Re: quantity database

• To: visad-list@xxxxxxxxxxxxx
• Subject: Re: quantity database
• From: Steve Emmerson <steve@xxxxxxxxxxxxxxxx>
• Date: Wed, 15 Apr 1998 10:09:29 -0600

Bill,

>Date: Wed, 15 Apr 1998 10:48 CST 
>From: "William L. Hibbard" <WHIBBARD@xxxxxxxxxxxxx>
>To: visad-list@xxxxxxxxxxxxx
>Subject: Re: quantity database 

In the above message, you wrote:

> OK, I'm starting to get the picture.  One thing to be careful
> about: many applications will need to create lots of similar
> RealTypes.  For example, some weather models generate densities
> for lots of different frozen states of water (e.g., pellets,
> grappel, sleet, big hail, medium hail, little hail - I've
> probably got the details wrong here, but do know that some
> models generate about 20 of these things) and need each one
> to be a different RealType to support visualizing correlations
> between densities of these different states.

I think this could be accommodated by a standard entry like the
following:

    "mass density" -> RealType("mass density", "kg/m^3", FloatSet(...))

together with some meteorological-specific entries like the following:

    "hail density" -> RealType("hail density", <some unit>, FloatSet(...))
    "rain density" -> RealType("rain density", <some unit>, FloatSet(...))
    ...

> RealTypes are sort of like variable names when you're doing
> mathematics.  You end up with so many that you need to start
> using subscripts and greek letters (hmmm, sounds like Unicode).

Sounds like we need a supreme dictator to bring order to this chaos.  I
volunteer!  :-)  :-)  :-)

Whether I succeed or go crazy first is problematical.  :-)

--------
Steve Emmerson   <http://www.unidata.ucar.edu>