|
Boost : |
From: Phil Richards (news_at_[hidden])
Date: 2004-01-09 08:30:51
On Fri, 09 Jan 2004 12:30:37 +0000, Paul A. Bristow wrote:
[in reply to Dan W. on the CAD problem]
> Should you being carrying out computations with mils (and perhaps mm) as
> _integers_, rather than as any sort of real?
I have to say, I would expect anybody who wants fixed precision quantities
to define a fixed precision value type and use that as the dimensional
quantity value type. Delegate the problem from the physical quantities
library to the underlying value type, where (as far as I'm concerned) it
really belongs. Not that there might not be a reason why it won't work,
but I can't think of one.
(After all, if it is possible to plug in the interval library as a value
type, I don't see what would be so much harder... and on my simple tests
it did seem to work OK - though having said that my code has moved on
quite a lot and I may have broken it again.)
phil
-- change name before "@" to "phil" for email
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk