[R-sig-ME] Lack of replies from me
bates at stat.wisc.edu
Mon Feb 9 20:20:40 CET 2009
On Mon, Feb 9, 2009 at 12:43 PM, Murray Jorgensen
<maj at stats.waikato.ac.nz> wrote:
> Dear Douglas,
> "Noli illegitemes Carborundum!"
> lme4 is a shining example of how statistical software should be
> developed with software, theory and practice all advancing together and
> in communication. I truly hope that this patch of nastiness, whatever
> it is, fades away and you get back on track. This is such a central part
> of statistics that the whole community is grateful for the time and
> efforts that you and your students have put into it. If others want to
> take a different approach they may surely develop their own software. I
> would be very sad if you took lme4 out of R. R does not belong to
> anyone, not even prominent members of the core group.
> Best wishes, Murray Jorgensen
My thanks to you, Murray, and to the many other people who offered
their support and encouragement privately. It is very gratifying to
me to know that people appreciate access to the software and the
discussions that we have been able to provide. In the last sentence I
wrote "we" referring to the entire mailing list because the exchanges
on the list and the suggestions from many people over the years have
helped in the process of what Murray describes as "software, theory
and practice all advancing together and in communication".
For example, I have said on several occasions (but may never have
written to this group) that the original suggestion to consider sparse
matrix methods for mixed-effects or multilevel models came Harold
Doran on the first day that I met him. He was attending a short
course I was giving and, at a reception before the course, we were
talking and he made this suggestion. Naturally my reaction was, "Hmm,
that's interesting but I really don't know much about sparse matrix
methods" and I planned to just go on with things as they had been.
Fortunately, Harold is persuasive and convinced me that I should learn
something about sparse matrices.
Right now I plan to take a break from acrimonious discussions (and let
me emphasize that it was not discussions on this list that made me so
despondent - Murray's last sentence is much more to the point) and
devote myself to developing code and writing a manuscript for a book.
When the code is developed I will consider in more detail exactly how
it should be released. For the time being I will retain the R-forge
Again, let me say how much I appreciate the many kind messages I have received.
> Douglas Bates wrote:
>> I regret that I have been absent from the list for some time. It
>> happens that in the last couple of weeks I have been involved in a
>> series of extremely unpleasant interactions on another, private email
>> list that have left me with little enthusiasm for developing and
>> supporting CRAN packages. One conclusion I have reached from these
>> interactions is that I will never allow lme4 to be a recommended
>> package in R. I am even having doubts about whether I want it to
>> continue to be a CRAN package at all, as opposed to, say, moving it to
>> Bioconductor or even switching development to another language.
>> I'm sure the last option would be "cutting off my nose to spite my
>> face" and I don't expect I would ever do that. There are many
>> wonderful aspects to R and many reasons why I want to continue to use
>> it. But right now I find myself forced to evaluate options other than
>> putting a package on CRAN.
>> R-sig-mixed-models at r-project.org mailing list
> Dr Murray Jorgensen http://www.stats.waikato.ac.nz/Staff/maj.html
> Department of Statistics, University of Waikato, Hamilton, New Zealand
> Email: maj at waikato.ac.nz Fax 7 838 4155
> Phone +64 7 838 4773 wk Home +64 7 825 0441 Mobile 021 1395 862
> R-sig-mixed-models at r-project.org mailing list
More information about the R-sig-mixed-models