svn commit: r403347 - head/security/keepassx2

[Cy Schubert]

In message <20151209144013.GA10737 at FreeBSD.org>, Alexey Dokuchaev writes:
> On Wed, Dec 09, 2015 at 06:15:23AM -0800, Cy Schubert wrote:
> > In message <20151209081342.GA41332 at FreeBSD.org>, Alexey Dokuchaev writes:
> > > On Tue, Dec 08, 2015 at 08:40:41PM +0000, Cy Schubert wrote:
> > > > New Revision: 403347
> > > > URL: https://svnweb.freebsd.org/changeset/ports/403347
> > > > 
> > > > Log:
> > > >   Update 2.0-beta2 --> 2.0
> > > > [...]
> > > >  
> > > >  PORTNAME=	keepassx
> > > > -DISTVERSION=	2.0-beta2
> > > > -PORTEPOCH=	1
> > > > +DISTVERSION=	2.0
> > > > +PORTEPOCH=	2
> > > 
> > > Why did you bump port epoch here?  DISTVERSION ensures that resulting
> > > PORTVERSION is correctly mangled, and used for -betas and -rcs because
> > > it allows to seamless update to .0 releases:
> > > 
> > >     $ make -V PORTVERSION DISTVERSION=2.0-beta2
> > >     2.0.b2
> > >     $ pkg version -t 2.0.b2 2.0
> > >     <
> > 
> > My mistake. I expected 2.0.b2 to sort after 2.0.
> 
> Understood; please please please guys, always check before bumping the epoch
> as it can never go back (unless you rename the package).
> 
> ./danfe (portepoch hater)
 
I'm not a big fan of portepoch either, which is why I'm always careful
to put alphas, betas, and other non-GA software into -devel ports.

This wasn't my port but to scratch an itch caused by an older release
I updated a port that IMO should have been -devel until it became GA.

I now understand that alphas, betas, and rcs are handled specially.
I for one will continue to make the distinction with ports I myself
maintain I think there's some value to distinguishing between GA and
non-GA software.

~Cy
(using nmh at the moment)