How to make GNU screen recognize UTF-8 characters [closed]

Question

I like to have different prompts in bash to recognize quickly on which machine I am currently working when I use ssh. I found that the default bash terminal in OSX Yosemite accepts a wide range of unicode characters, so I set up these PS1:

Mac PS1

PS1=$'\n\n\xf0\x9f\x98\x88'"  \t – \[\033[01m\]\u@\h\[\033[00m\]:\W > "

Raspberry π PS1

PS1=$'\n\n\xf0\x9f\x98\xBA'"  \t – \[\033[01m\]\u@\h\[\033[00m\]:\W > "

Unfortunately, GNU screen does not understand UTF-8 characters:

• F0 9F 98 BA 😺 http://unicode-table.com/fr/1F63A/
• F0 9F 98 88 😈 http://unicode-table.com/fr/1F608/

even when run in UTF-8 mode with screen -U

Any idea how to solve this problem?

Answer 1

1) In GNU screen type CTRL+A, then type this screen directive:

:utf8 on

(Not sure why -U does not already do it.)

2) Also check your locale by typing "locale -a" in a shell.

This should have UTF-8 near the top of the output.