I was working with utf-8 and i noticed these two bugs:
1) When you have utf-8 on more lines sometimes the bytes of the char are splitted between the lines (if the char has 2 bytes the first is at the end of the first line and the next at the beginning of the second one). If you change the width of the window the characters may or may not be showed correctly as you can see in these screenshots:
This shows the same text (as you can see in the timestamp) with different width of the window. Also when the characters are showed correctly, they don't fill the width of the line.
2) When you have some utf-8 text in your editbox and you autocomplete a nick with TAB, all the text will turn in question marks:
These is tested with mIRC 6.3 on Windows XP SP2 (even if I noticed these problems with the previous version of mIRC too).
In the font option I've set UTF-8 -> Display and Encode as default.
Taggnostr.
-Edit-
I tried to paste the text I used but I can't copypaste it from mIRC, btw the hiragana characters (U+3041 - U+3096) are showed correctly, and the first "set" of the katakana (U+30A1 - U+30FF) too, but just if they are alone. It seems that if there are some "strange" characters in the line (i can't undersand if they are the ones with 3 bytes or what else) all the line is showed in the wrong way (depending on the width of the window).
Here you can find
the list of unicode characters; my bot parse this list and return the character of the line that contain the given word (for example all the line with the word katakana).
In the mIRC Options i've selected, under IRC -> Messages, Multibyte display, Multibyte Editbox, Utf-8 display and Font linking.