Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 04 Sep 2005 16:06:19 -0500
From:      "Jeremy Messenger" <mezz7@cox.net>
To:        "Jean-Yves Lefort" <jylefort@freebsd.org>
Cc:        cvs-ports@freebsd.org, Alexander Leidinger <Alexander@leidinger.net>, alejandro@varnet.biz, cvs-all@freebsd.org, ports-committers@freebsd.org
Subject:   Re: cvs commit: ports/games Makefile ports/games/linux-ut2003-demo Makefile distinfo pkg-descr ports/games/linux-ut2003-demo/files ut2003-demo.in
Message-ID:  <op.swljktc29aq2h7@mezz.mezzweb.com>
In-Reply-To: <20050904222210.15eaa37f.jylefort@FreeBSD.org>
References:  <200509032046.j83KkpQL082427@repoman.freebsd.org> <20050904124420.545395cb@Magellan.Leidinger.net> <20050904171849.21d7613c.jylefort@FreeBSD.org> <20050904185122.29cc5e11@Magellan.Leidinger.net> <20050904204037.04bda915.jylefort@FreeBSD.org> <20050904215045.1fa907f5@Magellan.Leidinger.net> <20050904222210.15eaa37f.jylefort@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 04 Sep 2005 15:22:10 -0500, Jean-Yves Lefort  
<jylefort@FreeBSD.org> wrote:

> On Sun, 4 Sep 2005 21:50:45 +0200
> Alexander Leidinger <Alexander@Leidinger.net> wrote:
>
>> On Sun, 4 Sep 2005 20:40:37 +0200
>> Jean-Yves Lefort <jylefort@FreeBSD.org> wrote:
>>
>> > On Sun, 4 Sep 2005 18:51:22 +0200
>> > Alexander Leidinger <Alexander@Leidinger.net> wrote:
>> >
>> > > On Sun, 4 Sep 2005 17:18:49 +0200
>> > > Jean-Yves Lefort <jylefort@FreeBSD.org> wrote:
>> > >
>> > > > > >   Log:
>> > > > > >   Add linux-ut2003-demo.
>> > > > >
>> > > > > Why is it necessary to generate the plist at install time? As  
>> far as I
>> > > > > can see the port installs a fixed set of files and doesn't do  
>> any magic
>> > > > > to prevent the installation of files depending on a complex set  
>> of
>> > > > > options or the state of the moon.
>> > > >
>> > > > It reduces the size of the port by about 40kb, and eases future  
>> updates.
>> > >
>> > > We had a discussion on ports@ a while ago: As long as there are no  
>> very
>> > > urgent reasons to use a install-time generated plist a maintainer
>> > > should (as in: we point with fingers on you if you don't do it) use  
>> a
>> > > static plist since it is more beneficial for most people.
>> >
>> > The pkg-plist/PLIST_*/MAN*/PORTDOCS/... set is not meant to be
>> > human-readable. See x11/nvidia-driver/pkg-plist for an example.
>>
>> They are human-readable ATM and a lot of procedures rely on this. There
>> are always exceptions, but the ut demo ports don't need to be
>> an exception. Please reread the thread I mentioned before you come up
>> with other arguments since it's beaten to death already.
>
> I disagree with your interpretation of human-readability.
>
> The output of pkg_info -L is human-readable.

Isn't 'pkg_info -L' only can be seen if you installed port? What about  
non-install? I always support static plist by 100%.

Cheers,
Mezz

> A mixture of @exec,
> PLIST_SUB, PLIST_FILES, PLIST_DIRS, MANn, MANLANG, PORTDOCS,
> USE_RC_SUBR, GCONF_SCHEMAS (and so on, ad nauseam) is not.
>
>> [suggestions about new metadata handling]
>>
>> This is out of the scope of what I want to discuss, feel free to start
>> a new thread with a meaningful subject.
>
> No, my suggestions address all the concerns (minus the portlint one)
> you raised in your January post to ports@.


-- 
mezz7@cox.net  -  mezz@FreeBSD.org
FreeBSD GNOME Team
http://www.FreeBSD.org/gnome/  -  gnome@FreeBSD.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?op.swljktc29aq2h7>