[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Backporting R



Am Montag, 25. April 2016, 23:02:22 schrieb Andreas Tille:
> Hi,
> 
> On Mon, Apr 25, 2016 at 07:39:23PM +0200, Johannes Ranke wrote:
> > >   1. I'll create
> > >   
> > >        https://git.debian.org/git/debian-science/packages/r-base.git
> 
> Done.

Very nice.

> > >   2. I inject the current packaging from testing/unstable to this
> > >   
> > >      repository and create debian/jessie-backports branch.

OK, I had a look and the only changes you made are the changelog and the Vcs 
fields in debian/control.

> > 
> > Dirk, would you consider pushing your future packaging to this repo? It
> > seems to me then future backporting from there would be easier as I could
> > merge the changes from (my) master (which is you in that case :).
> 
> I'd happily move the repository I created above to any other location on
> git.debian.org Dirk might prefer.
> 
> > >   3. I add Johannes Ranke to the members of Debian Science team to
> > >   
> > >      grant him commit permissions
> > > 
> > > Thus Johannes can commit the changes for Jessie backports there.

Thanks! I added my ssh key to alioth and was able to clone the repo via ssh.
 
> > Actually, no changes were necessary for the backports of the last few
> > releases to Jessie.
> 
> I've taken this as good reason to upload to backports.

OK, I will test it when it is in the archive.

> > > Any better suggestion is welcome.
> > 
> > Sounds excellent to me. I will have a look at the repo and consider
> > setting up repositories for backports of the recommended packages pulled
> > in by r- recommended, so these could be brought up to date from time to
> > time as well.
> Fine.

I have had a look and I am starting to understand git-buildpackage, which 
looks like a very convenient tool indeed.

Kind regards,

Johannes

> 
> > If everything works out, this could give me the chance to skip the manual
> > work required today for the backports on CRAN.


> 
> I'd be delighted if this could save some work.
> 
> Kind regards
> 
>       Andreas.


Reply to: