1. When displaying diffs unicode characters (for example cyrillic 
characters) are shows as `<U+xxxx>` entities.

2. I expected to see characters as is.

3. 2.10.2

4. OS X 10.10.5

Attachments

• darcs-screenshot.png

Also, when viewing logs I see same entites in patch names (and I believe 
the same will be true for patch description).  This makes log information 
quite useless.

I thought that we had solved these issues some time ago. Apparently we
did not. I tried this here and it fails, too.

If I set

export DARCS_DONT_ESCAPE_8BIT=1

in my shell, I don't get the <U+xxxx> but some garbage, I guess the
bytes are interpreted as iso latin1 in this case.

I'll try to create a test for this.

I misunderstood. You are referring to the patch content, not the meta
data for the patch.

This is an open problem in Darcs. Can we enable printing patch content
in decoded form (user's locale) with an environment variable, perhaps? I
guess it would be extremely useful when e.g. managing prose with darcs.
This would assume that all patches have been recorded in the same locale
as the user that views them, but nowadays this might be the case more
often than not e.g. UTF8 locales should be compatible no matter what the
language.