From owner-cvs-user Tue Apr 9 12:24:08 1996 Return-Path: owner-cvs-user Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id MAA05661 for cvs-user-outgoing; Tue, 9 Apr 1996 12:24:08 -0700 (PDT) Received: from grumble.grondar.za (root@grumble.grondar.za [196.7.18.130]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id MAA05575 Tue, 9 Apr 1996 12:23:40 -0700 (PDT) Received: from grumble.grondar.za (mark@localhost [127.0.0.1]) by grumble.grondar.za (8.7.5/8.7.3) with ESMTP id VAA08651; Tue, 9 Apr 1996 21:23:27 +0200 (SAT) Message-Id: <199604091923.VAA08651@grumble.grondar.za> To: Mike Pritchard cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-user@freefall.freebsd.org Subject: Re: cvs commit: src/secure/lib/libcipher Makefile Date: Tue, 09 Apr 1996 21:23:25 +0200 From: Mark Murray Sender: owner-cvs-user@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Mike Pritchard wrote: > > >From what I understand crypt.3 is installed by secure/lib/libcrypt which > > is called from src/lib/Makefile. > > Ick. Everything else we have installs its related man pages > in their own makefiles. If crypt.3 is being installed by some other > makefile, then the MLINKS should reside in that makefile, not in > libcipher/Makefile. This is a bit difficult here... > Hmm. Looking, it appears the libcipher/crypt.3 and libcrypt/crypt.3 > are identical except for one small FreeBSD related blurb. > I know this stuff has gone through some revamping lately, > which is why I assumed that libcipher was simply missing > the MAN3=crypt.3 line in the makefile. If it should not > be installing crypt.3, then the MLINKS lines from that makefile > should be moved to the other makefile. Let me explain; the 'exportable' libcrypt has to be (?) compatible with the DES-enhanced one. That lib (with DES) originally also contained the calles that are now in libcipher. These had to be pared out to keep the two libcryps equivalent. > However, isn't libcipher usable even if libcrypt isn't installed? Which > should make it seem to me that libcipher's makefile is now > doing the correct thing. Debatable. Perhaps the crypt.3 should be split - part for libcrypt (crypt(3)) and part for libcipher (the_rest(3)). > Can someone who is more familiar with how all this stuff interacts > these days either fix everything right so that make world will function > if /usr/man was been wiped, or tell me if my fix is fine, or > how it really should be fixed. Hows the above? I'll do it; I'll just be a little slow because my ?roff is not too healthy. M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grondar.za for PGP key