svn commit: r469188 - head/ports-mgmt/pkg_replace

[Alexey Dokuchaev]

On Sun, May 06, 2018 at 12:34:20PM +0200, Kurt Jaeger wrote:
> > On Sun, May 06, 2018 at 10:15:03AM +0000, Koichiro Iwao wrote:
> > > New Revision: 469188
> > > URL: https://svnweb.freebsd.org/changeset/ports/469188
> > > 
> > > Log:
> > >   ports-mgmt/pkg_replace: Update to 20180424
> > >   
> > >   - Cleanup code
> > >   - Speed up and strictly check package's dependencies
> > >   - While here, pet portlint by using pkg-plist instead of PLIST_FILES
> > 
> > Could you explain what did you achieve by moving three-line PLIST_FILES
> > into pkg-plist?  How can "pet portlint" be a reason for a change while
> > we generally encourage converting <6-line pkg-plist's into PLIST_FILES?
> 
> There was a problem if the pkg-list stayed in PLIST_FILES
> (the build in poudriere failed with a signal 6 in make package).
> 
> Koichiro is working to reproduce/report this, until then...

*facepalm*.  This poudriere-driven development is more and more disturbing.
You didn't attempt to understand the problem, you didn't attempt to fix the
PLIST_FILES (remove bogus %%ETCDIR%% since this is Makefile), you didn't
ask for a peer review.  meta@ pessimized the port, provided wrong commit
log (it tells *nothing* about poudriere failing with a signal 6 in make
package, it says "pet portlint"), and you've approved it like there's
nothing wrong with all of this. :-(

./danfe