This was actually a further bug, reported as part of the same CVE - you could also overwrite .git/config by adding any of a number of zero-width Unicode characters that many filesystems ignore when checking for filename equality (but string comparison doesn't, of course).
http://en.m.wikipedia.org/wiki/IDN_homograph_attack
What seems really scary about this is that even Unicode has several different ways of comparing strings, and the correct one depends on the exact situation, so the common response of "just use a library" doesn't work; for example, if a user were searching for a filename it might make sense for full-width characters to compare equal to half-width ones, but not if opening a file where you wouldn't want e.g. the full width version of /etc/passwd to be equivalent to the half-width one.
If you don't perform canonical equivalence checking, you could search for "café" and not find a file named "café.txt" if it uses the other representation.
It's useful when I search for "café", if I also get results for "cafe" - Chrome's search does this. Not to mention searching for "don't" and getting hits including "don’t". But that should definitely be restricted to text data operations, rather than lower-level ones, including filenames.
The issue arises when this "text data" includes filenames. Having café.txt and cafe.txt be equivalent when searching is useful, but the real problem is if a filesystem decides that two "equivalent" filenames are essentially identical - to contrive an example, suppose it thought /étc/passwd was referring to the same file as /etc/passwd . It makes checking for and filtering out "sensitive" filenames far more difficult. For example, just take a look at all the ways Unicode homoglyphs and "special" characters can be used to bypass forum wordfilters, and you'll see how difficult that problem is.
(I know permissions, ACLs, etc. can help here with access control, but the problem of distinguishing between filenames still stands.)
http://article.gmane.org/gmane.comp.version-control.git/7076...
"Having programs that get different results back from what they actually wrote, that tends to be a security issue"
Directory-traversal attacks work just as well with ASCII or byte sequences.
[1] of course, what is large becomes less and less important over time, thanks to Moore's law. On embedded devices, this still may be quite significant, though.
I never really thought about this issue.
A path is an identifier. Just as it's infuriating and horrendous to have programming language identifiers act case-insensitively, it's a poor design choice to have path identifiers act case-insensitively.
I suspect that what you describe as a huge benefit to novice users is entirely seen in search functionality, where you are comparing a string (the search needle) against a number of paths (the haystack). Since this search should be conducted by converting the path to a string (never the other way around, that's nonsensical) it's perfectly natural to support CI operations on strings but forbid them on paths.
Sorry, but there's just no way around it.
The common complaint "but it's annoying to have to type in the filenames exactly" can be solved with tab-completion (I'm surprised how many CLI users don't know their shell has this feature), and using sane naming conventions like not GiVinG yOuR fIleS_stUpiD-NaMESLikEth1s.
(I am aware of all the - quite frankly ridiculous - complexity of Unicode characters that are visually identical and "should be treated as such for the purposes of comparison", but I think that's another example of excess complexity leading to things like directory-traversal attacks.)