[QUANTITY] Data Model for Quantity v0.5 - inheritance vs aggregation

[David Berry]

Martin,

> My view on this is that a Quantity might describe some pixel values,
> within a Frame that describes the pixel coordinates.  The process of
> dealing with relationships 'between' things then belongs to whatever is
> using the Quantity; ie, you 'apply' a Mapping depending on what you want
> to do, rather than relying on the Quantity having that Mapping.
>
> So for example if you have a Quantity that gives some pixel values in a
> 2D image described by the Frame, then that can be passed around from
> service to service without any extra mappings.  If one service wants to
> read in (Ra, Dec) coords, it applies that Mapping, that knows how to
> deal with this particular Frame.  If a new service comes along with an
> entirely new coordinate system, it has its own Mappings to deal with this.

Not sure I understand this. If a Quantity containing a 2D image had no
Mappings how would any service know how to convert pixel coords into
(RA,Dec) etc? The image Quantity has got to carry this information around
with it since it will in general be different for every such Quantity.
The Mapping in the Quantity is needed to encapsulate things like image
scale, orientation, etc.

David



----------------------------------------------------------------------
Dr David S. Berry    (dsb at ast.man.ac.uk)

STARLINK project		 |	Centre for Astrophysics
(http://www.starlink.ac.uk/)	 |	University of Central Lancashire
Rutherford Appleton Laboratory	 |	PRESTON
DIDCOT				 |	United Kingdom
United Kingdom			 |	PR1 2HE
OX11 0QX                                Tel. 01772 893733
                                             01257 273192