On 2020/06/03 16:39, Yasuhito FUTATSUKI wrote: > On 2020/06/02 14:52, Yasuhito FUTATSUKI wrote: >> On 2020/06/02 6:01, Karl Berry wrote: > >>> Also, py3c is apparently necessary. Again it would be nice if configure >>> bailed out if it's not present. >> >> In this case it should also stop with the message: >> >> SWIG python disabled at configure time: py3c library not found >> >> However, the 1.14.0 release tarball contains .swig_py_checked (not to be >> intended, I beleive), which is created if 'make swig-py' is allowed >> by the result of configure (by "._swig_py_checked" target in Makefile), >> so it didn't stop as we expected. >> >> The 1.14.0 release tarball contains also '.swig_pl_checked' and >> '.swig_rb_checked' files. Should they omit from release tarball? >> >> Also, should the configure script remove them? > I confirmed the configure script remove them. So I couldn't reproduce > the case that the confiugre script couldn't find py3c but allowed > building swig-py. > > Could you show me config.log in this case?
I got a response about it, off the list. There is no problem when the configure script couldn't find py3c: 'make swig-py' aborts immediately. Cheers, -- Yasuhito FUTATSUKI <futat...@poem.co.jp>/<futat...@yf.bsdclub.org>
