Please update 10 CRAN packages and 1 Bioconductor package
aitap opened this issue · comments
Thank you for your work on packaging so many R packages for Ubuntu! As you said in your e-mail, I'm reporting a few that are currently a bit behind CRAN and Bioconductor:
setRepositories(ind=1:4) # CRAN & BioC
subset(
merge(
available.packages(), installed.packages(),
by = 'Package', suffixes=c('.available', '.installed')
),
Version.available > Version.installed
)[,c('Package','Repository')]
# Package Repository
# 24 antiword https://cloud.r-project.org/src/contrib
# 154 coda https://cloud.r-project.org/src/contrib
# 163 commonmark https://cloud.r-project.org/src/contrib
# 197 data.table https://cloud.r-project.org/src/contrib
# 237 distrEx https://cloud.r-project.org/src/contrib
# 264 edgeR https://bioconductor.org/packages/3.18/bioc/src/contrib
# 511 igraph https://cloud.r-project.org/src/contrib
# 612 littler https://cloud.r-project.org/src/contrib
# 697 MplusAutomation https://cloud.r-project.org/src/contrib
# 1145 sp https://cloud.r-project.org/src/contrib
# 1225 tidygraph https://cloud.r-project.org/src/contrib
I can see r-cran-littler
being available in https://cloud.r-project.org/bin/linux/ubuntu jammy-cran40/
, but the one in https://r2u.stat.illinois.edu/ubuntu jammy/main
has a larger APT priority on the docker.io/rocker/r2u
container.
Please see CRANberries (https://dirk.eddelbuettel.com/cranberries/, or via RSS, or Mastodon toots, or Bsky skeets) -- most of these came to CRAN yesterday as I can tell from eyeballing: antiword
, commonmark
, data.table
, igraph
.
littler
is special, I am the author, it comes borked from p3m so we do it apart. It should still be working and current, and the fact that Michael also kindly covers it at jammy-cran40 is irrelevant. That is precisely why we pin the r2u repo at a higher priority.
So I am going to close this. When you claim "oh bad bad r2u these packages are behind" also included their CRAN release date. That is in either available.packages()
or utils::CRAN_package_db()
and I of course know that you are able to do an inner join.
I appreciate the effort in tryying to help and improve r2y but this fell short.
And to be plain: You cannot expect an hourly cadence from us as r-universe does. We give you over 22k packages, working, with full dependencies, for two LTS releases -- but updates need a day or two (or a weekend) or longer if p3m is behind. Such is life. You can always add packages from source if you need them immediately. I may add this to the FAQ.
I triggered a BioC update: basililisk
+ edgeR
+ tximeta
updated and are now available for focal and jammy.
Sorry, I should have stopped to think and check the dates.
No worries -- as I also lived off source on Linux for 25+ year I am also in the habit of expected new packages "immediately".
We may want to build ourselves a little package lister / checker with data. Inaki (in his bspm
) and I (in my dang
) wrote something for 'shadowed packages' (to identify when bspm
-installed binaries in /usr/lib/R/site-library
are "shadowed" by possible older packages in /usr/local/lib/R/site-library
. And I play a little with dates in eg issue 391 for RcppArmadillo which managed the long transition. So we could do a local package checker that lists 'newer available at CRAN' and flags when it is several day out. Would also help with QA.
Thoughts?
You must already have something similar to find the set you reported, no?
You should have most (all?) of those now as the daily run just finished.