Bug#880875: please rename source package to r-cran-scatterplot3d

[Sébastien Villemot]

On Sun, Nov 05, 2017 at 12:26:12PM +0100, Philip Rinn wrote:
> Hi,
> 
> On 05.11.2017 at 10:54, Sébastien Villemot wrote:
> > scatterplot3d is the only CRAN package maintained in Debian Science Team whose
> > source package name does not begin with "r-cran".
> >
> > Please rename the source package to r-cran-scatterplot3d, to facilitate the
> > identification of the package when making searches on source package names, and
> > also for consistency.
> 
> is it really worth the hassle? Carrying a transitional package around, going
> through new again, etc.?

No need to carry a transitional package. Just rename the source (and the git
repository), and upload it. I think it will have to go through NEW, though, but
I'm not sure. Then you'll have to request the removal of the old source
package once the new one is accepted. This has already been done for quite a
few other CRAN packages.

> There are packages around outside Debian Science which
> doesn't use r-*-<PACKAGENAME> as source package name.

This is not true. Out of the 120 CRAN packages maintained by Debian Science,
scatterplot3d is the only one whose source name does not have the "r-cran-"
prefix. You can verify this with:

 aptitude search '?maintainer(debian-science-maintainers at lists.alioth.debian.org)?name(r-cran-.*)' -F "%p src:%e"

> To identify R packages you
> have to look for (Build-)Dependencies anyway as some not-only-R-packages build
> binary R packages ... and there is still the section "gnu-r" to identify them -
> well, sadly not really [1].

Indeed that could be another possible avenue for identification, though as you
said this does not always work.

Also, my concrete use case is that I periodically go through the Debian
Maintainer Dashboard of Debian Science [1] to identify R packages that need
some love, and I search for "r-cran" on that page. Obviously scatterplot3d
won't show up, because the list is based on source package names.

> I'd rather ask why don't the other R packages in Debian Sciences follow the
> recommendation in the "Debian R Policy"[2] section 2.1?

This document is 14 years old and not maintained. I don't think it has any
authoritative value. Established practice is in my opinion stronger.

Moreover, CRAN packages sometimes have very short names (3 or 4 letters), and
I'm pretty sure there would be name conflicts if we enforced that policy
(though I did not verify).

> Well, if you have strong feelings about renaming the package I'll do it with the
> next upstream version but I don't really see the point.

This is up to you, but I hope I convinced you.

Best,


[1] https://udd.debian.org/dmd/?debian-science-maintainers%40lists.alioth.debian.org#todo


-- 
⢀⣴⠾⠻⢶⣦⠀  Sébastien Villemot
⣾⠁⢠⠒⠀⣿⡁  Debian Developer
⢿⡄⠘⠷⠚⠋⠀  http://sebastien.villemot.name
⠈⠳⣄⠀⠀⠀⠀  http://www.debian.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/debian-science-maintainers/attachments/20171105/24a8095f/attachment.sig>