From owner-freebsd-bugs  Tue Jun 24 19:10:03 1997
Return-Path: <owner-bugs>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id TAA23378
          for bugs-outgoing; Tue, 24 Jun 1997 19:10:03 -0700 (PDT)
Received: (from gnats@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id TAA23372;
          Tue, 24 Jun 1997 19:10:01 -0700 (PDT)
Date: Tue, 24 Jun 1997 19:10:01 -0700 (PDT)
Message-Id: <199706250210.TAA23372@hub.freebsd.org>
To: freebsd-bugs
Cc: 
From: "Jordan K. Hubbard" <jkh@time.cdrom.com>
Subject: Re: misc/3946: Errors in RELENG_2_2 
Reply-To: "Jordan K. Hubbard" <jkh@time.cdrom.com>
Sender: owner-bugs@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

The following reply was made to PR misc/3946; it has been noted by GNATS.

From: "Jordan K. Hubbard" <jkh@time.cdrom.com>
To: Conrad Sabatier <conrads@neosoft.com>
Cc: FreeBSD-gnats-submit@FreeBSD.ORG, brian@FreeBSD.ORG
Subject: Re: misc/3946: Errors in RELENG_2_2 
Date: Tue, 24 Jun 1997 19:02:02 -0700

 > I didn't realize this could cause a problem.  I had always thought make
 > depend was a good way to do a quick check to make sure everything needed to
 > do a build was there.  Apparently, this is not true in the case of building
 > the complete source tree?  
 
 No, it's never been the case that this will work when certain files in
 /usr are sufficiently out of sync with /usr/src (and when the files in
 question are include files, this is easy to do).
 
 You should always use the world target when you're even slightly
 unsure - only it can guarantee that the proper dependency ordering is
 achieved (and sometimes not even then, in which case you have to
 shepard the build process through by hand until the delta is
 small enough to allow world to work again).
 
 					Jordan