On Thu, Oct 08, 2026 at 03:59:10PM +0300, Eli Zaretskii wrote: > > Date: Thu, 8 Oct 2026 13:46:40 +0200 > > From: Patrice Dumas <[email protected]> > > Cc: [email protected], Bruno Haible <[email protected]> > > > > On Thu, Oct 08, 2026 at 09:51:31AM +0300, Eli Zaretskii wrote: > > > > This seems like a good hypothesis, probably MSYS2 uses nul and not NUL, > > > > so it does not match. I can fix that by making the comparison > > > > independent of the string case for nul. > > > > > > File-name comparison for the MinGW build of Texinfo should be > > > case-insensitive anyway, because the OS filesystem is. > > > > As far as I can tell, it is not what we decided in the past, see > > https://lists.gnu.org/archive/html/bug-texinfo/2023-01/msg00038.html > > https://lists.gnu.org/archive/html/bug-texinfo/2023-01/msg00057.html > > > > The user should set explicitely -c CASE_INSENSITIVE_FILENAMES=1 > > I meant file-name comparison in the test suite. The above discussions > talk about something entirely different.
Ok, I get it. Where we use CASE_INSENSITIVE_FILENAMES=1 in the tests, there is a comment that says exactly what you say, that it needs to be set otherwise on case insensitive file systems the tests results would be different. So I think that we are ok, and if it happens that there are other tests that need CASE_INSENSITIVE_FILENAMES=1, we'll set it. Having all the tests pass with native Perl on mingw clang ucrt and with your tests is an indicator that we are ok. And there is no evidence that the tests that still fail with MSYS2 mingw gcc ucrt fail because of filesystem case sensitivity. -- Pat
