Hi all, I'm decyphering an encoding scheme that was used to compress source code decks. It's from an old obscure IBM 1130 program (CMXP, written by Wil Baden, part of set of enhancements that let source code be stored, edited, and compiled directly from disk. Radical! Visionary!) The compressed format uses 6 bits to encode each character, but it's not any of the standard IBM 6-bit EBCD schemes I've found through Google. It *might* be unique to this compression/decompression program, whose source I haven't found yet. (And if the source does turn up it'll likely be compressed using this format -- so unless the character codes are in the source as hex constants, the source may not help). Here is what I've been able to map out so far, from a small fragment of an encoded source deck: Hex value 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 00 sp A B C D E F G H I . ) 10 + J K L M N O P Q R * 20 - / S T U V W X Y Z , ( 30 0 1 2 3 4 5 6 7 8 9 = where "sp" is space. I am absolutely certain of the encoding of the space, alphabet, digits, slash, comma, period, parentheses, and asterisk. I'm pretty sure of the plus, minus, and equals. For the ones I've left blank, I have no idea yet; either the codes aren't in the sample I examined or I can't guess what the character is from the context. If I read in more cards they might become clear. I've found some CDC codes where / precedes S, but the row ordering is different. Does this encoding ring any bells for anyone? (Again, it might not belong to any standard). Thanks Brian