Re: reading NaNs Thomas Bushnell BSG 27 Oct 2005 19:47 UTC

Jens Axel Søgaard <xxxxxx@soegaard.net> writes:

> Btw - I agree that #<not-a-number +> is a little more helpful than
> #<not-a-number> during debugging, but if there is no source location
> information to find the actual + producing the NaN then the
> helpfulness is limited.

Ah, but here's the genius of the source-tracking feature!

You could turn on a special "track source on NaNs" feature, which
squirrels away, say, a complete stack record (or something less
perhaps, depending on what you ask for), and then every NaN generated
now gets tagged with a *very* nice handle for debugging.

Of course the implementation of such a feature would be very slow, but
that's ok, the point is for debugging.

Thomas