Full unicode support in python 2

[sdvillal]

I would rather ditch python 2 than supporting this.

Two main issues here: python 2 default encoding being "ascii" (so implicit conversions simply fail and one needs to litter code with decodes and encodes) and whatami relying too much in using repr + unicode repr adding these pesky 'u' (which litter identifiers and therefore would break old identifiers and warrant a major version update). By whatami nature, forcing these "u" in identifiers would also split up ids in py3 and py2, which should be avoided at all costs.

So this is a wontfix and the solution should be to move downstream dependants to use python 3 as everyone should be already doing these days.

Some basic literature:

• https://pythonhosted.org/kitchen/unicode-frustrations.html
• https://docs.python.org/2/howto/unicode.html
• http://lucumr.pocoo.org/2014/1/5/unicode-in-2-and-3/
• http://python-future.org/unicode_literals.html
• http://flask.pocoo.org/docs/0.12/unicode/
• https://www.joelonsoftware.com/2003/10/08/the-absolute-minimum-every-software-developer-absolutely-positively-must-know-about-unicode-and-character-sets-no-excuses/

[sdvillal]

Another problem is with arpeggio and text, we would need to override how terminals are handled. There is a py2 unicode experimenting branch that shows how this could be done, but as anything else, it just introduces too much variance and corner cases.