On Sun, Mar 24, 2002 at 04:49:12AM +0900, Dan Kogai wrote:
On Sunday, March 24, 2002, at 04:18 , Jarkko Hietaniemi wrote:
The CN.t and TW.t are marked "do not go here if EBCDIC", so they are
just skipped. The respective CN/TW/KR.pm should be marked, too,
that's just an oversight, since as Encode now stands, it does
not work.
Okay, I will then mark CN|TW|KR as unsupported on EBCDIC.
The results I got are @15421, so it's Encode 0.95. I can retry
with later Encode, but compiling in this box is really slow.
Let me rephrase that: REALLY slow. I think Encode alone takes
hours to compile (the full Configure-make-harness cycle is more
than 24 hours).
It seems that I sadly have to conclude that any encodings that uses
MSB in bytes as multibyte character mark won't work....
The Golden Rule of Software Engineering: Adding Another Layer of
Indirection Helps. Usually.
Dan
--
$jhi++; # http://www.iki.fi/jhi/
# There is this special biologist word we use for 'stable'.
# It is 'dead'. -- Jack Cohen