Floating point trouble with x86's extended precision

[Robert Dewar]

> The current behavior sometimes results in objects which appear to
> spontaneously change their values between two read accesses (without
> any intermediate write access).  I find it hard to believe that such
> indeterminacy is permitted by any language standard.

By the way this is a well known and much discussed issue. But it certainly
is the case that it retains its capability to cause surprise and shock
when it is encountered for the umpteenth time :-)