A New Amiga Will Go On Sale In Late 2017 (theregister.co.uk)
An anonymous reader quote the Register:
The world's getting a new Amiga for Christmas. Yes, that Amiga -- the seminal Commodore microcomputers that brought mouse-driven GUIs plus slick and speedy graphics to the masses from 1985 to 1996... The platform died when Commodore went bankrupt, but enthusiasm for the Amiga persisted and various clones and efforts to preserve AmigaOS continue to this day. One such effort, from Apollo Accelerators, emerged last week: the company's forthcoming "Vampire V4" can work as a standalone Amiga or an accelerator for older Amigas... There's also 512MB of RAM, 40-and-44-pin FastIDE connectors, Ethernet, a pair of USB ports and MicroSD for storage [PDF]. Micro USB gets power to the board.
A school in Michigan used the same Amiga for 30 years. Whenever it broke, they actually phoned up the high school student who original set it up in 1987 and had him come over to fix it.
A school in Michigan used the same Amiga for 30 years. Whenever it broke, they actually phoned up the high school student who original set it up in 1987 and had him come over to fix it.
Wait, Will Smith used to write emulators?
The only stories we run on Amiga are about the release of a new computer. The comments then always boil down to the same thing:
a) My first computer was an Amiga
b) There's places where the original Amiga is still running
c) This company is a shell with nothing to do with the Amiga that made the Amiga great.
d) This product is too expensive and completely irrelevant.
e) This is a shameless Slashvertisement and is about the only Amiga related stories that gets run here anymore.
This post is in line with e.
680x0 Assembly was elegant, intuitive, and a crap-ton better than Intel's nonsense. :)
But there is a reason it is no longer used much. A single instruction could generate up to six page faults. The 68k put the C in CISC.
It isn't used much any more because it was beaten by x86, three times:
First, it lost to the x86 because x86 is assembly-source-code compatible with the 8080, making it easy to port CP/M software to PC/MS-DOS, giving the IBM PC a huge advantage over any 68k-based system in the early '80s, giving that platform a momentum which it never lost.
Second, it lost to x86 because Intel was able to pump far more money into developing their architecture than Motorola was able to spend on theirs (see my first point!), so despite x86 being weird and inefficient, Intel eventually managed to develop it to the point where they were beating 68k on raw performance; IIRC it was the Pentium that finally buried 68k in terms of speed.
Third, x86 beat 68k even worse, and started beating even the best RISC architectures, when clock speeds got so fast that the memory-to-CPU bottleneck became the determining factor of PC performance, and that bizarre-but-compact x86 instruction encoding became a major advantage.
Nobody cares about that extreme case of six page faults. For one thing, on architectures that don't have this behavior, that same instruction would cause a segfault, and even on 68k, every halfway sane compiler avoids that nonsense by aligning variables to natural boundaries.