Re: [boost] Interest in dimension/unit-checking library?

From: Deane Yang (deane_yang_at_[hidden])
Date: 2005-10-11 11:40:29

Matt Calabrese wrote:

> On 10/11/05, Deane Yang <deane_yang_at_[hidden]> wrote:
>> Although I appreciate your efforts to implement an efficient way to
>> automatically convert different units for the same quantity, I would
>> still prefer a library that requires an explicit "cast" to change units,
>> rather than implicit. Perhaps the library could be built in two layers,
>> with the lower level requiring explicit casts and a layer above that
>> that implements implicit casting for those who want it.
> 
> 
> When I first started with the library, I thought that might be a good idea,
> but as development went on I decided that it wasn't really necessary. Could
> you actually give any example as to why you would actually want such
> restrictions? Keep in mind that explicit unit casts are still available.
> 

I work mostly with calendar time. Although I want to view both years and 
months as units for the same vector quantity (time duration), I find 
that in any one code module, I always want to use a single consistent 
unit (which might be templatized). All the unit casting occurs only for 
inputs and outputs to the module. So if my code tries to add years to 
months, it's a coding error.

I can't speak for the people programming in physical units, but I just 
can't imagine wanting to mix variables representing the same dimension 
with different units in a single arithmetic expression. Even with a 
dimensions library, it seems like a recipe for trouble to me. To me it's 
much safer to convert all inputs (which may have different units) into a 
single set of consistent units when entering a module and converting all 
outputs into the desired units when exiting. By using templates, this 
can all be done very cleanly without hardwiring either the internal 
units or the external units.