On 8/26/25 15:05, Tony Harminc wrote:
...
Seems like a perfect use case for the much derided UTF-EBCDIC.
...
What need would that address?
It would provide compatibility with APIs that allow EBCDIC characters, as
UTF-8 does in the ASCII world. What need does UTF-8 address?
...
When I confronted UTF-8 on z, rarely, perhaps only an experiment,
I did:
iconv -f ISO_8859-1 -t IBM1047 <UTF-8.file
ssh performed the reverse conversion, and my
desktop system, knowing no better except that
my locale had iconv -f ISO_8859-1 -t IBM1047 <UTF-8.file
properly displayed the Unicode values of the
multibyte characters.
Would UTF-EBCDIC have availed me anything?
Does ssh grok UTF-EBCDIC?
Likewise, from my UTF-8 terminal I submitted
jobs via FTP with ;SBDATATACONn(819,1047)
If I had a UTF-EBCDIC JCL file with Unicode
characters in comments or SYSUTY, would they
display correctly in SDSF, subject to the
ability of my terminal?
The solution to incompatibility is not to add
complexity. That lesson should have been
learned from Genesis 11.
Why not let the access method handle it with
//SYSUT1 DD CCSID=1208?
--
gil
