The worst idea was definately AGA. More resources should have been put into making AAA available when AGA came out. The A3000+ should have been the A4000. A4000 was filled with poor choices and feels like it was rushed out. For one, the loss of the flicker-fixer were just idoitic. When I upgraded from my A3000 to an A4000, I was so upset that my newer computer couldn't use the nice monitor I was using on my A3000 and I had to spend all this extra money on an expensive 1942 monitor.
The decision to put a HD floppy in the A4000 was good the fact the drive was 1.5 height left you with a completely useless 0.5 height drive bay. But to make matters worse, since the A1200 didn't come with an HD drive, no commercial developers used HD floppys. This left the HD drive in A4000s useless except for personal purposes and game were coming on 14+ DD floppies instead of a more reasonable 7.
Also the A1200 should have come with an 68030 as standard (or a 44 or 50Mhz 020). 25Mhz 020's were too slow in comparison to standard PC's and Macs of that time.
But really, I think the biggest problem was that ECS was just too good and AGA was not enough of an improvement. Developers mostly just targetted ECS machines as that market was bigger and as a result there was not enough incentive for many people to upgrade to an AGA machine. ECS thrived for much longer than it should have and AGA machines never sold as well as they needed too.
@MPiva
when you say ECS you really mean the Original Chip Set, don't you?
I think that the most disappoining chipset was the ECS (the A3000/A600 chipset, relesed in 1990) not AGA.
ECS came out in 1990, 5 years later than OCS. The only improvements were genlock capabilities and the Shres and 31Khz modes. Since the latter were limited to 4 colors chosen from a 64 palette, they were not useful for games, and very limited for professional application. Also note that such modes on ECS (unlike AGA) use a palette encoding different from the standard Amiga palette making more difficult to have draggable screens with SHRES and other res. at the same time.
So in 5 years C= gave us only minor improvements, not suitable for the most successful Amiga applications, i.e., games.
AGA came out in 1992. I agree that is was not enough to keep up with other platforms, but with respect to ECS was a BIG improvement, released only 2 years later.
If ECS would have been released in 1988 and AGA in 1990, then AGA would have been a very good evolutionary step!
ok, my own 3 worst ideas/product
1. Not choosing a strategy, a market target. They want to market Amiga for professionals as well as for games. They also wanted to explore new markets (CDTV) too far ahead of itheir time
2. Not enough R&D, splitted between software or for hardware. Amiga started with the most advanced hardware and a very promising but still immature OS. Then C= developed a lot the OS (2.0 was a big improvement) and not enough the hardware (ECS and A600). Then tried again to do both. Probably it would have been better to concentrate either on
the OS (like Apple) or on the hardware (like console makers). best of all, of course, would have been to spend a lot more of their cospicuous profits in R&D :-D
3 ECS and (closely related) the A600
