What is preventing the merge of pqR into R?

[kindlychung]

As in title.

[radfordneal]

A full merge would be a fair amount of work, though not technically difficult. It might more easily be done by updating pqR than by updating R Core. The next pqR release, which is imminent, will move from compatibility with R-2.15.0 to compatibility with R-2.15.1.

There are, however, quite a few pqR changes, with substantial benefits, that could easily be moved to R Core versions, sometimes by just replacing entire modules. A few pqR features have been put into R Core, but only ones that involve a dozen or so lines of code, which R Core tends to rewrite in their own way, rather than taking unchanged from pqR. There are also easy changes involving a dozen or so lines of code that R Core hasn't put into the R Core version even though they would have large benefits, for minimal time investment.

The fundamental problem is that R Core is a dysfunctional group, in which no one sees it as their responsibility to assess what is best for the R community, and do that, regardless of whether or not the ideas originated in R Core.