On Dienstag, 23. März 2010, Junio C Hamano wrote: > Johannes Sixt <j6t@xxxxxxxx> writes: > > BTW, the real fix for this potentially problematic case is to teach > > pack-objects to create a pack file for a shallow clone. Then we don't > > need this instance of an async procedure. > > I keep hearing "shallow clone" here, but doesn't "bundle -n 1" essentially > do something quite similar using pack-objects? Do they make a call into > pack-objects in a different way, and if so can we update "clone" to mimic > how "bundle" drives pack-objects? Both *current* implementations of shallow clone and bundle feed pack-objects with rev-list output. In the case of non-shallow clone, upload-pack does not need rev-list because pack-objects can walk the revisions itself. I cannot tell what pack-objects does not have that it cannot walk the revisions itself for shallow clones. -- Hannes -- 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