Vol. 37, No. 4, 1998 - Nontopical issue 

Preface 
  

This issue contains five papers and two Technical Notes on a variety
of subjects: advances in software metrics, Component Broker support
for JavaBeans**, dynamic tables in relational databases, software 
development performance, management of intellectual capital, Web 
service aggregation, and the thirtieth anniversary of IMS*. 

As a result of interest in software measurement, software development
organizations have been using a significant number of metrics and
collecting substantial quantities of information about their processes, 
methods, tools, and results. As their methods, data quality, and needs 
change over time, these data collections and metrics should be re-examined. 
An organization can, for example, discover measurement goals that are 
not being met by the metrics or the data, goals that could be met by 
the collected data but where the data have not been exploited, and 
data that are collected and metrics that are in use but do not support 
organizational goals. Mendonca et al. combine top-down and bottom-up 
techniques to review, rationalize, improve, and rejuvenate the use of 
metrics and data within an organization. 

Codella et al. have developed an understanding of IBM's Component 
Broker and Sun Microsystems' Enterprise JavaBeans** that shows how 
these two component models can support, augment, and enhance each other 
in their respective roles as part of object-oriented systems. The 
authors demonstrate the combined ability of these two middleware layers 
to provide an object environment that has such valued features as broad 
scalability, transactional support, and persistence. They also describe 
and argue for further extensions to this cooperative and complementary 
situation that will cement the relationship and provide needed 
functions for both middleware systems. 

Recently developed and standardized additions to the Structured Query 
Language (SQL) environment provide users and programmers with powerful 
newcapabilities and generalized table concepts. As Fuh et al. point out 
in their paper, these include what they refer to collectively as dynamic 
tables, or run-time, explicitly defined derived tables: user-defined 
temporary tables, transition tables, user-defined table functions, and 
table locators. The authors have constructed an effective prototype in 
the context of IBM DATABASE 2* (DB2*) Common Server that demonstrates 
how to perform compile-time and run-time processing for these new tables, 
while creating the required linkages between the tables and their references. 

Sawyer and Guinan have studied the production methods and social 
processes of 40 commercial software development teams. The authors 
measured and assessed the effects of those methods and processes on 
software quality and team performance. They found little support for 
methods or processes as the fundamental drivers of quality and 
performance, although the data show social processes as more important 
than production methods. But about 75 percent of the variability in 
quality and performance is not accounted for by these two factors. 
Suggestions and speculations are made about the missing factors. 

Huang has authored an essay introducing a significant part of the 
knowledge management arena: the capture, exploitation, and management 
of intellectual capital assets. He describes IBM's program in this 
arena--the Intellectual Capital Management program--and provides its 
motivation in areas such as competitive responsiveness, team productivity, 
and core competencies. The author is particularly interested in the 
creation of, management of, and capitalization on collective knowledge 
across large organizations, such as IBM. The methods and tools for 
such capitalization have been and are being built and used today to 
provide competitive advantage to organizations, and they are described 
in this essay. 

In a Technical Note, Zhao presents Web service aggregators and, in 
particular, IBM's WebEntree. Web service aggregators serve the users 
as intermediaries that provide a single view of many diverse and 
distributed Web services. For example, a user can be provided with many 
services, select those of interest, use the services through a single 
interface, and have the aggregator perform such common, repetitious, 
and differing service functions as registration and authentication, 
without repeated user involvement. 

This year marks the thirtieth year since the first time the message 
"IMS READY" was displayed on a customer's computer screen. Few software 
systems enjoy that degree of longevity and success, and few have had 
such far-reaching and long-lasting effects on business data processing. 
In his Technical Note, Blackman retraces those 30 years, describing 
the capabilities, innovations, and history of the IBM Information Management 
System*, better known as IMS, from the computer environments of the 
late 1960s to the Internet of the late 1990s. The author notes that 
today over 90 percent of Fortune 1000 companies use IMS as their 
database management system of choice. 

The next issue of the Journal will be a special issue on Enterprise 
Solutions Structure. 

Gene F. Hoffnagle 
Editor 

*Trademark or registered trademark of International Business Machines 
Corporation. 

**Trademark or registered trademark of Sun Microsystems, Inc.