[GILS Logo]

gils: Re: Time Period Elements in the GILS Profile.

gils: Re: Time Period Elements in the GILS Profile.

Re: Time Period Elements in the GILS Profile.

Doug Nebert (ddnebert@usgs.gov)
Wed, 04 Sep 1996 11:11:13 -0400 

Message-Id: <322D9C11.1E33C37F@usgs.gov>
Date: Wed, 04 Sep 1996 11:11:13 -0400
From: Doug Nebert <ddnebert@usgs.gov>
To: gils@cni.org
Subject: Re: Time Period Elements in the GILS Profile.

Margaret St. Pierre wrote:
> 
> I have a question about Version 2 of the Application Profile for GILS.
> 
> 1. The element "availableTimePeriod and its subelements
> "availableTimeStructured" and "availableTimeTextual" are all listed as
> repeatable.
> 
> 2. The element "timePeriod" is repeatable and its subelements
> "timePeriodStructured" and timePeriodTextual" are non-repeatable.
> 
> In these two cases, the repeatability of the subelements appears to be
> inconsistent.
> 
> It seems to me that the two subelements are alternative
> representations for the same thing and should be non-repeatable in
> both cases.
> 
> Does anyone else have any opinions?
> 
> Margaret St. Pierre

Regarding "timePeriodStructured," perhaps there is some lack of
explanation of how structured time can be managed. Itis important that
these elements be searchable, not just presentable, to give users a
handle on the currentness of the data for a given purpose. To achieve
this, the subelement should be repeatible with an entry for each type of
time reference it supports. For our data we need to be able to store and
search for single date-time, multiple observations, and continuous
series of observations (date-time range). Can someone point to the
reference that verifies that these constructs are supported?

This set of elements also creates a problem for us in that it is in
conflict with the Geospatial "GEO" profile being developed that is
supposed to inherit many bib-1 and GILS elements.  To achieve meaningful
search across both types of servers (it is desired that GEO servers be
approachable as GILS servers) the semantics of the time content elements
should be the same. In GILS, the time reference appears to use a USMARC
convention that is not explained in the Profile. In GEO, time can be
represented as a single date-time, multiple discrete date-times, or a
range of date-times (beginning and ending times of coverage).  These
constructs are important in providing users with a sense of temporal
coverage that can be visualized relative to other data entries, and to
help in search for time-critical data.

Are the structured time fields being used as searchable elements by
default?  If not, they should be. If so, we need to assure that there is
a full and unambiguous mapping between the semantics of the GILS and GEO
elements for time if they are to be interoperable.

Your help in consolidating these elements, from the service perspective,
is appreciated.

Doug Nebert
FGDC

[CNI Home Page]