Re: pkgsrc-2013Q2 randomly not creating packages for me. What can be wrong?

[Joerg Sonnenberger <joerg%britannica.bec.de@localhost>]

On Tue, Jul 30, 2013 at 09:26:42AM +0900, OBATA Akio wrote:
> On Tue, 30 Jul 2013 08:28:49 +0900, Greg Troxel <gdt%ir.bbn.com@localhost> 
> wrote:
> 
> >
> >Thomas Klausner <wiz%NetBSD.org@localhost> writes:
> >
> >>On Tue, Jul 16, 2013 at 02:26:13PM +0200, Richard Palo wrote:
> >>>For info, I am experiencing this as well on current, for me it is
> >>>gcc48 and gcc48-libs (that I've noticed so far).
> >>
> >>Could this be related to this change?
> >>
> >>Module Name:        pkgsrc
> >>Committed By:       obache
> >>Date:               Thu May 23 13:18:56 UTC 2013
> >>
> >>Modified Files:
> >>    pkgsrc/mk: bsd.pkg.barrier.mk
> >>    pkgsrc/mk/install: bsd.install.mk replace.mk
> >>    pkgsrc/mk/package: package.mk
> >>    pkgsrc/mk/pkgformat/pkg: package.mk replace.mk
> >>
> >>Log Message:
> >>prevent to generate binary package in ${PACKAGES} from the side effect of
> >>`install' with USE_DESTDIR=yes.
> >>This changes prevent to unwanted overwite of existing binary packages with
> >>test installation (`stage-install', `replace' & `undo-replace', and so on).
> >>
> >>To do both `install' and `package', you can still use `package-install' 
> >>target,
> >>same as USE_DESTDIR=no.
> >
> >Have we resolved this?  "replace" with USE_DESTDIR=yes (default) used to
> >reliably generate packages.  While one can argue that packages shouldn't
> >be generated by default, it's a regression to change replace without
> >adding a package-replace target that guarantees both a binary package
> >and replace.  At least from the viewpoint that having binary packages
> >arund is good, which I think is > 50% of replace users, perhaps 90%.
> 
> binary packages is useful to replace installed package,
> but not required anymore after the replace.#

I disagree, but I find keeping packages around instead of pkg_tarup a
lot more useful and people insisted on keeping the latter.

Joerg