On 17 Nov 2022 08:18:39 GMT
Post by greymausthe amiga was doomed because it was fun.
The IBMPC had the name IBM on it.
Quite so, but alone that would not have killed the Amiga because it
was clunky, slow and expensive - which was fine for IBM's market.
However IBM published every detail of the PC (and AT) and the
market for a cheap IBM workalike was so obvious that it exploded almost
instantly and became brutally competitive. Then the chipsets appeared
making PC and AT clones really easy (and cheap) to make. The 80386
delivered the final nail in the coffin for everything else but by then
there wasn't much left to kill apart from the Apple who took some important
niches and hung on. After the 80386 the PC design started to take over the
data centre starting with minis - which by then were mostly unix boxes. The
BSDs and Linux accelerated that process dramatically (wot no license fee).
Notice that IBM no longer play in the market they created, they've
retreated to their comfort zone of high support mainframes.
This monocultural trend has shown no cracks until recently with the
emergence of ARM from the mobile and low power space into the desktop,
laptop and server world (there are some really impressive 48 core ARM
chips in use - don't ask about price). Even here there is convergence with
PCI-e, SATA, NVMe etc turning up in ARM chips now.
BTW ARM RISC ? Have you seen the size of the ARMv8 instruction set
documentation ? If that's reduced these days please don't show me complex.
High level languages used to be more complex than assemblers, I reckon it's
a toss up between C++ with STL and ARMv8 for complexity - Algol68 got lost
in the dust a long time ago.
--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/