From owner-freebsd-current  Mon Jul 15 20:02:01 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id UAA21809
          for current-outgoing; Mon, 15 Jul 1996 20:02:01 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id UAA21782
          for <current@freebsd.org>; Mon, 15 Jul 1996 20:01:53 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id MAA30766; Tue, 16 Jul 1996 12:57:05 +1000
Date: Tue, 16 Jul 1996 12:57:05 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199607160257.MAA30766@godzilla.zeta.org.au>
To: bde@zeta.org.au, imp@village.org
Subject: Re: Nits
Cc: current@freebsd.org
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>: Yes.  It shouldn't take long to redo it, especially if you don't run
>: `make depend' separately (`make depend all' sometimes uses out of date
>: dependencies).

>Then shouldn't the make world target do 'make depend && make all
>install' rather than 'make depend all install'?  Or is there a reason
>why make world doesn't that?

It should do something like that if necessary.  It usually isn't necessary
because of an intended(?) side effect of the subdirectory handling: the
`depend' target for each subdirectory gets done before any `all' targets,
so the `all' targets for subdirectories get handled by different `make'
processes, so there is no problem.  However this is reported to not work
in some cases.  It certainly doesn't work if you cd to a leaf directory
and run `make depend all install' there.  Note that the subdirectory
handling stops make `make depend all' from being an optimization for the
same reason that it makes it work.

Bruce