Surely, the lesson to be learned is that Codecs are more trouble than they
are worth, and that sooner or later we need to have a Final Solution to
the problem.
surely the lesson to be learned is that those who don't understand history
are doomed to repeat past mistakes. "Final Solutions" are neither.
And Codecs are one of those past mistakes.
wrong. what history shows us is that we can't come up with a single
universal solution for all cases, and that transitions to completely
new systems are very disruptive, so adapting is often a useful tool.
in fact utf-8 is a good example of what you are calling a codec -
it tries to make unicode fit into a box designed for 8-bit ASCII.
UTF-8 may not be a "final solution", but it is a much cleaner solution
than anything else on offer, and it has considerable extensibility built
into it, and a strong and authoritative body (the Unicode Consortium) to
keep it on track. I think is will hold the field at least as long as ASCII
did.
utf-8 is a good piece of design. I won't hazard a guess as to how
long it will last, but it seems presumptious and naive to expect that
it's a good enough solution to the universal language problem that
it's going to be usable without significant change for more than a few
years.
when someone comes up with something "better" - which is to say,
better at working with the then-current set of conditions -
are you going to insist that everyone move to that new charset,
or are you going to insist that they stay with utf-8 for the
sake of compatibility?
Keith