1997-10-15 - Re: CVS: cvs.openbsd.org: ports

Header Data

From: Warner Losh <imp@village.org>
To: uk1o@rz.uni-karlsruhe.de
Message Hash: 844d08b782701c79e4ce4c5e93bd972122e9730dfd3601e4de669f7b4ff774af
Message ID: <199710151405.IAA20043@harmony.village.org>
Reply To: <E0xLTBc-0001Bo-00@rover.village.org>
UTC Datetime: 1997-10-15 19:10:35 UTC
Raw Date: Wed, 15 Oct 1997 12:10:35 -0700 (PDT)

Raw message

From: Warner Losh <imp@village.org>
Date: Wed, 15 Oct 1997 12:10:35 -0700 (PDT)
To: uk1o@rz.uni-karlsruhe.de
Subject: Re: CVS: cvs.openbsd.org: ports
In-Reply-To: <E0xLTBc-0001Bo-00@rover.village.org>
Message-ID: <199710151405.IAA20043@harmony.village.org>
MIME-Version: 1.0
Content-Type: text/plain


In message <E0xLTBc-0001Bo-00@rover.village.org> uk1o@rz.uni-karlsruhe.de writes:
: How about suggesting a make variable HAS_LIBCRYPT to the FreeBSD people?
: We could set it in <bsd.own.mk> or something like this (to a value
: meaning "we don't have it and we don't need it either), and FreeBSD could
: do similar.

The problem isn't in the BSD make files, but rather in the makefiles
of the ports.  True, many FreeBSD ports' patches add -lcrypt blindly
to the makefile, but that could change.

: Another way were patching the third party program's Makefile accordingly.
: We should introduce a patches-${OPSYS} directory having precedence before
: the patches directory of a port. Thus we could maintain our special patches
: for ports which need them while continuing to use the FreeBSD ports where
: they work out of the box. I see now that this is already implemented in
: <bsd.port.mk>, so we could just USE that.

There already is the patches-${OPSYS} that can be used for this
purpose.

Warner





Thread