In Praise of Ranges AND Resolutions

[Robert E. Jackson]

Providing numeric ranges for Wavelength/Energy coverages
has the following benefits:

A query with ranges will return fewer hits than a query
without ranges.  Someone, somewhere will have to filter
out the hits which are not of interest.  If the hits
are filtered out before the actual resource is queried,
then resource providers will not have their cpu cycles
eaten up in returning data not of interest.  Better for
the user and better for the resource provider.

Having a standardized range specification at the top level
eliminates the need for the user to learn how each resource
provider specifies the range.  Better for the user.

Admittedly there is a cost to providing the range.  But, the range
can be as simple as two floating point numbers in some
standardized units, e.g., microns.  The user interface could
provide the necessary translation to other units, e.g., eV.

IMHO, the benefits of the user not having to sift through
undesired data are well worth the small cost of providing two numbers.

Perhaps as important as providing ranges is providing resolutions.
A user wanting optical images may not want to sift through all the
data provided by the LINEAR survey to get at an HST/ACS image.
A user wanting spectral data may not want to sift through the data
from an objective prism survey to get at the data from an echelle
spectrograph.

As an analogy, query Google for 'hanisch' and you get 53700 hits.
Query Google for 'hanisch AND bob AND stsci AND virtual' and you
get 132 hits.

Bob Jackson