Re: [Leopard-discussion] Module management phase

[Tom A. <ad...@ne...>]

bm...@te... wrote:
> On Fri, 12 Dec 2003 02:53:21 -0800
> "Jake" <ja...@jw...> wrote:
> 
> 
>>I will start digging on the database end of things and see what I can come
>>up with.
>>
>>Jake
> 
> 
> i'm working on a proposal for data structures / validation whitin xml
> and relaxng schemas.
> 
> if any idea on how best we can describe the 39.50  pragmatically,
> before we begin dealing with it please share with us.

This is my take on a pragmatic view of Z3950. Matthew will need to 
verify this and add or subtract from my take.

The Z3950 also called 23950 ISO provides a framework which looks like 
the most efficient way to handle interoperability. It's the basis for 
our workflow structure and it is not all inclusive for every government 
entity.

Different agencies have their own XML structures. We have to also 
consider the XML schemas at the Department of Justice Interoperability 
project.

http://it.ojp.gov/topic.jsp?topic_id=43

has ample informaion about their database vocabulary.

They've created a catalog of elements which will become part of our Open 
Government Interoperability standard along with z3950.

The database vocabulary will depend on the application (e.g. Juvenile 
services tracking or deeds of trust or business licenses). So that part 
of the database needs to be built in the application module which 
someone would download and install.

The implementation of Z3950 in Library Sciences provides API's which can 
map to existing databases at libraries. A library can also migrate it's 
data to the Z3950 server and use it exclusively.

A central server provides location information to any host attempting to 
connect to the Library network. Each library has a Z3950 server.

If I go a search, each Z3950 server in the Library network will respond.

They could have no entries or many entries.

As the responding servers begin their searches, a web page shows up and 
lists the reporting servers and then the results of the search.

When the user retrieves information from reporting servers, those 
records can become cached. In no case does one database sync with 
another and import or export records.

http://www.loc.gov/z3950/agency/zing/zing-home.html provides the specs 
for the Search/Retrieve for the Web (SRW) and the Common Query Language 
(CQL).

Those are important parts of the work flow.

The other part of the work flow we must consider deals with secure 
connections, rights and privleges.

To summarize

1. We use the XML schemas and database vocabulary of the application in 
question.

2. We provide a way to talk from one database to another.

3. We provide security.

The current application we want to create for leopard involves juvenile 
services. It's a tracking system and integrates different databases from 
different agencies.

http://www.ncsl.org/programs/lis/intjust/report01.htm

explains the kind of program we will need. I believe we will have a 
document from Newport News on Wednesdy whch will provide specific detail 
on the database. Matthew is also supposed to gain access to the Forest 
County-Hattiesburg system called SWORD developed by USM.

We will want to built a system that works for both using Z3950 for 
interoperability and then consult the Justice department data models to 
include the elements needed for this application.

http://it.ojp.gov/topic.jsp?topic_id=43