Re: Locate an underflow

[thompson(at)orpheus.nascom.nasa.gov (William Thompson)]

Sorry to chime in again.  I should have included this in my previous post.


Paul van Delst <paul.vandelst@noaa.gov> writes:

>... You state that there are cases where underflows are not harmless (via the
>"almost always" qualifier) but then go on to say you would like to suppress
>their reporting? Why not fix the error so you don't get underflows for that
>rare case where they CAN cause a problem? ...

That's because, while I can imagine in theory that there might be a situation
where one would want to be informed of a floating underflow error, I've never
actually run into one in practice.  But underflow error messages are
ubiquitous.

William Thompson