Commits to core *must* have reviewers

[Colin Walters]

Hi,

A lot of commits have been going by for the Windows work, and I'm glad
to see code being written in general.  However, I strongly believe
that a process of code review is essential for the long term health of
any important Free Software project, of which DBus is no exception.

It's a bit unclear to me which of the Windows commits are being
reviewed or how - could one of you guys fill me in on that?

I have to specifically draw the line though for commit
39aa8090241b2588eeb2df05880d71abaf8c526e (
http://cgit.freedesktop.org/dbus/dbus/commit/?id=39aa8090241b2588eeb2df05880d71abaf8c526e
)

This function is broken in multiple ways:

* Superficially, the documentation is not capitalized and punctuated
correctly, and does not conform with the coding style (space between
identifier in opening paren)
* More importantly, "isupper" and "tolower" are locale-dependent and
in general should never be used.  For more information, see:
file:///usr/share/gtk-doc/html/glib/glib-String-Utility-Functions.html#g-strncasecmp

I've reverted this commit, and would be happy to review a replacement commit.