Issue ERROR-CHECKING-IN-NUMBERS-CHAPTER

[Kim A. Barrett <IIM%ECLA@ECLC.USC.EDU>]

The question of whether NaNs and such cause TYPE-ERROR or ARITHMETIC-ERROR is
what I feel unsure about.  We currently signal ARITHMETIC-ERROR, but I don't
believe there was a lot of analysis that went into that decision.  I think
that any decision we make on this is likely to be pretty arbitrary.  I can't
think of any strong first-principled arguments either way, so it may be a
matter of which seems more convenient to the people who really care about
an error being signaled for such things.

kab
-------