Junio C Hamano <gitster@xxxxxxxxx> writes: > While that is all true, I do not think the posted patch as-is is a > good idea. After all, object missing on our side locally before you > start "git fetch" is a local repository corruption, which ought to > be a much rarer event than a still in development and in flux server > end sending nonsense packfiles, no? > > At least, "they didn't do X" would give the user somewhere to start > investigation (e.g. complain to the server folks about the error, > stating where they started from and what they tried to fetch). The > new message may be playing it "safe" by not saying anything you are > not absolutely sure about, but that is much less useful to the users > who got the message. Hmm...that's true. Let me try to understand the revision walking code a bit more, and if by then I still don't end up changing the revision walk to avoid prematurely parsing a parent, I'll make a patch with your suggestion.
