I wonder if Commodore could have done better by migrating the Amiga to Intel 486 when prices started to drop to something reasonable. Let’s say, managing to release something around 1991 or 1992. Specifically, put the Amiga team and Commodore PC folks together on a project to develop a new system. Give the thing the next iteration of Amiga custom graphics and audio hardware. Invest in the OS to take full advantage of its unique hardware capabilities, e.g., memory protection and a good GUI. Use software emulation for backwards compatibility with the full existing catalog of Amiga software. Finally, provide a compatibility layer allowing it to either run DOS programs in a window, or perhaps have the system dual-boot to DOS if the performance cost of the former is too high.
Would this have been enough to keep the platform, and the company, alive for another few years?
For many reasons, the fantasy I describe above was never going to be possible. My point was to wonder out loud about what the best possible course of action would have been, and how far that might have realistically taken them…
Commodore never understood the value of software compatibility. They would never provide backwards compatibility with the M68K Amiga, and a new machine would be DOA.
Apple struggled for years to develop a modern operating system with memory protection and preemptive multitasking. It’s a hard problem, and very expensive.
Other machines like the Tandy provided PC-compatible-ish machines with better graphics and sound than PC, but failed anyway due to network effects. Also, you can’t just ship a set of expansion cards for PC and drivers. People love standards and this kind of situation led to the deaths of many a sound card and graphics card company.
It seems likely Windows 95 would have killed the Amiga anyway. They were on borrowed time.