On 13 April 2025 at 02:10, Johannes Ranke wrote: | For what it's worth, I needed to add libzstd-dev to the build dependencies of | littler while backporting it to Debian bookworm (not for bullseye). Sorry, I | have no deeper insight but that is what I experienced. | | | Could it be related to this R 4.5.0 NEWS entry: | | | - There is some support for zstd compression of tarballs in tar() and untar(). | (This depends on OS support of libzstd or by tar.)
It very much could be, and that is what I tried to ascertain earlier in dependencies of r-base-core, but maybe it is more subtle. We will tie it down though. Thank you both for continued help. Dirk | | | Cheers, | | | Johannes | | | Am Samstag, 12. April 2025, 23:26:51 CEST schrieb Santiago Vila: | | > > | Could it be the case that the right fix would be | | > > | to add a binary dependency on libzstd-dev to | | > > | one of the current build-dependencies? | | > > | | > > That is what I did. | | > | | > Oops, didn't notice. Sorry. Nevermind, then. | | > | | > > The open question I myself have is whether R (now at 4.5.0) brought in | | > > zstd | | > > use (I think it can be use internally where available) or whether Python | | > > brought it in. I think it may be Python. | | > > | | > > But yes I also think about possibly adding it to r-base-dev, a | | > > (near-)virtual package ensuring an R environment. | | > > | | > > Thoughts? | | > | | > (Nothing to say on my side. Hope other R people who really know the stack | | > can answer). | | > | | > Thanks a lot. | | | -- dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org
