[ADMB Users] crappy software

[dave fournier]

OK at some point sure try bugs  But the first issue is if R really sucks 
at this
because a lot of its users presumably use this stupid test (I would
use likelihood ratio but that is another issue).  So first one wants a 
really clear and
correct comparison of R with something that does better. (SAS I don't 
have and
the R gurus will mumble about lack of source code etc anyway.)

So lets get a nice clear correct example based on Zhang et al.