[R] Superimposing y-variables in Lattice formulas

Deepayan Sarkar deepayan.sarkar at gmail.com
Mon Nov 17 21:03:01 CET 2008

On Sun, Nov 16, 2008 at 7:27 PM, Stavros Macrakis <macrakis at alum.mit.edu> wrote:
> Thank you for the pointer to your paper "Getting Started with Lattice
> Graphics" -- it looks very useful.
> I think we're talking past each other on the question of the semantics
> of formula operators, and it's probably not productive to continue.  I
> would guess that the underlying issue is that we come from different
> communities, and so have different background assumptions.  Perhaps
> it's the old MIT vs New Jersey story
> (http://www.jwz.org/doc/worse-is-better.html) -- after all, S/R does
> come out of the Bell Labs community, and I was in the MIT Lisp
> community many decades ago....

Thanks for the link; it was an interesting read. I can certainly
identify with the worse-is-better style. I believe "the right thing"
is essentially what Eric Raymond calls the Cathedral style, although
he contrasts it with something different.

One thought that pops into my mind is that "the right thing" approach
more or less implies design by committee, which while important for
standardization work, (1) rarely produces useful innovation, and (2)
raises the barrier to entry for new developers. Since new developers
are essential for Free Software projects to thrive, the
worse-is-better approach seems ideal for them.


P.S. I liked the bit about "losers" :-)

More information about the R-help mailing list