Without this, unicode in posts results in errors when running under py2.
This fixes, I think, EnigmaCurry/blogofile#122.
Note: I've not tested this under py3 (though I have attempted not to break things.)
Reintroduce the .decode() which was dropped in 8766e13
Without this, unicode in posts results in errors in py2.
Thanks for figuring this out! It does indeed resolve EnigmaCurry/blogofile#122.
The six library is already a dependency via the blogofile core, so I just used the change to post.py. So, I accepted this pull request in spirit even though I didn't actually merge it.