Really? Are you saying that C on a 370 isn't ASCII??
EBCDIC is a 16 bit character set. So, sizeof(char) is 2 bytes and all the C library funcitons (and the sorts since EBCDIC sorts differently from ASCII) have been rewritten to EBCDIC??
Really??
Last time I checked, EBCDIC was also an 8 bit character encoding. Have a look
at
this table. EBCDIC used 8 bits all the time where ASCII used only 7 bits;
the 8th bit came into play when they realized that they had to go 'international' ;-)
IBM never had those intentions; quite smart actually because ASCII lacks the
bitwidth for true Unicode anyway (except for those UTF/8 hacks).
This is a nice
link that clearly shows where EBCDIC came from.
And where it's going ;-)
kind regards,
Jos
ps. There's no need to re-implement anything w.r.t. the sort method, i.e. the
things just get sorted according to EBCDIC, that's all.