Arnon Yaari

README.rst should be ascii. This was fixed in 688f6cc by @tartley, but the latest release, 0.4.4, happened on a side branch that did not include this commit. This will be...

@cjerdonek I don't think this is it (although this could be a problem too, maybe). The code you mentioned is only called at program exit. This problem is caused by...

I pushed a fix to my personal forked repository that separates the `WinTerm` instances - https://github.com/wiggin15/colorama/commit/eccb87c747d8ac712aeab671ad7e1dba0f83b1bf This should fix the original issue. Unfortunately I don't have time to properly test...

Looks like it's broken in Python 3.5, probably related to this change: http://bugs.python.org/issue24402 The input function now doesn't print to the wrapped stdout because it's different than the original. It's...

Hi @al3xv3gas . I pushed the workaround with the 'fileno' method, and submitted a new bug in the Python issue tracker: http://bugs.python.org/issue28373 Please try the latest version in this repository...

The fix causes issues for dependent applications - see #130. I'm reverting the fix and reopening this issue. We'll find a different fix.

@Delgan you are correct, it should be stripped. Due to the nature of the CSI 'm' (color) command, multiple codes can be specified in the same command. This is why...

On Windows 7 this returns True. The check should include the major and minor, not just the build. A tuple comparison should work.

Hi. I think the check will fit more neatly in `AnsiToWin32.__init__` where the default strip/convert values are determined (I think we need convert=False and strip=False). Maybe we can add the...

Thanks for this pull request. This looks good. One thing I noticed is that on my Mac terminal, writing the swap ansi code (ESC 7m) twice doesn't re-reverse the foreground...