Hi Michael, Thanks for taking the time to write a bugreport.
I will try to explain why I am closing this bugreport below. On 13 November 2010 at 13:41, Michael Shuler wrote: | Package: r-cran-rcmdr | Version: 1.6-0-2 | Severity: wishlist | | On first start of Rcmdr after installation, I was presented with a notification | box with the following: | | The following packages used by Rcmdr are missing: | leaps, Hmisc, aplpack | Without these packages, some features will not be available. | Install these packages? | [Yes] [No] | | I selected no, as I am usually not fond of pulling in libraries outside of the | system package manager. | | r-cran-hmisc is available in the Debian repository and a manual install of this | package removes Hmisc from the warning notification, so a recommends (or | depends) for r-cran-hmisc should be included. Added in my package sources. The next r-cran-rcmdr will have that. | apt-cache searches for 'leaps' and 'aplpack' don't return any results, so it | would be nice for these libraries to be packaged for Debian, if they are not | included in some other R library - if they are, please recommend/depend on | those packages as well. I you find a maintainer, maybe. If you look at http://qa.debian.org/developer.php?login=edd&comaint=yes you will see that my package count is now at 128. Quite literally a dozen or so of these packages were added because Rcmdr at some point or other dependent upon them. Rcmdr changed over the years from 'hard' Depends on many package to much softer run-time detection which is better. What Rcmdr upstream now does is much better: Depends: R (>= 2.6.0), tcltk, grDevices, utils, car (>= 2.0-0), Suggests: abind, aplpack, colorspace, effects (>= 1.0-7), foreign, grid, Hmisc, lattice, leaps, lmtest, MASS, mgcv, multcomp (>= 0.991-2), nlme, nnet, relimp, rgl, RODBC Only car [r-cran-car] is needed as a Depends, the rest depends on how you use it. There is simply no way to manually maintain all the possible depends for core packages of the 2500+ CRAN packages. I am involved with the cran2deb effort so one day we may have full .deb repos. In the meantime, do what everybody else and install missing packages in /usr/local/lib/R/site-library. Or become a DM and package the missing one. It is not hard. [ That said, I am not always happy with fellow DDs/DMs who adopt some r-cran-* package and then let them become stale. For that, I actually prefer /usr/local/lib/R/site-library and a daily call to ~/bin/update.r (from the littler package examples/) ]. On 13 November 2010 at 13:56, Michael Shuler wrote: | I see there is an ITP for r-cran-aplpack - #588924 Sure, but there is no firm relationship between ITP and packages actually appearing. Dirk | | -- | Kind regards, | Michael | | -- System Information: | Debian Release: squeeze/sid | APT prefers testing | APT policy: (700, 'testing'), (300, 'unstable'), (1, 'experimental') | Architecture: i386 (i686) | | Kernel: Linux 2.6.32-5-686-bigmem (SMP w/4 CPU cores) | Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) | Shell: /bin/sh linked to /bin/dash | | Versions of packages r-cran-rcmdr depends on: | ii libc6 2.11.2-7 Embedded GNU C Library: Shared lib | ii r-base-core 2.11.1-6 GNU R core of statistical computat | ii r-cran-abind 1.1.0-4 GNU R abind multi-dimensional arra | ii r-cran-car 2.0-2-1 GNU R Companion to Applied Regress | ii r-cran-effects 2.0.10-1 GNU R graphical and tabular effect | ii r-cran-lmtest 0.9.27-1 GNU R package for diagnostic check | ii r-cran-mgcv 1.6-2-1 GNU R package for multiple paramet | ii r-cran-multcomp 1.1-7-1 GNU R package for multiple compari | ii r-cran-mvtnorm 0.9-92-1 GNU R package to compute multivari | ii r-cran-relimp 1.0-1-3 GNU R package for inference on rel | ii r-cran-rgl 0.91-1 GNU R package for three-dimensiona | ii r-cran-sm 2.2-4.1-1 GNU R package for kernel smoothing | ii r-cran-strucchange 1.4-1-1 GNU R package for structural chang | | r-cran-rcmdr recommends no packages. | | Versions of packages r-cran-rcmdr suggests: | pn r-cran-rodbc <none> (no description available) | | -- no debconf information | | -- Dirk Eddelbuettel | [email protected] | http://dirk.eddelbuettel.com -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
