L12@DHDURZ1.BITNET (04/21/87)
Date: 21 April 1987, 12:41:52 GMT
From: Wolf-Dieter Batz + 6221 547368 L12 at DHDURZ1
To: INFO-ATARI16 at SCORE.STANFORD
Here's some news concerning ASCII-EBCDIC character exchange:
Watching the C-source of "PAD61" I discovered strange signs where normally
brackets are found in C. On my IBM terminal C-function brackets are HEX-coded
as 'C0 and 'D0. In the source obtained from the net I found '8B and '9B.
new information comin' soon *** wo
Frat hmm@laura.informatik.uni-dortmund.de (Martin Mosner) (04/26/87)
Wolf-Dieter, your problem is to a big extent caused by the inconsistency in the various definitions of the EBCDIC character set which the IBM mainframes on bitnet/earn use. We got caught by the same symptom that you mention (different encodings for braces), and our investigations made clear that there is no consistent EBCDIC definition for some characters. So it's not the fault of the ASCII/EBCDIC translation, but the fault of IBM who sells terminals and printer chains which have different characters at the same code positions and vice versa. My advice: You should try to find out what the actual ASCII->EBCDIC encoding in your case is, and patch your IBM PC 3270 emulation to exactly reverse this encoding. It should be possible somehow... Hans-Martin PS: If you want to ask further questions, feel free to ask. I'm HMM at UNIDO on BITNET/EARN. D