IRC channel logs

<oriansj>well octal would just require 3 octal chars per byte; the rest can leverage the same !@$%&label pattern

<plasma41>oriansj: With hexadecimal, both nybbles of a byte can be parsed by directly converting any of the ASCII characters [0-9a-f] to the associated 4 bit value. With octal, though, you must check for [0-7] for the two least significant octal digits, but [0-3] for the most significant octal digit. IMO parsing octal on an architecture with 8- rather than 6-bit bytes feels awkward.

<oriansj>plasma41: completely fair, but if one assumes that the person encoding does everything perfect and always includes a leading zero character (when it isn't 1-3); then one can do a simpler parse. https://paste.debian.net/1317060/

IRC channel logs

2024-05-15.log