Looking at case insensitivity - question

[Paul Moore]

On 18/04/2008, Mark A. Flacy <mflacy at verizon.net> wrote:
> If the repository and all of its clones are on case-sensitive
> filesystems, you won't have any problems with the current
> implementation.

Correct, that's how Mercurial is designed.

>  If the repository and all of its clones are on
> case-preserving filesystems, you won't have any problems with the
> current implementation.

Assuming you mean case insensitive but preserving, wrong. Search the
bug tracker for topic "casefolding". Fixing these bugs is my key
interest, and where the whole discussion stems from.

> Mixing the environments create problems;

Probably. I don't have experience on this. My personal, uninformed,
opinion is that if people understand what's going on, it's not hard to
avoid issues. Problems arise when people take insular views (e.g. Unix
users creating multiple files whose names only differ in case, or
Windows users who freely change filename case without thought, and
without recording the change - either way creates problems).

> I personally would find mercurial treating all filesystems as merely
> case-preserving as a show stopper.

No-one is suggesting that, as far as I am aware. All I am suggesting
is removing the current places where Mercurial assumes filesystems are
case sensitive - but *not* by replacing that with an assumption that
they aren't, just by checking and dealing with the actual situation.

Paul.