[R-sig-ME] Resume terminated lmer fit if verbose=TRUE?

[Mike Lawrence]

I think I figured out how to resume a terminated fit, but I receive a
warning when the second fit completes, so I thought I'd post my work
here. To simplify, let's say I have the model:

y ~ (1|rand) + A*B

Where A and B are continuous numeric variables. The verbose=TRUE
output from lmer should look something like:

0:     826.14130:  1.07781 -6.28068  2.56251  1.44230 0.399600

where:
the "0:" labels the iteration
the "826.14130:" reports the fit statistic (I think?)
the "1.07781" reports an estimate of the standard deviation of the
random effect "rand"
the "-6.28068" reports an estimate of the global intercept
the "2.56251" reports an estimate of the slope of the A effect
the "1.44230" reports an estimate of the slope of the B effect
and the "0.399600" reports an estimate of the slope of the A:B interaction

One can store the last 5 values (all the estimates) in a vector called
"from_verbose" and then construct a list as follows:

to_start = list(
    ST = list(
        matrix(from_verbose[1],1,1)
    )
    , fixef = from_verbose[2:length(from_verbose)]
)
dimnames(to_start$ST[[1]]) = list('(Intercept)','(Intercept)')
names(to_start$fixef) = c('(Intercept)','a','b','a:b')

Now one can use "to_start" as the value of the start argument, as in:

fit = lmer(
    formula = y ~ (1|rand) + a*b
    , start = to_start
)

When I do this and additionally set verbose=TRUE on the resumed fit, I
indeed see that it picks off where where the terminated fit left off.
However, when the resumed fit finishes, I get the warning:

In sort(names(start)) == sort(names(FL)) :
  longer object length is not a multiple of shorter object length

Should I be concerned about this warning?