[Rd] Julia

[Douglas Bates]

My purpose in mentioning the Julia language (julialang.org) here is
not to start a flame war.  I find it to be a very interesting
development and others who read this list may want to read about it
too.

It is still very much early days for this language - about the same
stage as R was in 1995 or 1996 when only a few people knew about it -
but Julia holds much potential.  There is a thread about "R and
statistical programming" on groups.google.com/group/julia-dev.  As
always happens, there is a certain amount of grumbling of the "R IS
SOOOO SLOOOOW" flavor but there is also some good discussion regarding
features of R (well, S actually) that are central to the language.
(Disclaimer: I am one of the participants discussing the importance of
data frames and formulas in R.)

If you want to know why Julia has attracted a lot of interest very
recently (like in the last 10 days), as a language it uses multiple
dispatch (like S4 methods) with methods being compiled on the fly
using the LLVM (http://llvm.org) infrastructure.  In some ways it
achieves the Holy Grail of languages like R, Matlab, NumPy, ... in
that it combines the speed of compiled languages with the flexibility
of the high-level interpreted language.

One of the developers, Jeff Bezanson, gave a seminar about the design
of the language at Stanford yesterday, and the video is archived at
http://www.stanford.edu/class/ee380/.  You don't see John Chambers on
camera but I am reasonably certain that a couple of the questions and
comments came from him.