Re: [MoM] r-cran-nnls: please fix copyright
Hi Alba,
On Thu, May 21, 2015 at 04:39:28PM +0100, Alba Crespi wrote:
> > ERROR: dependencies 'ape', 'quadprog', 'igraph', 'Matrix', 'nnls' are not
> > available for package 'phangorn'
> >
>
> > which you stumble upon when building in a chroot. We know that ipraph
> > and nnls are in the new queue but please adjust the Build-Depends and
> > Depends of the package to be ready once the packages are available.
> >
>
> As you say igraph and nnls are new but the others are already packages on
> debian:
>
> https://packages.debian.org/stable/gnu-r/r-cran-ape (manteined by Debia-Med
> Team: http://anonscm.debian.org/cgit/debian-med/r-cran-ape.git/)
> https://packages.debian.org/stable/gnu-r/r-cran-quadprog
> https://packages.debian.org/stable/gnu-r/r-cran-matrix
>
> I've built the package by installing the other dependencies not yet in
> stable debian and it worked ok (using debuild). So I'm a bit confused.
You need to specify these packages as Build-Depends and Depends.
> > Moreover there is a related publication for this package which is
> > mentioned in inst/CITATION. You even mention this in the long
> > description but this is not the correct place to do this since it is not
> > machine parseable. Please instead create a file
> >
> > debian/upstream/metadata
> >
> I forgot about it. I've done it now.
Fine. But please also remove the now redundant information from the
long description?
> > Finally I removed debian/files since it is an auto generated
> > file and should not be in Git. Please pull.
>
> Ok, thanks.
> I've done the suggested changes. I hope now looks better. Let me know if
> more changes needs to be done.
The (Build-)Depends are evident for a succesful build in a pbuilder
environment. Please ask if you do not understand what this means.
Installing packages on your local machine is by no means sufficient.
> > Thanks for your work so far and do not hesitate to commit more. ;-)
>
> Sure. I'm working now on building another package from my own code (a tool
> to analyse protein sequences). However, this may take a while since it
> doesn't have any installation script yet. That's why I started with those
> packages first as my code depends on them.
Fine. We can't wait to see you packaging this. ;-)
> > BTW, if you (or somebody else) feel(s) the obligation on updating some
> > of our R packages - everybody in the team is more than welcome to push
> > updates to Git and I'll simply upload them. :-)
>
> Sure!
This would be great (also if others do as well!)
Kind regards
Andreas.
--
http://fam-tille.de
Reply to: