Running make bytecode VERBOSE=1 helped me to find the last command that is run, which is
fontforge -script feta11.pe feta11.pfb So I turned on FONTFORGE_VERBOSE and ran the same script. It shows the same warning about Mac Roman, but then continues, which was encouraging. It then runs four lines: Select all(); RemoveOverlap(); AddExtrema(); Simplify(0, 2); Then nothing happens. I don't actually get any error messages. Is this just a script that will take a very long time to run? The first time I let it run for probably 8+ hours. Am I just being too impatient? Thanks again for the help in getting this far. On Sun, Aug 25, 2024, 11:04 PM Werner LEMBERG <w...@gnu.org> wrote: > > > > > I replaced the system iconv with gnu-libiconv, so that when I did > > "iconv -l" I can see MAC ROMAN as an option. > > Interestingly, `iconv -l` doesn't show 'MAC ROMAN' on my GNU/Linux > box, which comes with `iconv` from GNU libc 2.31. Anyway, FontForge > checks for an encoding called 'Mac' (which should be the same as > 'MAC'). > > > Then I did a clean configure and rebuild of lilypond. But alas, at > > the same point I got the same message and it hanged. > > Please do > > ``` > make bytecode VERBOSE=1 > ``` > > and show the relevant warning and/or error messages. What exactly do > you mean with 'hanged'? > > > Are there any other thoughts on how to get past this? > > You could manually disable the offending error message in FontForge > (file `fontforge/encoding.c`). However, I can't imagine that such a > serious build problem hasn't been addressed already or at least > mentioned in FontForge's issue tracker (which is ont, AFAICS). > > > Werner >
