bioconductor rebuild todo list

[bgruening]

The following are packages that either need to be created or updated to have the gcc7 label for the bioconductor 3.8 rebuild to finish. Please note that packages in bioconda also need to have the gcc7 label attached to them. The only way to do this currently is by building them in the bulk branch, so please direct any PRs there. Help either adding recipes to the bulk branch or on conda-forge would be greatly appreciated.

Missing packages

• recipes/bioconductor-hilbertvisgui needs gtkmm
• libsbml package is missing

R packages that need to be rebuilt

• r-animation (depends on r-magick)
• r-clustersim Add r-clustersim and r-earth conda-forge/staged-recipes#7567
• r-dagitty (depends on r-v8 and, this, libv8)
• r-evmix (needs an r-gsl rebuild)
• r-feature (needs an r-rgl rebuild)
• r-hiver (needs an r-rgl rebuild)
• r-magick (needs an imagemagick rebuild)
• r-pca3d (needs an r-rgl rebuild)
• r-rcdklibs add new R recipes conda-forge/staged-recipes#7243
• r-v8

[dpryan79]

Thanks @bgruening, this will help a lot!

[federicomarini]

Hey @bgruening I'll be happy to try to throw in some time before Christmas comes, provided no colleagues go on rampage-submission mode in the last days 😨

[jenzopr]

Hi Björn,
for the packages intended to go to bioconda (r-gap, r-grain ,r-pcalg), the following dependencies are missing as well: r-grbase
Best,
Jens

[brainstorm]

As mentioned in #12732 (comment), r-bezier would be nice to have, if conda-forge/staged-recipes#7342 doesn't get it first.

[bgruening]

@bioconda/r any help appreciated.

[bgruening]

@nturaga do you have any insight in how often the bioc data packages are updated. They take a large amount of time and I'm wondering if we need to pin them to a specific R version. How many of them contain R version specific code. E.g. this one https://bioconductor.org/packages/release/data/annotation/html/Homo.sapiens.html could be probably unpinned to any R version, isn't it?

[nturaga]

Hi @bgruening, There are a few packages which are built with each new release, but there are others that are updated and built each time there is an update. Some packages include code to show how the data was gathered, cleaned ( or generated), and this could be R version dependent. I would just pin them to the latest release version of R-3.5.1, since it works well with Bioconductor version 3.8. But I believe there is no need to build them every few days or anything, just once a release cycle is fine.

…

On Sat, Dec 29, 2018 at 2:31 AM Björn Grüning ***@***.***> wrote: @nturaga <https://github.com/nturaga> do you have any insight in how often the bioc data packages are updated. They take a large amount of time and I'm wondering if we need to pin them to a specific R version. How many of them contain R version specific code. E.g. this one https://bioconductor.org/packages/release/data/annotation/html/Homo.sapiens.html could be probably unpinned to any R version, isn't it? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#12586 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACnoS1-v76-xJaHgcteMQkMUGrYDm5-6ks5u9xpugaJpZM4ZPLSb> .

[bgruening]

@nturaga I would like to avoid rebuilding all data packages with every bioc release. it takes a huge amount of space on the Docker side and if it does not change its useless.

[dpryan79]

We're mostly down to packages needing r-gsl and r-rgl.

[dpryan79]

Closing since this is effectively done (and has been for a while).