It would appear that copy/pasting utf8 strings of (at least) parts of Japanese text still results in it being possible for mirc to select the wrong byte as start of character, leading to garbage when copy-pasting a substring from a Japanese sentence.

It also appears that selecting only 1 or 2 Japanese characters and letting go of the mouse does not copy it into the clipboard, which for "normal" ascii is no problem.

- Pomax

It appears that pasting more than one line of Unicode text is also out. Only the first ever gets sent.