It would be interesting if the module could work per channel not only per server. To have #blabla in utf8 and #foo in ISOxxxx.
- yeah that's on my to do list, but I won't work on it any time soon. Flakes 20:05, 23 June 2010 (CEST)
Does it mean that correct HTTP output is possible only if server encoding is UTF-8?
You describe you module as "normalisation" or "recoding", but on this software there are really three encodings, not two: client’s, server’s and HTTP (which chosen to be always UTF-8). If my client speaks a codepage, then you propose to say something like «/znc loadmod charset codepage UTF-8» to get reasonable display at HTTP, but this implies that your module will put UTF-8 to the server. What should I do if I want not to have IRC traffic actually recoded (i.e. both client’s and server’s encoding to be, say, Windows-1251) but to have a reasonable display on HTTP? For use in my network, I could consider to use server-side Unicode, as my IRCd easily converts UTF-8 to internal representation. But this may lead to message size restriction and certainly adds extra complexity to ZNC account setup. On the other hand, I can throw UTF-8 away and use Windows-1251 on HTTP, but this will be buggy for browsing those accounts which owners prefer Unicode. Qq 22:05, 29 June 2011 (CEST)