R help - Sep 2012 - balanced two-way crossed balanced random effects analysis of variance

Hello, All,

I am wondering if there is a standard library/function for calculating variance
components (or the variance of the total mean) of a balanced two-way crossed
random effects problem a la equation 7.4.10 of Scheffe or section 14.5 of
Brownlee.  My model is 
s_ij = \mu + x_i + y_j + e_ij
x, y, and e are random and I need the variance of the mean of s.

I know that I can do this with lme4 or nlme, but I need the unbiased variance
calculated from the sum of squares, not a maximum likelihood estimate.  I know
this is easy enough to implement myself (I have already done it.), but I would
like to be able to reference an existing downloadable implementation.

Thanks for any help.

-Frank