[QUANTITY] The difference between quantities and measurements (Was: Re: [QUANTITY] Why quantities always have errors (Was: Re: [QUANTIT] Use-cases, role in larger scheme)

[Martin Hill]

>>And what I'm trying to say is: I don't think there is anything common
>>between the type 'flux' and the type 'position' that merits having a
>>'quantity'.
> 
> 	But there are commonalites. Think for a minute on the package 	
> 	that must handle passing around these things. Or searching for
> 	the values of these things, or breaking apart one of these things
> 	to make many children (big position quantity broken into smaller
> 	position quantities) and you will begin to see that the same operations,
> 	treatment are needed over and over again.

The bits that do the 'passing around' will be more general than 
[quantity] - such as XML (un)marshallers.  And to be of any use, the 
bits that break them apart will have to be specific to the breakable 
apart thing.  Similarly the things that find, compare and process 
quantities will all need (in practice) to be specific to the findable, 
comparable and processable things.

We run the risk of designing a data model so vague and all-inclusive 
that it offers us no extra value/meaning, but instead just puts an extra 
layer of restrictions between the values we want to describe (eg flux, 
position, error) and the tools we already have to describe them 
(integer, real, string, list).

But it is a useful discussion all the same to consider the different 
quantities that will go into the huge VO melting pot.  So I'm going to 
shut up now and all you can have some peace and quiet.  From me anyway!

Cheers!

MC

-- 
Software Engineer
AstroGrid @ ROE
Tel: +44 7901 55 24 66
www.astrogrid.org