🚀 go-pugleaf

I've been following the character set debate for years, and have finally 
arrived at the conclusion that in the long run, it doesn't matter--that the 
goal of constructing a universal character set, a universal mapping of 
numeric values to fundamental units of written language, can only be 
approximated and not achieved.

The problem is that units of written language are not as discrete as they 
seem, any more than units of spoken language are.  Just as no single phonetic 
alphabet can capture all nuances of spoken language, no single assignment of 
code points is going to apply to all representations of written language.  
This is not due to lack of diligence on the part of standards committees, but 
to the fact that different things are important in different circumstances.

Half of the heated argument about text representation seems to stem from the 
idea that we have to solve "the whole problem."  But we can't--the best we 
can do is improve on what we have now, so that we end up with fewer schemes, 
each of which covers more ground.

UNICODE will make some things easier, but it's not a universal panacea.
Neither will ISO 10646 be one, or any other scheme.  They're both just 
distinct improvements over the status quo for many applications.  However, 
the fact that they aren't suited for other applications is not a defect with 
them, but with the attempt to force-fit them into places they won't go.

The criterion against which we should be judging them is utility, not 
perfection.



Amanda Walker
InterCon Systems Corporation