Going back even earlier, back in 2011 when I was at Google we funded some work on bringing 64-bit integers into R:
https://opensource.googleblog.com/2011/11/bringing-64-bit-data-to-r.html Some of the use cases were language interop, differentiating between 64-bit quantities used as identifiers efficiently, etc. Around that time I also submitted a patch we had for a few releases of R to prevent 2^53 == 2^53 + 1 from returning TRUE which was unfortunately backed out of a later release of R. - Murray On Mon, Aug 3, 2026 at 9:05 AM Brian Repko <[email protected]> wrote: > > I just joined the listserv so apologies if this is not threaded in the > original discussion > https://stat.ethz.ch/pipermail/r-devel/2026-July/084631.html > > A restating of the use cases from Kevin Ushey: > 1. Language interop - C/C++/Fortran, Python via reticulate, Java via > rJava, Javascript / JSON APIs > 2. Data structure interop - arrow, DBI (BIGINT), parquet, hdf5, zarr > There are use cases in both single-cell / spatial transcriptomics - big > sparse array where we need interop above or the ability to handle more than > 2^53-1. > > [[alternative HTML version deleted]] > > ______________________________________________ > [email protected] mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > [[alternative HTML version deleted]] ______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
