R-beta: Re: S Compatibility

Ross Ihaka ihaka at stat.auckland.ac.nz
Thu May 1 03:42:25 CEST 1997

Bill Venables writes:
 >  > implementation.  But I may be wrong ...
 > My casual remark that started all this was not intended to hint
 > at any legal or ethical dilemmas lying just around the corner.  I
 > do agree these are serious issues, though.  One small anecdote I
 > can contribute was that both John Chambers and Ross Ihaka were
 > present at the SISC conference in July last year and John, with
 > great generosity of spirit, I thought, (or perhaps very acute
 > foresight) was publicly very supportive of Ross & Ron's efforts.

I would have to agree.  John has been extremely generous.  Not only at
the Interface, but in a variety of ways, some connected with R and
some not; and all this going back a long way.  I don't think it was
commercial "foresight", he is just extraordinarly supportive.  This is
the collegial side of things which I see as being hobbled by the
commercial situation.

 > It may well have been foresight, because my own feeling is that
 > the main effect of R will be to make available an S-like language
 > to people like students and enthusiasts who under normal
 > circumstances would not be MathSoft customers -- that is until
 > they graduate and go on to an honest job, of course.  S and R
 > will diverge, too.  That much is plain to anyone who has had a
 > glimpse of S version 4, as I have been priviledged to do.  So
 > ultimately if you want the genuine article you will have to pay
 > for it.  I suppose that also applies to the GUI (or gooey?) front
 > end to S-PLUS version 4, due out real soon now...

Perhaps we can have two genuine articles with differing strengths :-).
I personally am also concerned about:

1.  Availability.  NeXT users are S orphans now.  How long before
    Unix as a whole goes the same way because of "commercial
    realities"?  This won't happen with R.  At the very worst, you
    will be able to take the source and make it run yourself.
    This has already happened with NeXT users.

2.  A Reference Model.  I hope that someone will come along and
    find much better ways of doing this sort of thing.  (Good luck to
    them!  I hope I can help).  Having our mistakes (and successes)
    available in source form for public scrutiny has to be a good thing.
    In addition I hope that having a large body of free applications
    code might be a spur to language gurus to produce a nice core
    language which could then run (translated versions of) this

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list