From owner-freebsd-ports  Sun Aug 23 15:41:15 1998
Return-Path: <owner-freebsd-ports@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id PAA01898
          for freebsd-ports-outgoing; Sun, 23 Aug 1998 15:41:15 -0700 (PDT)
          (envelope-from owner-freebsd-ports@FreeBSD.ORG)
Received: from vader.cs.berkeley.edu (vader.CS.Berkeley.EDU [128.32.38.234])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA01893;
          Sun, 23 Aug 1998 15:41:11 -0700 (PDT)
          (envelope-from asami@sunrise.cs.berkeley.edu)
Received: from bubble.didi.com (sji-ca10-63.ix.netcom.com [205.186.214.63]) by vader.cs.berkeley.edu (8.8.7/8.7.3) with ESMTP id PAA19400; Sun, 23 Aug 1998 15:40:25 -0700 (PDT)
Received: (from asami@localhost)
	by bubble.didi.com (8.8.8/8.8.8) id PAA02331;
	Sun, 23 Aug 1998 15:40:20 -0700 (PDT)
	(envelope-from asami)
Date: Sun, 23 Aug 1998 15:40:20 -0700 (PDT)
Message-Id: <199808232240.PAA02331@bubble.didi.com>
To: ac199@hwcn.org
CC: vanilla@FreeBSD.ORG, ports@FreeBSD.ORG
In-reply-to: <19980823133519.A242@zappo> (message from Tim Vanderhoek on Sun, 23 Aug 1998 13:35:19 -0400)
Subject: Re: manpages (Re: cvs commit: ports/devel/p5-Term-Query p5-Term-Query-2.0.tgz)
From: asami@FreeBSD.ORG (Satoshi Asami)
Sender: owner-freebsd-ports@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.org

 * When I do use clean, it's because I want it to recurse.  I suspect
 * it's easier to type the appropriate ``rm -Rf'' most of the time
 * than to type the appropriate unset+reset the few other times---
 * often I've finished typing the following ``make target'' before
 * the rm even returns.

Well, I have NOCLEANDEPENDS set and explicitly type "make
clean-depends" at the rare cases when I want it to recurse.  (I often
have too many port builds running at the same time it's too dangerous
to recurse on clean....)

 * From a sampling of 33 ports(*): four appear to go out of their way to
 * patch-in support for NOMANCOMPRESS and ten appear to support
 * NOMANCOMPRESS in their virgin sources.

I think the patch support can be taken away, since bsd.port.mk handles
it now.

 * Add to this the two ports I manually changed to ".if
 * !defined(NOMANCOMPRESS) MANCOMPRESSED=t".

As for my side, I've found six that will require MANNOTCOMPRESSED
(i.e., defines USE_IMAKE but doesn't install compressed manpages).
I'm fixing them as I see breaks, and the build is now running in
games/lang/net/x11.  (It's going roughly alphabetically but I started
x11 first because it's so big.)

 * (*) Not a random sampling---all 33 were not previously using the
 *     MANx variables, and all 33 installed only one manpage.  I
 *     suspect that the remaining ~200 ports make much less use of
 *     NOMANCOMPRESS (in their virgin sources, at least).

Maybe it's not worth it, but what do you think about the following?

MANCOMPRESSED: manpage is compressed (implied by USE_IMAKE)
MANNOTCOMPRESSED: manpage is not compressed even though USE_IMAKE is defined
MANCANBECOMPRESSED: honors NOMANCOMPRESS

Eek.  This is getting to long. ;)

 * [Incidentally, tip-offs that the port may honour NOMANCOMPRESS are any
 *  occurences of "NOMANCOMPRESS", "bsd.prog.mk", and "bsd.man.mk" in patches/,
 *  scripts/, and the work/ (after ``make extract'')].

Maybe I'll later try do a full build with NOMANCOMPRESS set in
/etc/make.conf.  We'll see how many things break. ;)

Satoshi

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ports" in the body of the message