RE: [boost] Physical Quantities revisited

From: Paul A. Bristow (boost_at_[hidden])
Date: 2003-10-11 10:57:08

I hate to be negative about your proposal (unlike many Boosters who seem to take
great delight in shooting/shouting anything down) but having glanced at the code
examples, I really don't think that this is the right way to go.  I still think
that everything should be held internally in SI units - now that double has
enough range to cover almost everyones needs.  I'm worried about conversion
loss - your first example warns of the danger.  And I am worried about the
expansion to other all the types of units - many are unecessary (feet even!) but
widely used.

But we definitely still need a good system for Quantity units etc.

Paul

Paul A Bristow, Prizet Farmhouse, Kendal, Cumbria, LA8 8AB  UK
[phone]   Mobile [phone]
mailto:pbristow_at_[hidden]

| -----Original Message-----
| From: boost-bounces_at_[hidden]
| [mailto:boost-bounces_at_[hidden]]On Behalf Of Andy Little
| Sent: Friday, October 03, 2003 4:55 PM
| To: boost_at_[hidden]
| Subject: [boost] Physical Quantities revisited
|
|
| Hi,  I am a Newbie to Boost :-)
|
| I am currently working on an upgrade to a
| point-and-click small wind turbine rotor blade design app,
| which involves various physics.
| The original did its math in terms of inbuilt types ,
|  but significant time is spent looking up (say)
| whether a length is in metres or millimetres
|  or an angle in degrees or radians.
| I have also had requests from users for ability to work
|  in (say) feet and inches etc rather than s.i. units.
|
| I have had a look at various physical quantities libraries
|  including the "quantities" library at boost,
| and SIunits at
| http://www.fnal.gov/docs/working-groups/fpcltf/Pkg/SIunits/doc/0SIunits.html
|
|  but they have an odd syntax.
|
| ie  both do something like:
|
| 	length mylength = 20 * metres(); //  here length is a type
|
| whereas what would seem to be more useful would be:
|
| 	metres myLength(20); 	// here metres is the type
| 					//*1 see footnote on explicit ctor
|
| The critical difference here is that the type is the
|  unit of dimension, rather than the dimension itself.
| As a programmer if I want to know the units
| of a variable they are encoded in the type
|  and if I can't decipher them I can do a conversion:
|
| metres  mylength = metres(length_with_unknown_units);
| // ok... whatever units length_with_unknown_units was ,
| // mylength is its value in metres.
|
| I have spent some time creating a physical_quantities class template,
|  with this approach, and have put the source code and
|   a demo programme (and sample output) on my web site at:
| http://www.servocomm.freeserve.co.uk/Cpp/physical_quantity/index.html
|
| I would be interested to hear any views
|  as to whether this approach is worth pursuing, or suggestions or
| improvements.
| ( Alternatively there may be a library that does this that I havent found
| ... ? )
|
| note:  I have tried to answer some obvious questions and criticisms
|  in the Potential FAQ on the page.
|
| ---
| *1 / this example is actually an error if  coded as:
|
| metres myLength =1;
|
|  in my version of physical_quantities
|   because it violates dimensional analysis.
|
|
|
| _______________________________________________
| Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
|
|