[QUANTITY]: gather requirements for quantity model

[Brian Thomas]

On Tuesday 28 October 2003 04:46 pm, Patrick Dowler wrote:
> At some low levels we are just doing CS; if we are not then we are
> totally screwed because we are mixing domain-specific knowlegde
> and concepts with basic fundamentals of math and comp sci. That
> path leads to disaster.

	I dont accept that. Scientific numbers are *not* CS, they are 
	scientific expression, a different domain. FORTRAN, as much as
	I hate it, is the only language which actually keeps many scientific
	numerical concepts coorectly (for example, scientific 
	signicant digits correct across calculations (at least some of the
	time)). Hence many scientfic programming applications are written
	in it because it crosses "domains".
	
	For me, the "domain" of quantity is partly defined by its re-usability
	by other DM components. For example, will all components have a
	need to use scientific numbers?, if yes then clearly that belongs.
	Will all components have a need to share these numbers easily, even
	if in complex arrangements? if yes, then ideas like "set" and "datacube" 
	*also* belong in the base level package. I also don't see how you can 
	create a vector field (tensor), which *is* a quantity, without having 
	n-dimensional quantities.

	And just when was "quantity" ever defined by us (the VO DM community)? 
	If it has, then we would have no need for requirements. 

	What is the CS "quantity" definition, btw?

				-b.t.