Wish/to-do list

Logistics/dissemination

* set up r-forge tracker (e-mail r-forge admins/bug HS); import this list to r-forge tracker
* (BMB) provide package from www.math.mcmaster.ca for times when R-forge build is absent or out of date (source package only should be OK, since there is no compilable code in the package).  Set up a mirror or alternate repository 

Glitches

* fix warning about non-pos-def corr matrix with norand
* prettify warning about non-pos-def corr matrix otherwise

Links and families

* allow alternative parameterization (i.e. var=phi*mu) of NB? ["NB1"]
* implement cloglog link, identity, ... test!
* allow alternative RE distributions?

Capabilities

* methods: terms(), predict() [with newdata, se.fit], simulate(), model.matrix(), confint() [from quad or MCMC]...
* residuals() options
* multinomial family?
* plot method; fortify method?
* fix VarCorr method for lme4 compatibility; print.VarCorr.admb method
* MCMC: add mcmc2.  Better examples/documents


Documentation
* warn on summary() that p-values are based on Wald Z tests
* work on vignette

ADMB end/interface
* Build logic to speed up single-RE case (i.e. achieve speed of old version in special case)

Data and examples

* add some sort of progress bar or "working ..." message in the absence of voluminous messages?  progress bar is difficult since ADMB doesn't know how long it will take
* update epil2?  draw from MASS instead?

Other
* switch to using R2ADMB as interface layer?? Compare R2ADMB and glmmADMB interface code, take the best of both ...

Long-term
* "R-side" random structure (difficult/long-term)
