Hello Dirk,
Thank you for the advice. Currently I get a bug whenever I try to use R Builder on my computer, no matter which browser I launch it from, but I will try the Rocker images, as I have worked with Docker before and have some familiarity. Gratefully, Julia ________________________________ From: Dirk Eddelbuettel <dirk.eddelbuet...@gmail.com> on behalf of Dirk Eddelbuettel <e...@debian.org> Sent: Monday, May 28, 2018 6:55:25 PM To: Julia A. Pilowsky Cc: r-package-devel@r-project.org Subject: Re: [R-pkg-devel] RcppArmadillo-based package works in every tested OS except Debian Hi Julia, On 28 May 2018 at 16:33, Julia A. Pilowsky wrote: | I am the maintainer of the colorednoise<https://github.com/japilo/colorednoise> package, which employs RcppArmadillo. The package passes all checks on my local OS X install, on win_builder, and on Travis CI's Ubuntu distribution<https://travis-ci.org/japilo/colorednoise/builds/384377138>. | | However, when I submitted the package to CRAN, I got an error message<https://win-builder.r-project.org/incoming_pretest/colorednoise_1.0.2_20180527_174354/Debian/> for Debian and only Debian � all the other systems appear to run it fine. The error is for the unit test the package has for a RcppArmadillo-based function � all the other unit tests are for R or Rcpp functions. | | I am really scratching my head over why this RcppArmadillo function is working on every OS except Debian, and I'm at a loss as to how to fix it, since I don't have a Debian install on my computer, and Travis CI only tests in Ubuntu. Any advice would be greatly appreciated. 1) You get easy access to Debian systems via RHub, either using the web interface at https://builder.r-hub.io or via the CRAN package rhub. A highly recommended service. 2) Another possibility is Docker. It is pretty easy to set up, and I often use Docker instances to test things I do not want more permanently on my system(s). Also, our Rocker images happen to already use Debian and give you R. Maybe start with our recent intro paper to learn more: https://journal.r-project.org/archive/2017/RJ-2017-065/index.html 3) As Rcpp maintainer, I happen to run a lot of reverse dependency checks for Rcpp, RcppArmadillo, ... and yes, colorednoise also failed for me on the last three Rcpp reverse depends checks. I log these in a GitHub repo at https://github.com/RcppCore/rcpp-logs -- that won't give you details but it simply re-affirms the CRAN failures. Hth, Dirk -- http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org [[alternative HTML version deleted]] ______________________________________________ R-package-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-package-devel
