Download View Free PDF - IRMA International

Journal of Database
Management
INDUSTRY AND PRACTICE
Associate Editor Editorial
Object-Oriented Database: How Wide is
the Application Range?
The information systems field is a dynamic
one in which we are constantly moving the state–
of–the–art forward. Database management systems are very much a part of this movement. We
have never been satisfied with the current DBMS
paradigm, always looking for a new and better
approach. Navigational DBMS was followed by
Relational DBMS and now the latest embodiment
of this trend is Object–Oriented Database Management Systems (OODBMS).
The OODBMS paradigm is striking in that it
includes several truly advanced features including
abstract data types, encapsulated code, and inheritance hierarchies, plus, of course, the concept of the
object itself. Quite naturally, the direction that
applying the new OODBMS concept has taken is in
the exploitation of the new features in application
areas that did not lend themselves well to the
relational or navigational models. Thus, we have
seen OODBMS applications in rather complex
data storage applications such as electric power
grid management, computer–aided design for VLSI
circuits, computer–aided software engineering,
geographic information systems, medical decision
support, and document browsers, among others.
While these are interesting and important applications, I cannot help but wonder if on a broader
scale we are missing the point and missing the boat.
Database management, as a technology, has been
successful at least in part because of its very broad
applicability. While in the past it has perhaps not
been applied to the relatively exotic applications
mentioned above, it has been applied to the wide
range of common, “bread–and–butter” applications in accounting, inventory control, customer
account management, and so forth, in virtually
every conceivable industry. In fact, it was so
successful in the navigational era, that when the
relational era emerged, there was widespread resistance – and there still is – to converting the existing
data and programs to relational technology because
of the huge investment that had been made originally in, what are now, smoothly running applications. In most companies today, new applications
are based on relational technology and there certainly has been activity in converting some navigational applications to relational technology, but
there has also been a great deal of effort expended
in devising ways to continue to use the so–called
“legacy” systems in the modern environment.
Nevertheless, it is clear that relational technology,
barring the factors just mentioned, could replace
navigational technology. Yes, there are some comparative advantages and disadvantages (the ease–
of–use of the relational query interface) and some
lingering controversies (relational versus navigational performance in some applications) but overall, everyone recognizes that relational technology
is capable of handling common applications.
The question now is whether object–oriented
Summer1995
31
Journal of Database
database technology is to be relegated to exotic
applications only or if it has the capability to
handle the wide range of common applications, as
well. If object–oriented database is truly going to
wear the mantle of the next database generation,
then perhaps it is time to investigate its use in
common applications. I suggest that this effort
should proceed in two directions. One is to examine the ways that the OODBMS can efficiently
store the data of common applications as those
applications operate today. This will have implications in database design for object–oriented databases and then in the conversion of data from the
relational or navigational form to the object–oriented form. The other is to explore whether object–oriented database has the ability to add value
to common applications in uncommon ways, using
its innovative features.
To expand on this latter idea with an example,
consider a standard banking application in which
tellers and customer service representatives interact with customers using terminals to access their
account data. Also, consider the object–oriented
database abstract data type facility which permits,
among other things, the storage of digitized pictures in the database. Ordinarily, the banking
application requires only numeric and alphabetic
32
Vol. 6 , No. 3
Management
data. But, with the OODBMS, the teller could call
up a photograph of the customer or the customer’s
signature, along with the normal account data.
Does this constitute added value sufficient to justify the expense of the conversion to the OODBMS?
Maybe. Clearly there may be value judgements of
intangibles to be made. A value judgment would
also be required of the worth of the increased data
integrity afforded by the data encapsulation facility of the OODBMS. And so on.
Historically, database processing began with
only one choice: navigational DBMS. Currently,
some companies do all of their database processing
with relational DBMS while others split it, in
various ways, between navigational and relational.
The question now is whether object–oriented
DBMS is going to serve as a niche type of system
in partnership with navigational and relational
DBMS or whether it will be capable of being the
one database system for all applications, including
providing new, added value to the array of common applications. This is a fertile area for new,
applied research in which the academic database
community could make significant contributions.
Mark L. Gillenson
Associate Editor
Journal of Database
INDUSTRY
Management
NEWSBRIEFS
This section includes information collected from several sources, including press releases. The information was chosen by the Journal
of Database Management editorial staff for its interest to our readers. Inclusion does not imply endorsement of the products or services
by either the Journal of Database Management or Idea Group Publishing.
IISI and Infosoft Pilot Test IMS
Migration
generates a list that match those criteria. The user can then
browse through those files to find the one needed. For more
information, contact Martin Steinberg at 303/689–9998.
Climbing out from 15 levels of an IMS hierarchical database
and climbing into the free information flow of a relational
database is today’s impossible dream— until now. In a recent
pilot test, 25% of an IMS/DB application was successfully
migrated to an HP 3000 business server using the HP
ALLBASE/SQL relational database. Test performance levels proved comparable to the original system. The new data
structure allows application development in a relational database environment.
System integration and consulting firm, Innovative
Information Systems, Inc. (IISI) of Norwood, MA. worked
with Infosoft of Dortmund, Germany and Hewlett–Packard
Company to develop and apply new modules for the Conveyor migration tool that would support HP’s internal IMS
conversion project. The effort produced the ability to migrate
applications and data. Converting to a relational database
such as HP ALLBASE/SQL reduces system costs and avoids
rewriting code. For more information, contact Lyn S.
Branscomb at 508/435–1585.
New SQR3 Windows Report Writer
Designed for Electronic Documents
ISYS/NetSpace Introduced
ISYS (Odyssey Development) and Avail Systems recently introduced ISYS/NetSpace, an integration of the ISYS
information retrieval application with Avail’s NetSpace hierarchical storage management (HSM) system, which expands
the storage and retrieval capabilities of all NetWare file
servers.
NetSpace increases the amount of space available on a
file server’s primary storage devices (hard disks) by migrating rarely accessed and old files to secondary storage such as
optical jukeboxes and tape libraries. This migration is controlled by the age of the files, the amount of space available
on the file server, and other rules determined by the system
administrator. Files can be migrated to different media to
optimize storage costs and access time.
ISYS is a high performance information retrieval and
management system. It reads files in their native formats and
generates indexes that allow rule–based queries. Any file can
then be retrieved without knowing its file name. The user
supplies some information about the file’s content; ISYS
Customers of the new SQR3 Workbench for Windows, The
Universal Production Report Writer for the Windows environment, can disseminate even the most complex reports via
e–mail to other PCs, saving time and eliminating paper
documents. And, according to SQR3 developer, MITI,
buyers can send reports electronically without additional
charge because with each copy of SRQ3 Workbench for
windows, MITI will license the purchaser for unlimited use
of the SQR3 Viewer.
SQR3 Workbench, The Universal Production Report
Writer, is a multidatabase report writing environment which
operates on multiple platforms, including PC/Windows. SQR3
is based upon a powerful, 4GL–like procedural language
which allows reports to be written in half the time and with 50
percent less code than previous–generation report writers.
While most report writing tools are focused on either sophisticated developers or low–end general users, SQR3 bridges
the gap with a combinations of true productions capability
and ease of use. SRQ3 Workbench for Windows is priced at
$500 in single units. Everything required for unlimited use of
the SQR3 Viewer is supplied with the product. For more
information, contact Tara Stoutenborough at 714/957–8880.
Windows Functionality Comes to the
AS/400
Business software developer, ITS, Inc. has introduced
400/Manager, the first Microsoft Window–based database
administrator for the AS/400 System on the market. 400/
Manager brings the Windows environment with all its inherent benefits to the AS/400. System administrators and
NewsBrief section was prepared by Tricia Gladney, Idea Group
Publishing.
Summer1995
33
Journal of Database
nontechnical users can search libraries, browse physical or
logical files, and their members through tiered graphical
views. Then users can browse, view, add, delete, copy, paste,
and sort data in one or multiple files with point and click, drag
and drop Windows functionality. AS/400 data can also be
copied “on the fly” between multiple files or into any other
Windows–based applications.
400/Manager is a plug and play application, which
resides entirely on the PC/LAN and does not require any
change or additional software to an existing AS/400. Working with AS/400 files and data entirely at the PC workstation
level creates a client/server environment that significantly
minimizes the day to day administration workload on the AS/
400. 400/Manager security authorization is based on AS/400
log on, permitting users to execute only previously authorized functions. System administrators and nontechnical
users with proper security authorization will now be able to
easily access, manipulate, and sort data “on the fly”. For
more information, contact Jae Y. Hong at 504/539–9300.
WatchWorks 3.0 Provides
Comprehensive Event-Monitoring and
Expert Advice for ORACLE Databases
Eventus Software, Inc. is now shipping WatchWorks
3.0, an event-monitoring and expert-advice tool for ORACLE
RDBMSs. Developed for the proactive management of
ORACLE databases, it helps eliminate costly downtime and
performance problems by forewarning administrators of
potential database and systems troubles.
This latest release of WatchWorks combines all of the
various monitoring disciplines, including event detection,
automatic corrective actions, notification, storage of historical data, trend analysis, on-line monitoring, reporting and
expert advice.
Through both pre-defined and user-defined events and
automatic corrective actions, WatchWorks 3.0 will automatically call a beeper number, send e-mail messages and
faxes, or print specific information to a log file in response to
a specified event. It will even restart a failed database or run
a batch program without user interaction. For more information, contact Mike Burkland, Eventus, 415/871-0700.
Look into the Architectures of Databases
An examination of the architectures of database, text
retrieval, and knowledge-based machines and how specialpurpose hardware addresses current database issues to im34
Vol. 6 , No. 3
Management
prove their functionality is the focus of a new book published
by IEEE Computer Society Press. Parallel Architectures for
Data and Knowledge-Based Systems, edited by L.L. Miller,
A.R. Hurson, and S.H. Pakzad, presents several new topics
and discusses their impact on the next generation of specialpurpose architectures.
The book’s introduction details the present and future
of database machines, as well as the effects of artificial
intelligence on database technology. Its text examines data
models that are receiving special attention today and provides an overview of the current trends in modeling data and
knowledge-based machines. The book introduces several
database machine designs and describes recent research
activities in these areas. It also examines text retrieval machines, the current status of the database machine in the
marketplace, the link between AI and database systems, and
architectures that are designed to support knowledge bases.
The book concludes with a look into several new directions
that are influencing software system research in the database
arena. For more information contact IEEE Computer Society at 714/821-8380.
Virtual SQL—Open Three Tier
Architecture
Plans to build Virtual SQL have recently been announced by software industry veterans, David Taylor and
Max Rosenblatt. Virtual SQL is the key missing component
in an open, three-tiered, client/server development environment. The developers claim that Virtual SQL can bring 80%
of the benefits object-oriented three-tier architectures to
today’s standard SQL tools.
Virtual SQL is a virtual database sandwiched between
standard client application development tools such as
Powersoft’s PowerBuilder and standard SQL database servers such as Sybase. The virtual database includes tables with
extended attributes, updatable views, data integrity tools,
and methods. It looks like another SQL database to the
application software and like another application program to
the database. The virtual database can deliver many of the
benefits of closed, proprietary object-oriented development
environments without giving up standard SQL tools. These
benefits include application partitioning, automatic data validation, and code re-use. Virtual SQL is targeted at large,
complex, mission-critical, transaction processing applications built using client/server technology. For more information contact Max Rosenblatt at 617/259-9594.
Related Content
Syntactical and Semantical Correctness of Pictorial Queries for GIS
Fernando Ferri and Maurizio Rafanelli (2005). Encyclopedia of Database Technologies and Applications
(pp. 671-676).
www.irma-international.org/chapter/syntactical-semantical-correctness-pictorial-queries/11222/
A Crash Course in Metaphysics for the Database Designer
John M. Artz (1997). Journal of Database Management (pp. 25-30).
www.irma-international.org/article/crash-course-metaphysics-database-designer/51186/
Enhancing Decision Support Systems with Spatial Capabilities
Marcus Costa Sampaio, Cláudio de Souza Baptita, André Gomes de Sousa and Fabiana Ferreira do
Nascimento (2007). Intelligent Databases: Technologies and Applications (pp. 94-116).
www.irma-international.org/chapter/enhancing-decision-support-systems-spatial/24231/
Business Information Integration from XML and Relational Databases Sources
Ana María Fermoso Garcia and Roberto Berjón Gallinas (2009). Database Technologies: Concepts,
Methodologies, Tools, and Applications (pp. 1959-1983).
www.irma-international.org/chapter/business-information-integration-xml-relational/8014/
Applying Ontologies to Business and Systems Modelling Techniques and Perspectives: Lessons
Learned
Peter Green and Michael Rosemann (2004). Journal of Database Management (pp. 105-117).
www.irma-international.org/article/applying-ontologies-business-systems-modelling/3312/