https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120621
--- Comment #21 from James K. Lowden <jklowden at schemamania dot org> --- On Fri, 01 Aug 2025 13:02:50 +0000 "ro at gcc dot gnu.org" <gcc-bugzi...@gcc.gnu.org> wrote: > I think so, yes. However, COBOL has been so much trouble that I've > decided to stop testing on my platforms. Hi Rainer, [off list] I want to make sure COBOL is no trouble. Your testing revealed some real problems. I'm sorry if anything I've done has discouraged you. There is IMO an unresolved question about %z in diagnostic messages. I think we disagree on that point. No one has explained to me -- in words I understand -- why any system, 32-bit Darwin or other, can't be bootstrapped and build the diagnostics support for it. I understand there are C libraries that don't support %z. For any messages that go through printf(3) and friends, I've been careful to remove %z. We made 100s of changes to satisfy cppcheck. As of now, it produces no warnings for the COBOL FE (using our configuration). If there's something I'm doing to make your life miserable, please tell me so I can correct it. Kind regards, --jkl
