[Bioc-devel] Recreating nebbiolo1's build environment

sergi sayols puig @erg|@@yo|@pu|g @end|ng |rom gm@||@com
Tue Jan 11 11:21:53 CET 2022


Dear all,
Our package rrvgo has recently started to fail when building for BioC
3.15 (devel branch).

I'm trying to replicate the build environment as in nebbiolo1 in order
to reproduce the issue locally, with little success so far. I'm
following carefully the instructions provided in the build report, but
I must be missing something because I can build the package
successfully.

I tried both compiling R from source (snapshot
R-devel_2022-01-05_r81451.tar.gz) + installing the devel version of
Bioconductor, and also using the docker image of
bioconductor_docker:devel. Here I provide the steps I followed to
build from within the container, using vanilla .libPaths():

* pull docker image
    $ docker pull bioconductor/bioconductor_docker:devel
* open shell in the running container:
    $ docker exec -it bioconductor/bioconductor_docker:devel /bin/bash
* install package + dependencies (within the container):
    # R
    R> install.packages("BiocManager")
    R> BiocManager::install(version='devel')
    R> BiocManager::install(c("rrvgo", "org.Hs.eg.db"))   # the second
is required in the vignette
    R> q()
* clone the devel branch of the package:
    # git clone https://git.bioconductor.org/packages/rrvgo
* get nebbiolo1's .Renviron:
    # wget -qO-
http://bioconductor.org/checkResults/devel/bioc-LATEST/Renviron.bioc >
Renviron.bioc
* and build as described in the build report:
    # R_ENVIRON_USER=Renviron.bioc R CMD build --keep-empty-dirs
--no-resave-data rrvgo

This will successfully build the tarball of the package, without complains.
Does anyone have any idea what I could possibly do to troubleshoot the error?

Thanks a lot in advance!

Best wishes,
Sergi



More information about the Bioc-devel mailing list