[boost] Re: Re: Dimensional Analysis

From: Robert Ramey (ramey_at_[hidden])
Date: 2005-02-01 13:57:06

My main motivation is to divide the idea of the units library from a large
"thing" that requires much effort at consensus, and developement into two
smaller "things" - one dependant upon the other.  In now way do I want to
suggest that units is not important.  Its just something that should be
separate and sit on top of a dimensional analisys library.

This view is motivated by my personal view that the effort involved in all
apects of software development grows disproportionatly to the size of the
project involved.  If the project can be divided into layers, or losely
coupled components, the effort is reduced significantly.

Also, it results in a better, more elegant, more widely applicable design.

So lets make a minimalist dimensional analysis library whose function is to
check arithmetic operations at compile time for consistency.  I'm not even
convinced that fractional dimensions are needed.  If the really are, maybe
the library can take a template paramter the specifies either  an integer or
compile time rational type.  This would cleave off another tricky piece -
compile time rationals.  This would diminish the load on the writer of the
dimensional analysis library and leave the issue of fractional dimensions to
someone who needs this sufficiently to create a compile time rationals
package.

Robert Ramey

Andy Little wrote:
> "Phil Richards" <news_at_[hidden]> wrote
>> On 2005-01-30, Andy Little <andy_at_[hidden]> wrote:
>>>> It's all we need because *all* our interfaces to the outside world
>>>> are defined entirely in terms of SI units.
>
> 1)Unfortunately not every one has the luxury. Many incoherent units
> exist in many real world situations, where errors cost money and
> worse. 2)Even within SI a quantity can be expressed in various units.
>
> regards
> Andy Little
>