[QUANTITY] Review of Quantity suggestions

[Ray Plante]

On Thu, 24 Jul 2003, David Barnes wrote:
> Just quickly in response:
> 
> >   -  Question: specify a datatype?  I feel we should be pretty restrictive 
> >      here.  At most, we should allow floating point (not distinguishing 
> >      between "float" and "double"), integer, and complex.  I'm not even 
> >      positive about integer (someone have an example out there?).  
> 
> I doubt you can throw away integer types.  Many detectors record
> counts which were integer last time I checked.  Other examples
> include image mosaic weights ...
> Antenna number...

This may be a question of scope.  The type of quantities that I had been 
thinking about are those associated with measurements and calculated 
values that can potentially have errors associated with them.  We normally 
don't expect calculate an error for the number of antennae in an array and 
to apply the kind of data model we've been talking about to antenna number 
would be overkill.  

Counts is a good example, though.  (Is it typical render statistics on 
integer values, such as counts, as integers as well?  Do we say 350 +/- 
2.5 or 350 +/- 3 or 350.4 +/- 2.5?)  

(Incidentally, in AIPS++, the Quantum class is a template, so it is 
possible to support integers, etc.  However, floating point quantities are 
the only type available to end users in Glish.  I was suprised that 
couldn't create a complex quantity.)

> I can give dozens of examples of quantities which really should
> only ever allow integer values.  

Keep them coming.  It would be good to know what we're dealing with.

cheers,
Ray