Date: Wed, 18 Jun 1997 18:23:30 -0400 (EDT) From: Chuck Robey <chuckr@glue.umd.edu> To: Ruslan Shevchenko <rssh@cki.ipri.kiev.ua> Cc: FreeBSD Ports <FreeBSD-Ports@FreeBSD.ORG> Subject: Re: PLISTs Message-ID: <Pine.BSF.3.96.970618182125.305C-100000@Journey2.mat.net> In-Reply-To: <33A7E86A.716C@cki.ipri.kiev.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 18 Jun 1997, Ruslan Shevchenko wrote:
> Hm, why you can not simple write ${PKGDIR}/PLIST.new in
> Makefile and then compare it with old ?
Because NO other file in a port is ever modified by building or packaging
a port, and I'd like this to continue. Doing make clean just means
deleting the work subdirectory, and this can continue without
modification. Otherwise, amek clean would have to be modified to make
sure that no old, modified PLIST, COMMENT, or DESCR files were hanging
around.
>
> Chuck Robey wrote:
> >
> > I want to make another suggestion, dealing with packaging problems, for
> > options that make ports want to change the PLIST. This suggestion does
> > several different things, and I'll _try_ to make sense coherently here.
> >
> > I want to make changes so that during the early phase of a build (say, at
> > the end of the extract phase) everything in pkg gets copied over to a new
> > subdirectory, underneath work (work/pkg). There would be a new variable,
> > MOD, which would be set explicitly to "".
> >
> > During any subsequent part of the build, things in work/pkg could be
> > modified, based on defaults, or explicitly required inputs, or environment
> > variables. These changes would affect the build, the work/pkg/PLIST, and
> > the MOD variable.
> >
> > At the beginning of the package phase (if it goes that far), the original
> > PLIST would be compared to the work/pkg/PLIST. If they were identical,
> > MOD would remain as it was (normally empty, altho it could have been
> > modified also). If the new and old PLISTs differed, then one of two
> > things would happen:
> >
> > 1) if MOD was still empty, it would be set to "mod".
> > 2) if it wasn't empty, it wouldn't be changed.
> >
> > The name of the final package would be changed to stick the value of the
> > MOD variable before the "tar.gz".
> >
> > This would give a method to allow a port to change it's PLIST, and allow
> > such manufactured changes to be signalled in the final package name as the
> > port author would want it. The author could design in a port that builds
> > distinctly different, and the changes could easily be signalled in the
> > package name. As an example, MOD could be set to A4 or LTR for the a2ps
> > port, to signal that the package has the default paper size set to either
> > A4 or LTR. Changed stuff would install correctly, because the PLIST could
> > be made to reflect a new reality.
> >
> > The clean target would wipe out anything in work, just as it does now.
> >
> > I'd like to get an idea if this might be fould appealing. If folks like
> > this, I could make a diff that does it, or maybe include any changes that
> > folks like, or if you all think it's a dumb idea, ....
> >
> > ----------------------------+-----------------------------------------------
> > Chuck Robey | Interests include any kind of voice or data
> > chuckr@eng.umd.edu | communications topic, C programming, and Unix.
> > 213 Lakeside Drive Apt T-1 |
> > Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD
> > (301) 220-2114 | version 3.0 current -- and great FUN!
> > ----------------------------+-----------------------------------------------
>
>
----------------------------+-----------------------------------------------
Chuck Robey | Interests include any kind of voice or data
chuckr@eng.umd.edu | communications topic, C programming, and Unix.
213 Lakeside Drive Apt T-1 |
Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD
(301) 220-2114 | version 3.0 current -- and great FUN!
----------------------------+-----------------------------------------------
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.970618182125.305C-100000>