Ancient problem with a code 255 or "\377" or TLN_IAC
At present the situation looks so (I bring the calculations):
To eggdrop two types of clients can be connected:
1. IRC clients (mirc, trillian, etc.)
2. A telnet clients. (PuTTY, secureñrt, etc.)
Between them there is a basic difference, IRC clients do not support operation with command characters of the protocol a telnet, and a telnet clients, naturally, support operation on the protocol. And itself eggdrop it is sharpened for operation about a telnet by the protocol, but it is not finished. He catches command characters of the protocol and generates some sequences of these codes, but he does not know how to distinguish a control sequence from the text if in the text there is a character which code coincides with the command character. Also does not know how distinguish clients which can to work with command characters and those clients which cannot do it. (the command character is meant as the character which code is equal 255).
Radiating from this the task was designated:
1. To train eggdrop to distinguish clients
2. Knowing how to distinguish such clients, he should know how work correctly with them.
When it became clear in what a problem and what to make, all appeared simply.
At connection to a bot a telnet the client, it is especial at the first stage (login, password), a telnet puts out itself that that he dispatches control codes in the beginning the text, it is the key moment, it are necessary to flag this junction by a certain flag, and in the further, basing on this mark, accordingly and to carry on the dialogue with such client as with the client which knows how to work with command characters. And IRC the client does not dispatch such digits, and eggdrop, therefore, should not catch for such client command characters.
I offer a patch for localization of this problem
http://v-ufal.chel.su/vd/irc/fix255.egg ... 6.diff.zip