I guess asking on the devel mailing list rather than sending mail to a particular individual means that you get the benefit of whatever global wisdom is available.
All packages are built nightly, in both release and devel, and across three architectures. The build system is more complicated than building individual packages -- reverse dependencies need to be built, too, and successful builds synchronized with public repositories -- so the nightly builds will not change in the near future. The problem is likely that your local configuration is using the current version of R and Bioconductor, whereas as a developer you should be using the version of R and Bioconductor used in the 'devel' branch. The version of R in use on the build machines is given at the top of the build page http://bioconductor.org/checkResults/3.3/bioc-LATEST/index.html so R-3.3.0-beta, available from cran.r-project.org. Once you have the correct version of R installed, then source("https://bioconductor.org/biocLite.R";) BiocInstaller::useDevel() sets you up to install bioc devel packages (the second step is not always necessary, depending on the version of R and Bioconductor; it is not necessary in the current 'devel', so the error -- 'devel' version already in use -- can be ignored). See a variant of these instructions at http://bioconductor.org/developers/how-to/useDevel/. The specific commands used for building are given in each section of the build report, so the specific build command (editing paths, of course) can also be invoked. I guess you are talking about MultiDataSet, with the build report at http://bioconductor.org/checkResults/3.3/bioc-LATEST/MultiDataSet/zin2-buildsrc.html The error ** R ** preparing package for lazy loading Error : object ‘rowRanges’ is not exported by 'namespace:GenomicRanges' ERROR: lazy loading failed for package ‘MultiDataSet’ * removing ‘/tmp/RtmpzWv6vq/Rinst36c3606d6b5a/MultiDataSet’ is because in bioc devel rowRanges is now in the SummarizedExperiment package. I see from SVN that you have discovered this, too. The warning * installing *source* package 'MultiDataSet' ... Warning in .write_description(db, file.path(outDir, "DESCRIPTION")) : Unknown encoding with non-ASCII data: converting to ASCII is probably complaining that the DESCRIPTION file does not declare a particular encoding (so defaults to ASCII) but contains non-ASCII characters. These can be spotted with > tools::showNonASCIIfile("DESCRIPTION") 8: person("Juan R.", "Gonz<c3><a1>lez", ,"jrgonza...@creal.cat", role = "aut")) and you could either forgo the accents or declare the encoding, probably by adding the tag Encoding: UTF-8 to the DESCRIPTION file. People will also point to travis for use with github, but for me this introduces two tools that play only sort-of ok with the Bioconductor build infrastructure (trying to bridge git via github with svn in particular seems to cause problems that can be very tedious and time consuming to solve). Martin ________________________________________ From: Bioc-devel <bioc-devel-boun...@r-project.org> on behalf of Ruiz Arenas, Carlos <cr...@creal.cat> Sent: Friday, April 22, 2016 6:06 AM To: bioc-devel@r-project.org Subject: [Bioc-devel] Custom build of packages Hi Bioconductor community, I am a recent Bioc developer. I uploaded a package just before the deadline of this BioC 3.3. During the admission process, the build machine was run with my packages almost every two hours which allowed me to improve the package more efficiently. However, now that I am in the devel trunk, the build machine is only run nightly. I am experiencing some problems due to some changes and I have to wait until the following day to check that the changes had worked. One of the people of Bioc tried to help me running the machine with my package. However, he is in Seatle while I am in Barcelona, so when I receive the mail he is not working anymore and the other way round. I am wondering if it could be possible for developers to generate a build report for our package. Otherwise, I think it could be interesting that the build machine was run more often during the last days of Bioc release, at least for packages with changes. Are any of these options feasible? Thanks, Carlos Ruiz _______________________________________________ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel This email message may contain legally privileged and/or confidential information. If you are not the intended recipient(s), or the employee or agent responsible for the delivery of this message to the intended recipient(s), you are hereby notified that any disclosure, copying, distribution, or use of this email message is prohibited. If you have received this message in error, please notify the sender immediately by e-mail and delete this email message from your computer. Thank you. _______________________________________________ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel
