From owner-cvs-all Mon Mar 16 20:53:04 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA19485 for cvs-all-outgoing; Mon, 16 Mar 1998 20:53:04 -0800 (PST) (envelope-from owner-cvs-all@FreeBSD.ORG) Received: from bubble.didi.com ([133.11.9.67]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA19439; Mon, 16 Mar 1998 20:52:52 -0800 (PST) (envelope-from asami@vader.cs.berkeley.edu) Received: (from asami@localhost) by bubble.didi.com (8.8.7/8.8.8) id NAA03454; Tue, 17 Mar 1998 13:20:03 +0900 (JST) (envelope-from asami) Date: Tue, 17 Mar 1998 13:20:03 +0900 (JST) Message-Id: <199803170420.NAA03454@bubble.didi.com> To: obrien@NUXI.com CC: chuckr@glue.umd.edu, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, cvs-ports@FreeBSD.ORG In-reply-to: <19980316121213.28949@nuxi.com> (obrien@NUXI.com) Subject: Re: cvs commit: ports/print/c2ps Makefile From: asami@FreeBSD.ORG (Satoshi Asami) Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk * WRT building both A4 and US-letter packages of all the print utilities, * did we come to a conclusion? Well, as far as I can tell, I had the final word in the previous discussion. (Actually it was more like I came back from a 1-day vacation, saw a bunch of messages, posted my opinion, and that was the end of it. But anyway.) Here's that post for your reference. Any comments? Satoshi ------- Date: Sun, 1 Mar 1998 01:01:49 -0800 (PST) To: hoek@hwcn.org CC: chuckr@glue.umd.edu, obrien@NUXI.com, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, cvs-ports@FreeBSD.ORG Subject: Re: cvs commit: ports/print/c2ps Makefile From: asami@cs.berkeley.edu (Satoshi Asami) * If we add ports such as a2ps-letter, etc. being wholly * self-sufficient (ie. don't call `a2ps') then most of these * problems solve themselves. Cookies aren't needed since * a2ps-letter can only do -letter. INDEX isn't a problem. * Packages are built properly. If we lose the single "a2ps" port, * then the whole PAPERSIZE question becomes unnecessary. : (pretty good summary deleted) I think I give up on the "DTRT with PAPERSIZE if the user types `make' in ports/print" idea. People doing that can easily use DUDS or some other method. It's getting too confusing. On the other hand, I really don't like the idea of having every single port be self-sufficient. We've already got too many "oops" commits, and things like having someone update one pkg/PLIST but not its counterpart is extremely hard to check. We should accept the fact that downloading a single port has a very good chance of breaking (think *_DEPENDS) and recommend users keep up with the whole ports tree. I can add something like this to bsd.port.mk to help unconfuse the users: === .if defined(INCLUDE_MAKEFILE_FROM) .if !exists(${INCLUDE_MAKEFILE_FROM}/Makefile) IGNORE= "This port requires another port (${INCLUDE_MAKEFILE_FROM}). Please get that one too" .else .include (${INCLUDE_MAKEFILE_FROM}) .if ${REQUIRED_MAKEFILE_VERSION} != ${MAKEFILE_VERSION} IGNORE= "Version of ${INCLUDE_MAKEFILE_FROM} disagrees with required version. Please get an updated version first" .endif .endif .endif === Then we have something like: a2ps-letter/Makefile: === REQUIRED_MAKEFILE_VERSION= 1.3 INCLUDE_MAKEFILE_FROM= ../a2ps-a4 .include === a2ps-a4/Makefile: === MAKEFILE_VERSION= 1.3 === Satoshi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message