Closed jaraco closed 8 years ago
The IRC lib is opinionated about the expected character encoding, requiring UTF-8 from all clients (which is not a requirement in the specification). The IRC library provides several techniques for overriding the encoding, which are detailed in the Decoding Input section of the homepage for the library.
The reason different clients have different behavior probably has to do with the way that each of those clients handle errors, as the IRC library itself is consistent in its error handling.
Original comment by: Jason R. Coombs
Intermittently, on certain (unknown) kinds of network input, IRClib crashes with backtrace similar to the following:
Several different IRC bots in the same channel, running same IRClib, respond differently; one crashes as above, one apparently loses server connection but automatically reconnects, one appears to quit normally with no exception or backtrace. Unknown how exactly this is happening.