Date: Tue, 19 Oct 2004 07:44:32 +0400 From: Andrey Chernov <ache@nagual.pp.ru> To: Peter Wemm <peter@wemm.org> Cc: cvs-all@FreeBSD.ORG Subject: Re: cvs commit: src/gnu/lib/libreadline/readline Makefile Message-ID: <20041019034432.GA84326@nagual.pp.ru> In-Reply-To: <200410181755.37423.peter@wemm.org> References: <200410180836.i9I8afRS060144@repoman.freebsd.org> <200410181755.37423.peter@wemm.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Oct 18, 2004 at 05:55:37PM -0700, Peter Wemm wrote: > No. This is wrong, and a NOP change. Please back this out. How NOP change can be wrong in the same time? > Any dependencies will be silently "fixed" by ld, so no matter what, you > end up with a DT_NEEDED of libncurses.so.5. I am aware of that. It is internal ld/libraries/ldconfig things application shuld NOT know about. Because it is unneded details for it. Dependences created at _applications_ level, not at (often changed) system architecture level. > Except that by using the wrong name, you obfuscate things. I use the right name - the name of library libradline really _depends_ on. The fact it becomes ncurses by chance is implementation details. It can be anything in some day, f.e. libc. Even it will be libc, there it is no reason change -ltermcap dependance to that. -- Andrey Chernov | http://ache.pp.ru/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041019034432.GA84326>