Re: [gentoo-user] rng-tools

[Alan McKinnon <alan.mckinnon@gmail.com>]

On Sat, 3 Mar 2012 22:04:29 +0700
Pandu Poluan <pandu@poluan.info> wrote:

> On Mar 3, 2012 9:05 PM, "Alan McKinnon" <alan.mckinnon@gmail.com>
> wrote:
> >
> > On Sat, 3 Mar 2012 08:48:59 -0500
> > Michael Mol <mikemol@gmail.com> wrote:
> >
> > > So I've been making extensive use of rngd on one of my Debian
> > > servers, and I wanted to make use of it on a couple of my Gentoo
> > > boxes. Only to find out that two parameters I need, -T and -R,
> > > aren't available, even when I unmask version '3' in portage. Even
> > > the manpage contains 'FIXME' where the Debian manpage contains
> > > much more information.
> > >
> > > The version of the rng-tools Debian package I'm using is
> > > '2-unofficial-mt.14-1~60squeeze1'. Are the -T and -R parameters
> > > unique to Debian, or is the Gentoo package simply out of date?
> > >
> >
> > You should download the .deb and see if it contains custom patches
> > (rather likely actually). If you want those patches applied, stick
> > them in /etc/portage/patches/ in the proper place.
> >
> > I can never remember how all that stuff works, I have to Google it
> > each time :-)   Start here if you haven't heard of user-custom patch
> > support in portage yet:
> >
> > http://forums.gentoo.org/viewtopic-t-791464.html
> >
> http://klaig.blogspot.com/2011/07/how-to-apply-user-patches-in-portage.html
> >
> >
> 
> Remember that it depends on whether the ebuild for the package
> support user patches or not.

That's true enough. epatch_user() must be called for it to work.

However, it is defined in base.eclass and part of EAPI-2. How many
ebuilds does that exclude these days?

I did a quick back-of-the-envelope check:

# find /var/portage/ -type f -name "*.ebuild" | wc
  29846   29846 1730549
# grep EAPI=[234] /var/portage/*-*/*/*.ebuild | wc
  11138   11273  719796
# grep EAPI=1 /var/portage/*-*/*/*.ebuild | wc
    336     341   21938
# grep EAPI=0 /var/portage/*-*/*/*.ebuild | wc
      5       5     275

37.3 coverage of >=EAPI-2. Wow. I really thought it would be more.

-- 
Alan McKinnnon
alan.mckinnon@gmail.com