Re: Conflict between R 4.5 and Bioc 3.20
On 17 April 2025 at 05:52, Dirk Eddelbuettel wrote:
|
| On 17 April 2025 at 14:43, Charles Plessy wrote:
| | I just noticed that r-base 4.5.0-1 cannot migrate to Trixie because it breaks
| | Bioconductor 3.20, which was developed and tested upstream against R 4.4.3.
|
| Are you really there is actual 'breakage' in any of the packages in 3.20?
|
| While they were built under R 4.4.3 they continue to work with R 4.5.0 as there
| is no API / ABI change.
|
| You can easily test that under the r-base Docker container which already has
| R 4.5.0 and points at testing/unstable as usual. I just ran 'apt update -qq;
| apt install r-bioc-alabaster.sce' and it of course loads as expected. [1]
|
| I also usually mix and match on my own box and don't immediately update to
| new BioC there and never had issues. So I am not sure the situation is as
| dire as you say.
I should re-clarify that packages such as BiocManager will indeed insist
matching pairs of BioConductor and R releases (ie 3.20 with 4.4.*, 3.21 with
4.5.*). R itself cares less. As Charles outlined this does not leave us
with many great options given the release date of BioConductor and our soft
freeze. Imposing that R must stay at 4.4.3 is also heavy.
Dirk
--
dirk.eddelbuettel.com | @eddelbuettel | edd@debian.org
Reply to: