Jeff King <peff@xxxxxxxx> writes: > On Thu, Aug 27, 2026 at 10:28:32AM -0700, Junio C Hamano wrote: > >> +void die_for_incompatible_opts(bool opt1, const char *opt1_name, ...) > > I'm mildly negative on this, just because there's no compiler support > for making sure there is an EOF somewhere. Keeping patch 1 and using > LAST_ARG_MUST_BE_NULL would be preferable, IMHO. Let's discard this topic for now. I do not like the second iteration very much, and I do not like the 1/2 preliminary step of the first iteration even less so. Thanks.