> > An argument could be made that we should not merge patch 2 just yet due > > to the fact that some server implementations (such as Git and JGit) > > still exhibit the old behavior, and the resulting clones (albeit failing > > fsck) are still usable, because when attempting to load the blob, Git > > will automatically fetch it. I'm on the fence about this, and have > > included patch 2 in this patch set nevertheless for completeness. > > I think the latter is probably a good thing to nudge the server > implementations in the right direction by gently poking them ;-)
OK, I'll keep patch 2 then. > The patches textually apply cleanly on 'master' but apparently it > needs some fixes in jt/partial-clone-fsck-connectivity topic to > function correctly? I forgot to mention in the original e-mail that this is on bw/ref-in-want. (I mentioned it in an e-mail I sent later [1].) I've checked and basing it on both bw/ref-in-want and jt/partial-clone-fsck-connectivity (which is based on bw/ref-in-want itself) both work, so you can choose either one. [1] https://public-inbox.org/git/20180706193847.160161-1-jonathanta...@google.com/
