"Previously [comparing Unicode with a non-ASCII string] would raise a UnicodeDecodeError exception." Python 2.5 became more weakly typed!
miniblog.
Related Posts
Today I learnt that null bytes are valid in UTF-8 and ASCII strings!
It was a surprise -- \0 doesn't feel like text -- but it'd be even more awkward when using C-style strings.
Today I learnt that Emacs will do a best-effort parse of expressions in comments too, unless you set parse-sexp-ignore-comments.
That's all well and good until org.el has ASCII smileys! https://github.com/bzg/org-mode/blob/4d8d7d6cb42e388572b4f5d227e9b3c9da6ca4a7/lisp/org.el#L4815-L4816 ("unbalanced paren")
It's hard to demonstrate how difftastic works in screenshots, so I've prepared a <60 second ASCII screencast of usage!
(Feedback welcome.)
