Hi guys,

I've been trying to setup my Git repositories to work with Review
Board. I'm using Windows Server 2003 as the OS. I'm using mSysGit to
use Git under Windows.

The problem I have is that Review Board always fails to complete
GitClient's is_valid_repository function.

My repository is in drive D, in folders Repository\Git\sdu-test.git;
therefore, the path I use is file://D:/Repository/Git/sdu-test.git

I modified GitClient's __init__ method:
if url_parts[0] == 'file':
            self.git_dir = url_parts[1] + url_parts[2]
As you can see, I changed it from url_parts[2] to url_parts[1] +
url_parts[2], otherwise it would crop the path string and remove the
'D:' from it.

So far, it seems that _run_git is running fine. Within __init__, it is
used like this:
p = self._run_git(['--git-dir=%s' % self.git_dir, 'config',
            failure = p.wait()
And failure does not occur.

However, when is_valid_repository is reached, the failure occurs:
p = self._run_git(['ls-remote', self.path, 'HEAD'])
        errmsg = p.stderr.read()
        failure = p.wait()

I don't understand why this fails. If I open a command prompt and type
'git ls-remote file://D:/Repository/Git/sdu-test.git HEAD' (based on
the value of self.path), it works and prints:
991bb49dbacd5bf4123ab9ac5c2285836a01af35        HEAD

If you're curious, here is the error message that appears in debug:
error: cannot spawn sh: No such file or directory fatal: unable to

Now, if my Git executable weren't properly set in the PATH, then I
assume the code would fail at the first call of _run_git. It is not
the case here.

What's going on?

Want to help the Review Board project? Donate today at 
Happy user? Let us know at http://www.reviewboard.org/users/
To unsubscribe from this group, send email to 
For more options, visit this group at 

Reply via email to