not listened to, and overall there was a growing lack of understanding of the very industry they were a part of.

Even if there was an A1200 2MB/4MB EC030/25MHz in 1993 with AA+ chipset for chunky pixels at $999 USD, and an A4000 2MB/8MB 040/25MHz with a video card that supported SVGA-level of chunky pixels at $3,999 USD, I doubt it would've saved Commodore.

At best it would've delayed the inevitable demise. Commodore were just not prepared for the PC hardware boom that came in the mid '90s. Apple was another company that was not prepared, but was luckily bailed out by Microsoft.

And today, somebody would still be slagging AA+ chipset and calling it crap because it didn't have 3D pixels, Gouraud shading, and z-buffer. Since the PlayStation console and the 3Dfx Voodoo card for PCs had those things, despite being released in 1995 and 1996, respectively.

_________________
All the way, with 68k

BigD Quote:

A 68030 at 50Mhz AGA Amiga plays ADoom with Doom 1 & 2 WADs just fine with minor slowdown. The Final Doom WADs are more pleasant with an 060 & RTG but a Walker spec machine but with a full 030 50Mhz in early 1994 would have been ok as a Doom machine. Obviously C= would have had to have paid id Software to port it like they did with Origin and Wing Commander but it would have secured medium term survival and I think A500 owners WOULD have upgraded if that option had existed! Obviously, the US Amiga owners gave up far earlier

I was trying to say that AGA was not a major bottleneck for Doom and other chunky gfx games when a processor with adequate performance to move chunky data around was present which is not true of many low end VGA cards in 1992.

Some 68030@50MHz Amiga accelerators came later as they were favored by some Amiga users who wanted maximum early Amiga compatibility at the expense of performance. Some have better and more expensive memory than was common or available in 1992. The 68030@50MHz was expensive and of limited quantity and therefor unlikely to be used in a large line of affordable mass produced products. The Apple Macintosh IIfx introduced in 1990 used a 68030@40MHz and dual ported 80ns 64 pin SIMMs but was introduced at $8,969 for the base system.

https://en.wikipedia.org/wiki/Macintosh_IIfx

A while after the 68040 came out in 1990, A 68040@25MHz was a more available and affordable CPU alternative with generally better performance than a high clocked 68030 except for CBM accelerators which were made as cheap as possible. Part of the problem was CBM was focused on providing the cheapest hardware possible instead of the best value. CBM missed out on high profit margin high end hardware which was sold by 3rd parties and was expensive and customers missed out on more affordable mass produced high end hardware. Customers in the U.S. were willing to pay more for high end Amiga hardware but the value was often not there. European customers often payed for accelerators but they were generally low to mid clocked 68030s with fast memory upgrades for the Amiga 1200 which sold in large enough quantities for mass production by 3rd parties and offered better value. The lack of value was in the high end of the Amiga market and that was clearly evident when the high end PC clone gaming market took off and left Amiga in the dust. The Amiga affordable efficient general purpose computing and 2D gaming markets were still healthy in Europe when CBM went bankrupt although PC clone hardware was offering increased value through more performance and lower prices where the Amiga improved value primarily by dropping prices and offering bundles.

agami Quote:

According to at least one boisterous participant, what defines an Amiga is the lack of native chunky pixels/gfx support. As the decade rolled over from the '80s to the '90s, no one denies that Commodore was making a lot of mistakes. Many things were planned and scheduled to come out earlier, many corners were cut, technology people were not listened to, and overall there was a growing lack of understanding of the very industry they were a part of.

Some people would say mistakes were made by CBM before the late '80s. Jack Tramiel had created a bad reputation for CBM with distributors and suppliers. The takeover of CBM by Irving Gould and exit of Jack Tramiel resulted in the Atari ST which weakened the Amiga introduction. There was the "We made it they f***ed it up" rift between Los Gatos and CBM. The upper management lost time during transitions of CEOs and cronyism as Gould forced out good people and wasted valuable cash. For such a friendly name, the Amiga history has been marked by division and conflict instead of cooperation.

agami Quote:

Even if there was an A1200 2MB/4MB EC030/25MHz in 1993 with AA+ chipset for chunky pixels at $999 USD, and an A4000 2MB/8MB 040/25MHz with a video card that supported SVGA-level of chunky pixels at $3,999 USD, I doubt it would've saved Commodore.

The SVGA/VBE standards were nothing special providing standard 800x600 up to 1280x1024 in 4 or 8 bit depths. AGA could already support high resolutions like 800x600 and 1280x400 in bit plane depths only limited by gfx mem bandwidth and the 2MiB chip memory size though refresh rates were poor or interlaced was used. AA+ doubled the gfx mem bandwidth over AA (8 times over ECS) and increases chip memory to 8MiB allowing for an eye friendly non-interlaced 800x600@72 Hz with an 8 bit depth similar to SVGA. The later Amiga chipsets are flexible allowing for a programmable pixel clock and customizable resolutions. Amiga chipset upgrades (ECS to AA+) were also more compatible than PC clone gfx upgrades.

The early advantage of expensive dual ported VRAM to increase bandwidth disappeared as mem bandwidths increased over time with cheaper memory. Around 2001, DDR memory doubled bandwidth by transferring data on both the rise and fall of the clock. SAGA today supports 720p and even 1080p resolutions using a similar programmable Amiga like chipset with many times more gfx mem bandwidth while using older affordable hardware in similar CBM Amiga philosophy. SAGA does not have the bandwidth to support the highest resolution TVs and monitors using 32 bit chunky modes currently but it is just a matter of lack of mem bandwidth using cheap lower end hardware the same as in CBM days.

In any case, the Amiga chipset could not have saved the Amiga alone. At best, it avoided major bottlenecks and limitations which even AGA did in 1992 and I expect AA+ would have in 1993. There was more to be gained by the CPU providing a better performance/price ratio and this was a problem which CBM needed to address. It didn't help that CBM (and Atari) had always bought the cheapest processors and made the cheapest accelerators which became a problem for Motorola upgrading the 68k as well.

agami Quote:

At best it would've delayed the inevitable demise. Commodore were just not prepared for the PC hardware boom that came in the mid '90s. Apple was another company that was not prepared, but was luckily bailed out by Microsoft.

Add Motorola to the list of companies that did not transition the Moore's Law tech acceleration well. They were slow to realize they were in competition with Intel and AMD and needed to accelerate 68k development and lower prices more aggressively to keep the 68k from being left behind. They then gave up the bread and butter 68k market, and with it the fast growing embedded market, to put all their apples in one rotting basket and take the lazy route with PPC and IBM doing most of the ISA and design work. The once mighty Motorola became a Chinese cell phone company and the spinoff chip company Freescale nearly went bankrupt and is now owned by the Dutch company NXP.

agami Quote:

And today, somebody would still be slagging AA+ chipset and calling it crap because it didn't have 3D pixels, Gouraud shading, and z-buffer. Since the PlayStation console and the 3Dfx Voodoo card for PCs had those things, despite being released in 1995 and 1996, respectively.

Hombre supported 3D and could have been added to AA+ for the next Amiga chipset after AA+ if CBM had survived. It would have been more efficient if a 68k CPU with SIMD support had been added instead of the memory wasting PA-RISC CPU with weak SIMD support. In an alternate history with better management, Motorola and CBM could have worked together to make the next generation Amiga. Today, we could follow the path that could have led to with FPGA Amiga development although the full potential would never be realized without an ASIC.

Last edited by matthey on 11-Nov-2021 at 04:08 PM.
Last edited by matthey on 11-Nov-2021 at 09:18 AM.
Last edited by matthey on 11-Nov-2021 at 09:11 AM.
Last edited by matthey on 11-Nov-2021 at 09:06 AM.
Last edited by matthey on 11-Nov-2021 at 09:05 AM.

@agami

The thing that made lack of chunky mode graphics such a screw-up for Commodore is that IBM had that capability since VGA mode 0x13 in April, 1987. That gave Commodore five years before Wolf-3D came out to get it done, but they didn't. After Wolf-3D—and especially after Doom—it was too late.

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

bison Quote:

The thing that made lack of chunky mode graphics such a screw-up for Commodore is that IBM had that capability since VGA mode 0x13 in April, 1987. That gave Commodore five years before Wolf-3D came out to get it done, but they didn't. After Wolf-3D—and especially after Doom—it was too late.

VGA was introduced in the IBM PS/2 in 1987.

PS/2 Model 25 286 to 386 CPU Upgrade. runs DOOM!
https://youtu.be/k4Tm2Mrh5g0?t=186

The model 25 originally came with a 8086 and this PS/2 model 25-286 originally had a 80286 so would have been released in 1990. IBM released a PS/2 model 25 with a 80386SX in 1992 so this CPU upgrade is about what you had in 1992 except the video shows a 80386SX@33MHz where the original from IBM had at most a 80386SX@20MHz. This hardware is no where near playable for Doom and is described in the video as a slide show. The IBM VGA card at the time was also a bottleneck. The following quote is from the video.

Quote:

I suspected that the PS/2 model 25s on board VGA graphics system mite be holding the processor upgrade back so I tested my theory and it proved to be true. Installing an external graphics card yielded me a huge increase in performance. Now keep this in mind you model 30 owners who are probably going to use an external graphics card anyway. I got closer to a 200% increase in performance with both benchmark utilities when I used my external STB Nitro ISA video card.

Unlike VGA, the Amiga chipset provided hardware assistance when dealing with data. With similar CPU performance to a 80286, the 68000 CPU with Amiga chipset blitter, copper and sprites could do much more in the '80s.

Shadow of the Beast Longplay (Amiga)
https://youtu.be/y9BltSvKMlQ?t=217

That IBM PS/2 model 25-286 that can't play Doom with it's chunky VGA had a list price of $2,215 in 1990. The Amiga 500 was introduced at $699 in 1987.

IBM PS/2 (Model 25-286) - Technical Specifications
https://www.ardent-tool.com/qtechinfo/8525-286.html

Wolfenstein was the highlight of what '80s and early '90s IBM VGA could do as far as gaming. It was more powerful x86 processors starting in about 1993, partly because of Doom which came out late that year, that finally allowed IBM compatible hardware to move data around with brute force fast enough to play descent fps games. Even then, the hardware was very expensive at first until economies of scale kicked in as more and more people bought the same hardware.

Last edited by matthey on 11-Nov-2021 at 08:09 PM.
Last edited by matthey on 11-Nov-2021 at 08:06 PM.
Last edited by matthey on 11-Nov-2021 at 08:04 PM.
Last edited by matthey on 11-Nov-2021 at 08:03 PM.

To everyone who cites how this or that tech was introduced in some year prior to AGA:
There is a big difference between technology being available and technology has large adoption.

When it came to game development, in many ways the Amiga was like a game console. Developers could rely on a known quantity of the install base. The move away from sprites had not began in earnest prior to 1992. But when it did happened, it happened quickly. Far too quickly for a struggling Commodore to keep up.

_________________
All the way, with 68k

@matthey

c2p is not for free.
c2p times for 030 are easily avaible so dont lie that
"AGA was not a major bottleneck for Doom and other chunky gfx games"
c2p on 030 50 MHz take almost half of cpu time in 320x200 8 bit 25 fps.

@agami

As a software developer I don't care about Amiga chipset.
It was slower than cpu even in Commodore times.
AGA was many times slower because it has not chunky pixels.
But not only AGA.
Even my Amiga 500+ connected to VGA monitor works like NG.
No custom copper list, cpu only because blitter is too slow.
This Amiga 500+ is 30 years old hardware from Commodore.

agami Quote:

To everyone who cites how this or that tech was introduced in some year prior to AGA: There is a big difference between technology being available and technology has large adoption.

Yes, that was the point I was just trying to make. There was x86 hardware that was more powerful than the Amiga but most x86 personal computer hardware in the home wasn't overall better than the Amiga until the mid '90s. I showed that even the original genuine from IBM PS/2 VGA hardware which first came with a pseudo 8 bit chunky mode didn't have the performance for Doom because of lack of CPU performance and VGA gfx performance. High performance hardware starting around 1990 could play Doom but was very expensive. A 80486@33MHz in 1990 cost $953 which is more than than a low end Amiga computer cost. Because x86 hardware won the Clone Wars, there is this idea that x86 hardware was superior back in the late '80s and early '90s when it was not on average. There was very expensive x86 hardware for businesses and universities but it was rarely used for gaming until the Doom gaming boom. The performance/price of x86 hardware was generally poor until the gaming boom and brute force using the CPU was the only way to move gfx data around unlike the Amiga. Amiga did not really have high end hardware but the low end hardware was generally better than most x86 hardware up to and including the Amiga 1200 in 1992.

Doom was a clever game in many ways. Before Doom, game producers generally created games for the lowest common denominator of hardware which was low and slow to progress for PC clones. Doom was scalable in performance by shrinking the screen size so customers could see the difference in hardware as a sort of visible benchmark and wanted more performance. This is when large amounts of gold could be found in the hills of the dumps where old, and even not so old, x86 hardware ended up.

agami Quote:

When it came to game development, in many ways the Amiga was like a game console. Developers could rely on a known quantity of the install base. The move away from sprites had not began in earnest prior to 1992. But when it did happened, it happened quickly. Far too quickly for a struggling Commodore to keep up.

It is surprising the Amiga didn't try to introduce an ECS based Amiga game console after the Amiga PC was introduced (only the CDTV device but dropped even standard CD support after it failed). NES turned around the console market and the next generation SNES and Sega Genesis were big successes also. The Sega Genesis used a 68000 and was similar to the Amiga. The Genesis which was introduced in 1988 (1989 for U.S. and 1990 for PAL version) had better sprite hardware but was otherwise inferior. The Amiga chipset hardware could have added more console like features but even AGA only added minor improvements in this direction and was late at that. It was easier for a computer like the Amiga to drop into the console market than for a console to move up into the computer market as an OS is required. Selling into both the computer and console markets would have increased volumes for mass production and made CBM more defensive by being in two different cyclical markets. CBM should have been developing and marketing products for the embedded market as well which is a defensive (non-cyclical) market that was starting to boom. CBM could have added higher performance processors along with the chunky AA+ they were planning and had both console compatibility and higher performance CPU brute force chunky which maybe could have been adequate if a little behind the performance curve provided prices were lower. The 2D console sales actually tapered off gradually. The SNES with just a 16 bit 6502 CPU wasn't discontinued until 2003 with nearly 50 million units sold and the Sega Genesis was discontinued in 1999 with 32 million units sold. CBM could have added 3D support with Hombre to the Amiga chipset if they wanted to compete with the newer consoles. CBM was well positioned with the Amiga in 1985 to become a tech power house but they botched it. CBM died a one trick pony failing to diversify into other tricks and failing to improve their one trick enough to remain competitive. Only at the end did they almost succeed with a new trick when they introduced the CD32 console but it was too late to market with upgraded technology and for CBM's finances.

ppcamiga1 Quote:

c2p is not for free. c2p times for 030 are easily available so dont lie that "AGA was not a major bottleneck for Doom and other chunky gfx games" c2p on 030 50 MHz take almost half of cpu time in 320x200 8 bit 25 fps.

Actually, c2p is free on the 68040 and 68060 with fast memory as the conversion is done at copy speed. In other words, the CPU can do the c2p conversion while waiting on chip memory writes. Yes, chip memory back then was slow and the chip memory bandwidth was low as the Amiga chipset was made for the masses but it was still better than many VGA implementations, including the original IBM PS/2 VGA, as I showed above.

coder76 Quote:

Both chunky and planar modes have their own advantages. But there exists chunky to planar routines (c2p's), which converts a chunky screen held in fast ram to chip ram planar format. On a 68040/68060 its copyspeed, so this conversion does not slow down copy speed to chip ram. With Doom, you really see that planar format does not slow down faster Amigas. A 68030/50 MHz even outperforms a 386/40 MHz running Doom in benchmarks by a frame or two, despite doing the c2p. Also, a 68040 outperforms a 486 PC running Doom at same clockspeed. The chip ram access speed is more of a bottleneck on the Amigas, rather than the planar format.

http://eab.abime.net/showpost.php?s=5875a681b8c4cd80838d34b4f39428fc&p=1192120&postcount=5

Last edited by matthey on 13-Nov-2021 at 06:49 AM.
Last edited by matthey on 13-Nov-2021 at 06:27 AM.

@matthey

Quote:

Yes, that was the point I was just trying to make. There was x86 hardware that was more powerful than the Amiga but most x86 personal computer hardware in the home wasn't overall better than the Amiga until the mid '90s...

I know you're making that point. I made the same point in post #33, and others have also made the same point. We're on the same page you and I.

I was more talking to Mr. Revisionist History (@ppcamiga1) with that last one. He keeps bringing up crummy reasons why AGA was crap in 1992 because it lacked technologies which were 2x and 3x the cost, and were not widespread in the user base.

Commodore had many faults, but I challenge anyone to put together and market a profitable personal computer in 1992 with a sub $700 USD target price, that is compatible with the OCS/ECS Amiga's that came before, and is materially better than the A1200 we got. Keeping in mind that hardware is usually priced at cost of components x 3.

Quote:

It is surprising the Amiga didn't try to introduce an ECS based Amiga game console after the Amiga PC was introduced (only the CDTV device but dropped even standard CD support after it failed)...

I agree. The CDTV was such a misstep. The interactive CD appliance was so out of character for Commodore. If they wanted to put an Amiga-based device under the living room TV, it should've been a cartridge based console made for the masses. Then the CD32 would've been a welcome upgrade to this secondary market. Like you said, with the planned Hombre chipset, it could've even been a competitor to Sony's PlayStation.
If only.

_________________
All the way, with 68k

@matthey

Pc run version at 640x480, but Amiga ports runs at 640x200 (or 320x200), or something like that. 1/2 screen = 1/2 cpu needed.
(you almost never see dbpal or dbntsc being used in games, ufo enemy unknown being an exception.), at least AmiHeXen does (640x200).
the Amiga port assembler optimized so it run on 68030, the PC version is only written in C code.
it was one advantages DOOM had it easy to port many different platforms, it even runs some calculators.
I know CHIP ram is slow to write, the CPU has to wait many cycles before it given access, with faster CPU that problem increases. the c2p has do lot of work, I have hard time it image can do all that work in that time. It has to change the layout of all bits in pixels after all.

what I'm saying is Amiga port, saves times but not doing 100% of the work, it runs hand optimized code in asm, you can't compare just fps, the same work is not being done.

a static test only test pixel plotting, memory copy, rectangles, and basic draw test will give better indication of how slow the AGA chipset is, compared to VGA.

Last edited by NutsAboutAmiga on 13-Nov-2021 at 10:52 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 09:37 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 08:45 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 08:44 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 08:43 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 08:42 AM.
Last edited by NutsAboutAmiga on 13-Nov-2021 at 08:40 AM.

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

agami Quote:

Commodore had many faults, but I challenge anyone to put together and market a profitable personal computer in 1992 with a sub $700 USD target price, that is compatible with the OCS/ECS Amiga's that came before, and is materially better than the A1200 we got. Keeping in mind that hardware is usually priced at cost of components x 3.

CBM was generally good at cost reducing products but poor at upgrading and marketing products. From the acquisition of the Amiga, they were cost reducing the hardware and waiting for it to come down in price enough to turn it into another C64 rather than upgrading it to stay ahead. Sadly, it wouldn't have cost much to upgrade the chipset more while they integrated and cost reduced it. I believe AGA could have been out in 1990 and AA+ in 1993. This may sound aggressive by looking at x86 and Mac hardware but look at the x68000 hardware which came out in 1987.

https://en.wikipedia.org/wiki/X68000

Amiga console hardware needed to be cheaper than that but the Sega Genesis showed what could be done on the low end using a similar amount of hardware to the Amiga.

agami Quote:

I agree. The CDTV was such a misstep. The interactive CD appliance was so out of character for Commodore. If they wanted to put an Amiga-based device under the living room TV, it should've been a cartridge based console made for the masses. Then the CD32 would've been a welcome upgrade to this secondary market. Like you said, with the planned Hombre chipset, it could've even been a competitor to Sony's PlayStation. If only.

Carl Sassenrath, who was brought back to help develop the CDTV, tells that the CDTV was not only supported by, but being prioritized by, the highest levels of management. They were still out of step with Amiga owners most of which didn't buy this Amiga based hardware. It wasn't labeled an Amiga, it was expensive and a keyboard and mouse cost more, it didn't come with an upgraded CPU, chipset or ROM (still used AmigaOS 1.3 in the 1990 release) and couldn't easily be upgraded with more memory or a hard drive. The biggest advantage was the CD-ROM which Amiga owners expected CBM would make a standard for the Amiga and bring to existing Amigas (only happened in a limited and inconsistent way). The concept of pairing cheap CD-ROM disc capacity with the multimedia features of the Amiga was a good one but they attempted to create a whole new market that was too high end and not gaming oriented enough for children. CBM eventually realized this when they dropped the CDTV2-CR and went in the right direction with the CD32 just too late.

CBM released a low priced cartridge based game console in Europe in 1990 before the CDTV. It was the 8 bit C64GS (Game System) which was a failure.

https://en.wikipedia.org/wiki/Commodore_64_Games_System

The price was similar to the NES and Sega Master System but they were high quality products with high quality games in an entrenched 8 bit market. The next generation 16 bit Sega Genesis had already arrived while considerably more expensive. The C64GS was low quality, not expandable even with a keyboard and too late. CBM was wasting time chasing the console market with the C64GS instead of leading it with the next generation Amiga. The big question is whether an Amiga game console in the late '80s or very early '90s should have been cartridge based. At some point, the price of CD-ROM drives was cheap enough that the cheap price of CD-ROM discs, huge disc capacity and resistance to piracy outweighed the higher cost of the CD-ROM drive itself. Certainly by the time of the CD32 in 1993, the cheap audio CD mechanism found by Jeff Porter as used in Sony CD Walkman mass production was compelling. The Sony Discman, which was later renamed as CD Walkman, was released in 1984 and audio CD sales surpassed audio cassette sales in 1991 so I believe this cost reduction could have been done earlier. The Sega (Mega/Genesis) CD was launched in 1992-1993 for most of the world so the CD32 in 1991-1992 at less than half the price of the CDTV should have been very competitive. If CBM had been able to get their foot in the door of the console market at that time, then yes, adding 3D support to the Amiga AA+ chipset with Hombre while retaining good Amiga backward compatibility all the way to OCS would have been interesting, more interesting than trying to start over with a new console based on the Hombre chipset alone. Eventually, the main 68k CPU would have needed to be upgraded and SIMD unit support added to provide maximum compatibility which CBM was looking at licensing. If they had had console success early enough and upgraded their CPUs for both PCs and consoles, Motorola may have even stuck with the 68k but CBM would have wanted to integrate the 68k into a one chip Amiga SoC eventually. That is where consoles eventually ended up using powerful CISC CPUs just with the less power efficient x86-64 instead.

Last edited by matthey on 13-Nov-2021 at 09:44 AM.
Last edited by matthey on 13-Nov-2021 at 09:31 AM.

@ppcamiga1

Quote:

AGA was many times slower because it has not chunky pixels.

So why didn’t you do like you’d do with a PC or Mac, and get an Amiga with a graphics card capable of chunky graphics?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

So why didn’t you do like you’d do with a PC or Mac, and get an Amiga with a graphics card capable of chunky graphics?

It is simple. Price.
Amiga 4000 with graphics card in 1994 cost few times more
(after adjusted inflation) than Amiga X5000 today.
It was way too much.
In semptember 1996 I switch to pc.
Upgrading Amiga 1200 to 030 cost as much as change to Pentium 90 with S3Virge.
So I sell Amiga and buy pc.
S3 Virge even if not so fast in pc world was two hundreds times faster
than orginal Amiga blitter almost not changed since 1983.

@agami

Maybee in Africa and Australia people use ega and hercules at end of 1992.
Rest of world use SVGA with at least 8 bit chunky pixels.

@matthey

In Amiga community asylum many people compare original DOOM writen in C
with Amiga version with hand optimised assembly.
Suprise no one made working DOOM on 030 back in the half of 90.
This is best proof that lack of chunky pixel in AGA was what killed Commdore.

I don't care how fast c2p is on 040 or 060.

Back in the half 90 040 was too expensive.

060 was even worse.
060 was and is ridiculously expensive.
When I try to return to Amiga at the end of 90 it cost more than ppc plus bvision to Amiga 1200.
This was pure BS.

@ppcamiga1

So why didn’t you switch to PC or Mac in 1992, and instead waited more than two years, till way after death of CBM, before you got a "cheap" PC? That’s full four years of self loathing?

So what did I do? I got an A3000 with CV64 for money I earned from making animations with that substandard AGA chipset that you detest - after first spending half of it going on interrail for a month.

Last edited by kolla on 13-Nov-2021 at 06:09 PM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

GTFO with that question. We all suffered because of this obsession of ours but there's a limit and everyone reaches his or her own. He used it for as long as it was hospitable to his needs and way of living and so did I. Like I said before, you guys act like your upgrading method was some smart way of keeping an Amiga. So far PPC has said nothing but truth as far as this thing goes - those upgrades for Amiga were very expensive, when getting an A4000 made more sense long term and I used it even after getting a pc until I burnt the modem chip and I didn't care enough at that point to keep going. I didn't understand why upgrading a computer which doesn't help me with my enveloping needs, was a good thing. I rather invest in one swing and try to use it as long as I can rather than buying cheep and paying expensive later. What he's saying that in 1992, the cheap Amiga model was no longer at that favourite spot of being cheap enough to get compared to more expensive alternatives and being ahead of the curve compared to your run of the mil pc or a non-attractive counterparts. It was about using what you got and hoping for the best as a good way which was as good as getting a A1200 and spending more on it than I spent on the A4000, and without any real advantage other than a hobby, which is exactly like it is at the current time.

Last edited by Srtest on 13-Nov-2021 at 06:37 PM.

ppcamiga1 Quote:

Maybee in Africa and Australia people use ega and hercules at end of 1992. Rest of world use SVGA with at least 8 bit chunky pixels.

You conveniently left out VGA which was likely the most common IBM compatible gfx hardware in 1992.

ppcamiga1 Quote:

In Amiga community asylum many people compare original DOOM written in C with Amiga version with hand optimised assembly.

The original Doom code used assembler optimizations and hardware tricks for the x86.

ppcamiga1 Quote:

Surprise no one made working DOOM on 030 back in the half of 90. This is best proof that lack of chunky pixel in AGA was what killed Commodore. I don't care how fast c2p is on 040 or 060. Back in the half 90 040 was too expensive.

Doom was first developed on NeXT hardware.

https://fabiensanglard.net/doomIphone/doomClassicRenderer.php Quote:

Since Doom was developed on a NeXTSTEP system using a flat virtual memory model, id Software decided against using EMS or XMS like most games of the time. Instead they used DOS/4G: a memory extender that allowed a software to access protected mode RAM on a real mode operating system (DOS). The NexT workstation was so powerful that it was able to run the editor, the game and a debugger. When the game was stable enough the code was sent of the network to a PC where it was compiled for DOS/x86 by a Watcom compiler. Thanks to DOS/4G the code ran with the same memory model on PC and NeXT.

What was this amazing NeXT hardware using? Either a 68030 if 1st gen or 68040 if 2nd gen NeXT hardware which I believe is more likely. Yes, Doom ran on the 68k first while multitasking with the development tools. Non-multitasking DOS based x86 hardware couldn't do that and used another hack to get a flatter memory model like the 68k used. The NeXT hardware had fancier graphics than the Amiga but they used dual ported VRAM which was very expensive. The original NeXT with 68030 in 1987 had a base price of $6,500 and the 2nd gen with 68040 released in 1990 had a retail price of $9,999. The NeXTdimension graphics card was $3,995 sold separately.

ppcamiga1 Quote:

060 was even worse. 060 was and is ridiculously expensive. When I try to return to Amiga at the end of 90 it cost more than ppc plus bvision to Amiga 1200. This was pure BS.

Amiga hardware was cheap and really didn't go that high end. The 68060 was pretty cheap until the Amiga retro revival and all the new 68060 boards. A full 68060 could be found for under $50 and rev 6s for less than $100. Now regular 68060s may go for closer to $100 and rev 6s for over $200 in rare cases they can be found. Much more advanced SoCs on Raspberry Pi hardware can be produced for only a few dollars but PPC Amiga like hardware is now for the classes where CBM hardware was for the masses.

@Srtest

Quote:

What he's saying that in 1992, the cheap Amiga model was no longer at that favourite spot of being cheap enough to get compared to more expensive alternatives and being ahead of the curve compared to your run of the mil pc or a non-attractive counterparts.

What you and I guess he is saying is that unlike the A500, the A1200 could no longer be used to make people who bought a PC or Mac feel like idiots for spending twice as much on a machine that does half as much?

Upgrades for the Amiga were pricier than common ISA/VESA cards while Commodore was still around, and definitely more costly than common PCI cards after Commodore was gone and the Amiga market was shrinking. Common sense.

But upgrades for the Amiga were the least disruptive path for productivity going forward.
1. They just worked.
2. Users didn't need to spend additional money on software that would be needed when switching to another platform.

Also, as I already said in post #33, the Windows ecosystem didn't have as many creative applications as Amiga and Mac did. If I wanted to run spreadsheets, then a 386+VGA PC would've been better in 1992, but otherwise a machine that would've cost 50-100% more than a similarly spec'd A1200 was useless to me.

I too did end up buying a Pentium 90 PC in late 1995, with 8MB RAM, 2MB SVGA card, SoundBlaster sound card, IDE card, IO card, FDD, 500MB HDD, CD-ROM drive, Windows 95 + Encarta on CD, KB+Mouse. Without the 15" monitor it cost me $2,200 AUD. If ppcamiga1 is correct, that would've been 500 in rich European money at the time.

In 1992 I spent $1,900 AUD on my A1200 HD with 8MB + 68882 expansion. Even when I include the $500 AUD I spent in 1993 to upgrade the HDD to 200MB, I still consider the A1200 more useful between 1992 and 1995 than the Pentium 90 PC was between 1995 and 1998.

Sure, the Pentium 90 could play the latest PC games like Doom II and Quake, but it sucked at most creative tasks, especially video. I didn't sell the A1200. I used it side by side with the PC until about 1999. PC's still sucked at video in 1999. I bought a used PowerBook G3 333MHz, which ran circles around a Pentium III 800MHz when it came to DV editing. It rendered slower, but the end-to-end production experience was way better, and the rendered videos never had any stutter.

If it was just about gaming for some of you, then I understand: The Amiga hardware was borderline out of date when it was released in 1992. Spending money on costly upgrades just to be able to play games does not make economical sense, especially when all these cool games where coming out almost exclusively on the PC after Commodore was gone.

_________________
All the way, with 68k