I recall one forum poster was asking about our Unicode support - the game itself used only ASCII but it supported players e.g. chatting with non-ASCII characters. The main troubles I foresee are renormalising the incoming text so the trigger pattern matching works on the "right-thing" and on handling strings where counting the bytes is not a way of determining how much space they take up on screen or of how many "characters" they are - for characters I mean of course graphemes .
And as for supporting Chinese - that may be a couple of Major Mudlet releases away. What got me looking at the replay code in the first place was when I started to look at non-ASCII file name support in the TLuaInterpreter code (for version of Windoze latter than XP I think) which allow non-ASCII for user names and by default uses their user name at for their home directory name - POSIX specifications don't permit this and whilst trying to set up a non-ASCII named user on my system I found a bug with the Debian adduser(9) script that meant the --force option to permit such naughty things didn't work... :frowning:
We could the tab character as only taking up 1 'character' in some cases and 8 in others. :(
This is where we add a configurable option for setting how many spaces a tab should have and allow it to be set via the API
The issue is with the view, which doesn't do the right thing