Encoding trouble revisited

Searching on my blog for articles on encodings and how to deal with them will yield quite a few results.

As if there could be a project without encoding trouble. So I’ve been searching for the reason why unix diff finds differences where none are and those are of course caused by some funky encoding weirdness.

If I am lucky my single “good” blogpost on the topic will help me out because I can only reproduce the bug on a machine which does not really have locales configured. I really hope – I know one shouldn’t hope to fix bugs but just find the solution and fix them – that this is it.

But the exact same test acts different on 2 different deployments of my project. So maybe I’ll be done with that one soon.

Update: it’s definitely not simple trouble with locales …