What you define as a UTF8 bitstream can easily be a bunch of ANSI chars to me.
I guess that's the reason why mIRC behaves like that.
And as I already mentioned, I've never seen such character combinations appearing randomly in a text without being meant to be UTF8 encodet.
It's not only "my special scenario", many people using high ascii char themes got the same problem, thats not limited my community.
Of course you could treat every word seperately, but in my opinion, that does not replace the old behavior.
(Correct me if I'm wrong, but I think mIRC behaves like that:)
If UTF8-Display is activated, mIRC checks a line for ansi-chars to know if the client the text came from encodet that line.
If it's able to find any high ascii chars in that text which do not belong to a properly utf8 encodet character, it assumes that the whole line is not utf8 encodet.
In case of text events, this is only possible if the client does not encode at all.
Checking word-seperately could now result in the same mistakes as decoding the whole line regardless of high ascii chars. (only less often)
I would really appreciate a possiblity to simply decode and display all appearing utf8 codes in the line without limitations or checkings before.
Be honest, did you ever type € ?
There is a reason why utf8 codes are not made out of common characters...
And I also think, that additional option is not hard to realize, would not hurt anyone (you dont have to make it the default option) and solves those widely spread utf8-display problems easily.
