[R-sig-Debian] R 2.5.0 packages for Debian stable/etch

[Deepayan Sarkar]

On 5/14/07, Johannes Ranke <jranke at uni-bremen.de> wrote:
> Hi Ivailo, Dirk and Deepayan,
>
> (copying in Deepayan Sarkar, the author of rcompgen)
>
> * Dirk Eddelbuettel <edd at debian.org> [070515 00:00]:
> >
> > On 14 May 2007 at 21:42, Ivailo Stoyanov wrote:
> > | Johannes Ranke <jranke <at> uni-bremen.de> writes:
> > | >
> > | > Dear list,
> > | >
> > | > R 2.5.0 packages for Debian stable (i386 and amd64) are now available
> > | > from CRAN.
> > | >
> > | > The new recommended codetools and rcompgen packages haven't completely
> > | > made it to Debian yet, but can of course be installed in the standard
> > | > non-Debian ways.
> > | >
> > | > Please consult the README file in the Debian directory, and report
> > | > problems to me directly or to this list.
> > |
> > | Dear Johannes,
> > |
> > | your first R-build (without codetools and rcompgen) uploaded at CRAN
> > | worked just fine, and when I discovered that you have provided another
> > | build including both above-mentioned packages, I have upgraded flawlessly.
> > | However, I have noticed that rcompgen installed under
> > | /usr/lib/R/site-library, unlike all the other recommended packages.
> > |
> > | Is this intended behaviour or a bug in the backport-packages?
> >
> > Can't speak for Johannes, but I recently altered debian/rules such that this
> > location should be automatically determined based on whether or not
> > 'Priority: Recommended' is found in the upstream DESCRIPTION file.  So it may
> > also just be inconsistency at the CRAN end...
>
> Yes, I didn't touch that mechanism. I just saw that this line is indeed
> missing from the upstream DESCRIPTION file. I believe this is not on
> purpose.

Yes, my bad. I removed it at one point to install on 2.4.0, I must
have forgotten to put it back in.

-Deepayan