On 20 September 2017 at 16:44, Martyn Plummer wrote: | On Wed, 2017-09-20 at 09:45 -0500, Dirk Eddelbuettel wrote: | > Thierry, | > | > This has always been a considerable puzzle, and the source of some amusement, | > to me. It is a clear design bug in Rscript because _code that would | > otherwise work_ will not if it happens to rely on library(methods). S4 comes | > to mind. This has also been answered at least a few times on SO. | > | > Now, 'r' from littler always did this differently and lods methods, just like | > R itself does. And 'r' still starts (way) faster as I once documented in a | > blog post. | > | > If someone from R Core reads this and nods silently: please put this to rest | > and make sure Rscript also loads the methods package. | | We have been here before, most recently in 2016 | | https://bugs.r-project.org/bugzilla/show_bug.cgi?id=16921 | | There are opinions on both sides, of course, but the decision rests in | the hands of those who use extensively for its original purpose (i.e. | running short snippets of R code) and for whom the overhead from | loading the methods package is still unacceptable, despite being | substantially reduced since this decision was made 10 years ago.
Sure, I don't care. I wrote littler for my purposes. But to be honet I still find it strange that you all troll your users by withholding one (base) package on one platform only, but your call. Dirk -- http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org ______________________________________________ R-package-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-package-devel
