Junio C Hamano <gitster <at> pobox.com> writes: > Ah, I overlooked that codepath, but if that if statement triggered, that > would mean fetching is still 0, which in turn means that you never sent > any "want", so "Once all the 'want's' (and optional 'deepen') are > transferred" is not even true, is it? If you want to get pedantic, it IS true that "all" the 'want's are sent; the correct set of wants to send just happens to be empty. What DOES seem incorrect is the part about the 'deepen' (as well as the 'shallow's I'm proposing to add); that part of the code isn't even reached if this termination happens. So, either I'm mistaken and that is NOT the right codepath, or this is a mistake already present in the documentation. Alternatively, the problem there is that it's just deceptively worded. It implies that a 'deepen' can be sent and that termination can still happen afterwards; but I don't believe this is possible. If a depth argument is present, everything_local is not called and COMPLETEness is not set, so it's impossible to skip any refs except in the corner case where there aren't any to begin with. The second version of this patch addresses this better. -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html