Stefan Beller <sbeller@xxxxxxxxxx> writes: >> So I think this patch is insufficient, and needs to at least change >> the "submodule.active" codepath to return !!ret; otherwise, a caller >> that now expects 0 (not active), 1 (active but can lose URL) and 2 >> (active and the presence of URL makes it so) will be confused when >> one of the MATCHED_* constants from dir.h comes back. > > Yes. > > I'll resend when appropriately. Thanks.
