[Quantity-users] Re: [Quantity-devel] temperature conversion

[<ber...@t-...>]

Bernd Speiser wrote:

> 
> What possible solutions do we have?
> 
> (1) decide for one of the above - but which one? - and accept the situation
> (2) remove DegreeCelsiusTemperature from TemperatureGroup and put it into a
>     new GT, simultaneously disallow operations between objects of same 
> dimension
>     but different variable type. This would require an explicit 
> conversion if
>     arithmetics between the types are desired, which would not be 
> possible with
>     the present constructors.

I tried to solve the problem by separating Temperatures into 
ThermodynamicTemperature (in K and DegreeRankine, according to the Mills book),
CelsiusTemperature, and FahrenheitTemperature, by updating the conversion and 
assignment operators in Variable (which was good to do anyway; before copy 
construction and assignment was certainly not strict enough), and by providing
explicit conversion methods for the *Temperatures.
Now, one can e.g. explicitly convert CelsiusTemperature into 
ThermodynamicTemperature before performing an addition. Then, of course, 
addition becomes commutative, although, e.g. the additions in

   CelsiusTemperature Tc2 (0.);
   ThermodynamicTemperature Tt3 (298.15);
   ThermodynamicTemperature Tt4a = Tt3 + Tc2;  // prints 298.15 K (wrong)
   ThermodynamicTemperature Tt4b = Tc2 + Tt3;  // prints 571.3 K
   CelsiusTemperature Tc4a = Tt3 + Tc2;        // prints 25 oC
   CelsiusTemperature Tc4b = Tc2 + Tt3;        // prints 298.15 oC (wrong)


yield different results! On the other hand,

   ThermodynamicTemperature Tt5 = Tt3 + ThermodynamicTemperature (Tc2);
   ThermodynamicTemperature Tt6 = ThermodynamicTemperature (Tc2) + Tt3;

yield the same result (571.3 K).

So, in my opinion, in lieu of a possibility to disallow mixed expressions 
completely, there should be extreme care with such mixed cases, and explicit 
conversion should ALWAYS be used.

Bernd
-- 
=======================================================================
Bernd Speiser
Institut f"ur Organische Chemie
Auf der Morgenstelle 18
D-72076 T"ubingen
Germany
phone: +49-7071-2976205 (office) +49-7071-2976242 (laboratory)
fax:   +49-7071-295518
e-mail: ber...@un...
Internet: http://www.uni-tuebingen.de/speiser
=======================================================================