[R] Conflicts using Rcmdr, nlme and lme4

[Douglas Bates]

Kjetil Brinchmann Halvorsen wrote:
> Douglas Bates wrote:
> 
>>
>>
>> I can guarantee that this is the last time that I will fundamentally 
>> redesign the computational methods and data structures used in fitting 
>> mixed-effects models.  This is the fifth time I have done it from 
>> scratch and after this one I am quitting the business.
>>
> 
> Hope that does'nt mean you are quitting completely from statistical 
> computing and R?
> 

No, no.  Not at all.  It just means that I am tired of spending my time 
  thinking about computational methods for this particular model.  I 
think that doing five separate implementation of methods for the same 
problem, where each of the implementations is thousands of lines of 
code, is perhaps a sign of an obsessive compulsive disorder.  :-)

And I probably won't even keep that promise not to do it again.  As the 
title of the James Bond film cautioned, "Never say `never again'".

It's like the way that the omegahat project got named.  Several people 
including John Chambers, Duncan Temple Lang, Luke Tierney, Ross Ihaka 
and Robert Gentleman were discussing the idea of doing yet another 
implementation of an S-like language and Ross was feeling a bit drained 
from having worked on the last one, which is the core of R.  He insisted 
that he didn't want to mess around with alpha, beta versions, etc.  He 
wanted to go straight to omega, the final version.  We found that the 
domain omega.org was already taken so we settled on a statistical 
approximation to the last version, which was omegahat.