[R-sig-ME] glmmADMB fails to fit poisson data

[dave fournier]

If you look at the code for the negative binomial model in glmmadmb you 
will see that it
initially fits the

    log(Y_i+1) transformed data.  This is by far the most stable model 
numerically and good for
getting initial parameter estimates.

However for the Poisson case the code  never got around to writing itself.

I'm sure the code will rectify its oversight in good time.