Date: 20 Mar 2000 10:20:45 -0800 From: asami@FreeBSD.org (Satoshi - Ports Wraith - Asami) To: obrien@FreeBSD.org Cc: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: ports/devel/ElectricFence Makefile ports/lang/Sather Makefile ports/print/a2ps43-letter Makefile ports/textproc/agrep Makefile ports/news/aub Makefile ports/shells/bash1 Makefile ports/lang/bc-gcc Makefile ports/print/bibcard Makefile ... Message-ID: <vqcu2i1ec5e.fsf@silvia.hip.berkeley.edu> In-Reply-To: "David O'Brien"'s message of "Mon, 20 Mar 2000 08:47:14 -0800" References: <200003200237.SAA95508@freefall.freebsd.org> <vqchfe2f0vg.fsf@silvia.hip.berkeley.edu> <20000320084714.A96452@dragon.nuxi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
* From: "David O'Brien" <obrien@FreeBSD.org>
* > This is part of the required header, as specified in the handbook. If
* > you want to remove it, you need to discuss and get consensus first.
* DO WHAT?!?
* We did discuss it, and there was consensus -- billf, kris, reg (did not
* object to), "Akinori -Aki- MUSHA" <knu@idaemons.org> (by not objecting
* and by saying we need real make variables), myself, and you.
I'm the one that suggested the VERSION variable. The following quote
(in "* |") is all mine.
* Below is your email from Date: 13 Feb 2000 18:51:50 -0800.
* Did I misunderstand this part:
* But you are right, we should probably remove the comment for obvious
* cases (which will be most). After the release, though.
Yes. Sorry if I wasn't clear, but it was not a decision, just a hint
on what might come in the future. The handbook and portlint still
clearly object to this. I certainly didn't mean people can go change
some subset of ports, that will make the ports tree inconsistent with
itself.
When a change like this is proposed, this is how we proceed:
(1) we discuss the change, propose modifications, review patches, etc.
(2) when concensus is reached (or it is hopelessly split and there has
to be a decision), an definite announcement will be made to tell
people of the change of policy
(3) we coordinate changes to the infrastructure (bsd.port.mk,
handbook, portlint, etc.)
(4) commit
We are still in (1), and you jumped to (4).
By the way, since you quoted the mail from the discussion I was
meaning to ressurect, I'm going to take this opportunity to continue
where we left off.
* | I'll be very interested in hearing how they changed. I've also
* | thought about adding a mandatory VERSION variable separate from
* | DISTNAME/PKGNAME, to have something like
* |
* | PORTNAME= foo
* | VERSION= 1.2
* |
* | These two variables will be mandatory, and DISTNAME and PKGNAME will
* | be defined in bsd.port.mk as ${PORTNAME}-${VERSION}. (The PKGNAME
* | will not be overridable; it will be an error for a port to define it.)
* |
* | At least this will help ensure that the ${PKGNAME} will have correct
* | formats (well, you can always put a "-" in VERSION, but that's far
* | easier to find than going through a list of packages/All or
* | packages/Latest).
Knu-san, any progress on this? (You said you'll look into NetBSD's
changes.) From my customary glance, they don't seem to have anything
like this, but they just removed the "version required" comment (in
fact, they removed the entire header section other than $NetBSD$).
Their "update" support seems interesting, and with a new VERSION
variable, and some modifications to pkg_* tools, we might be able to
do a more seamless upgrade (emacs-20.5a -> emacs-20.6, etc.).
Satoshi
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?vqcu2i1ec5e.fsf>