Jeremie Courreges-Anglas <j...@wxcvbn.org> wrote: > On Wed, Dec 21 2022, "Theo de Raadt" <dera...@openbsd.org> wrote: > >> But yeah, maybe we'll just flip the default option in LLVM and then > >> we'll just not use that warning... at all? > > > > is this specific warning finding dangerous bugs? is it finding a > > substantial > > number of dangerous bugs? > > No, but if we remove -Wdeprecated-non-prototype from -Wall, software > built with clang -Wall on OpenBSD won't hit this warning and may later > fail when built with clang -Wall on other systems. *I* don't care that > much, but I suspect that this is not the additional warning that will > hurt us most...
well, i'm not sure it is bad if some other people get the opposite effect. I don't think anyone disagrees with the move towards more strict C. But the pain must be evenly distributed, and every maintainer must adopt the practice on their own schedule, well mostly. it is the maintainers who must adopt the practices, one after the other, without workarounds imposed by downstreams. If we end up with 100 ports Makefiles containing these -W, even after upstreams adapt to the new reality those -W options are going to stay around and it will be 2038 before the last one is removed by someone digging for bones
