Roughly in order of importance:

- allow SE estimation of parameters with R's nlm optimizer
- expand bifactor and full information factor analysis to polytomous items (Gibbons 2007, Muraki 1995)
- add adaptive quadrature for nfact > 5
- allow for specific parameter constraints
- add more extensive simulation data functions (different distributions, skew/kurtosis, better selectivity of dims)

	