Some characters aren't correctly handed by the eggdrop

[KnuX]

Hi all,

I have an eggdrop with the following caracteristics :
- eggdrop v1.6.17: 50 users (mem: 1958k).
- OS: FreeBSD 6.1-PRERELEASE
- Tcl librairies: /usr/local/lib/tcl8.4
- Tcl version: 8.4.11 (header version 8.4.7)

I notice that it replaces 2 caractères to an other one, as these examples show us :

Code: Select all

.tcl putlog "#blablaѥ"

tcl: builtin dcc call: *dcc:tcl KnuX 39 putlog "#blablae"
tcl: evaluate (.tcl): putlog "#blablae"
#blablae

Code: Select all

.msg KnuX "#blablaѥ"

tcl: builtin dcc call: *dcc:msg KnuX 39 KnuX "#blablae"
#KnuX# msg KnuX "#blablae"
[!h] PRIVMSG KnuX :"#blablae"

I am wondering about the reasons of this phenomenon. In this case, if an IRC message contains the "Ñ¥" characters, eggdrop will think it's a 'e' and in fact, its ambiguous.

For example, for the eggdrop, the "#Ñ¥" channel will be the same as the "#e" channel...

I know that "Ñ¥" isn't used everyday, but I am disturbed by this behaviour.

If anyone can help me to understand...

[DragnLord]

the bot's environment most likely defaults to ISO-8859-1 encoding for text, eggdrop may use those characters if you change that (and possibly recompile after changing default language)

[KnuX]

What can I do if I want the eggdrop to distinguish "Ñ¥" and 'e' ?

I don't understand your reply, sorry but thanks you

[De Kus]

what hex codes are these? Are they D1A5?
I have seen similar problems involving non-alphamumeric iso-8859-1 bytes. No idea why it transforms them since eggdrop is supposed to do no encode changing.
Only thing I can say is that it happens same for ^-^.

.tcl putdcc 15 ѥ
[13:06:59] tcl: builtin dcc call: *dcc:tcl De_Kus 15 putdcc 15 e