Rick Graham Sat, 13 Sep 2014 23:01:11 -0700
I suppose it's not too bad.
my_nextfloat(x)=with_bigfloat_precision(()->nextfloat(x),x.prec)