RE: Inconsistent treatment of EBCDIC Newline (NL) character in codepage.c
| Newsgroups | gmane.comp.emulators.hercules390.general |
|---|---|
| Message-ID | <[email protected]> |
Max wrote: > When translating from EBCDIC to ASCII, not all of the > translation tables in codepage.c appear to have the > correct translation of EBCDIC NL (x'15') to ASCII > (IMHO the correct value should be x'0A' -- LF). If you search the web (and even IBM's own web site and reference manuals), *most* translation tables have X'0A' labeled as ASCII "LF" and X'25' as EBCDIC "LF" (and "NL" is not listed anywhere). Other tables however, label X'0A' as ASCII "LF" and X'15' as EBCDIC "NL". Very confusing! But the consensus seems to be that ASCII X'0A' is "LF" and EBCDIC "LF" is X'25' (not X'15'). That is to say, most ASCII to EBCDIC tables translate ASCII X'0A' to EBCDIC X'25' (not X'15'), and most EBCDIC to ASCII tables translate EBCDIC X'25' (not X'15') to ASCII X'0A'. Make of that what you will. > This causes trouble when using the IBM 3767 device support > in comm3705.c.... The problem seems to be persistent up > through the latest Hyperion... There is nothing wrong that I can see with Hyperion's codepage tables. Perhaps you are simply using the wrong one? If none of them seem to meet your needs, you can of course define your own (i.e. you can override the unwanted translation) via Hercules's "cp_updt" command. HTH -- "Fish" (David B. Trout) Software Development Laboratories http://www.softdevlabs.com mail: [email protected]