svn commit: r495383 - in head/java: . wildfly16 wildfly16/files

[Alexey Dokuchaev]

On Thu, Mar 14, 2019 at 04:38:24PM +0000, Alexey Dokuchaev wrote:
> On Thu, Mar 14, 2019 at 05:29:40PM +0100, Torsten Zuehlsdorff wrote:
> > On 14.03.19 17:19, Alexey Dokuchaev wrote:
> > > On Wed, Mar 13, 2019 at 06:17:19PM +0100, Kurt Jaeger wrote:
> > >>>> Log:
> > >>>>   New port: java/wildfly16
> > >> [...]
> > >>> Its name suggests it should've been repocopied from one of the earlier
> > >>> versions but it was not, why is that?
> > >>
> > >> I missed it, that is common for arkane rules.
> > > 
> > > There's nothing arcane about repocopies Kurt.  I'm honestly surprised
> > > why people make this mistake again and again.  When you resurrect a
> > > port you make a repocopy.  When you spin-off a new branch based on a
> > > previous version you make a repocopy [...]
> > > 
> > > Upstream renames their software, you make a repocopy, etc.
> > 
> > A rename should be a "svn mv" instead of "svn cp"? Or am i mistaken?
> 
> Well yeah, I didn't go to the very details.  However, svn move is
> equivalent to an svn copy followed by svn delete.

To make myself 100% clear: repocopy != always "svn cp", it's about
copying of the previous history.  Naturally, for software renames, or
when a port is moved from one category to another, one would usually
use "svn mv" (which still implies "svn cp" under the hood).

./danfe