Re: Cone 0.57 now available.
Sam Varshavchik <[email protected]>
| Newsgroups | gmane.mail.cone,gmane.comp.lib.ncurses.bugs |
|---|---|
| Message-ID | <[email protected]> |
Jim Gifford writes: > Sam, > With the latest patch of ncurses, I get the following build error. > Ncurses with patch-5.3-20040110.sh from > ftp://invisible-island.net/ncurses/5.3 > > In file included from cursesscreen.C:9: > cursesscreen.H:42: error: conflicting return type specified for `virtual > NCURSES_BOOL CursesScreen::writeText(const char*, int, int, const > Curses::CursesAttr&) const' The problem is that for some reason your configuration now has ncurses.h #define-ining bool to something else. I'm not exactly up to speed on what ncurses settings trigger that, but you can fix this now by manually sticking "#undef bool" right after #include <curses.h>, in this header file. But the bigger question is this: If curses.h does something as invasive as "#define bool somethingelse", I think curses.h should also undefine the macro when it's all over. Otherwise you'll end up affecting a whole bunch of stuff elsewhere.
signature.asc
(application/pgp-signature, 189 B)
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (GNU/Linux) iD8DBQBAAzEgx9p3GYHlUOIRAuM0AJ4jLQWj9/szOCsLkOcIYyoECl8JuwCfVG0j CCtIDvHmgUsj1vVOGLgJuhk= =UY2b -----END PGP SIGNATURE-----