Re: Type conversion (Was: Brain Dump of Towards FIT 1.0OpenSpace session)

On Friday, August 22, 2003, at 12:55  PM, [email protected] wrote:

> One of
> the major issues is, of course, floating point compare precision,
> which I have covered with a special entry in my type dictionary.

The CalculatorExample also shows that it is possible to imply precision 
in the table data by determining how many digits have been specified (a 
big deal in my son's science class). The example uses a subclass of 
number called ScientificDouble. These objects remember how many digits 
of precision are implied by the string representation and then compare 
results to that precision.

Note, ScientificDouble is not part of the Fit core, but an example of 
the capability of the core as it exists in Java.

I consider this an elegant solution to the problem, though possibly not 
the only one. I think it is important that no fit implementation 
preclude this solution (if it is possible in the host language) but am 
not at all sure how to write acceptance tests for this desire.

--
Ward Cunningham
503-245-5633 v	mailto:[email protected]
503-246-5587 f	http://c2.com