Vol. 37, No. 2, 1998 - San Francisco Frameworks 

Preface


Rapid development of business software for critical distributed 
applications is, and has long been, a key aspect of IBM's support for 
businesses and for independent software vendors (ISVs). IBM's San 
Francisco* project, which is the subject of most of the papers in this 
issue, provides that support and is an excellent example of joint IBM
and ISV efforts. More than 130 ISVs have participated in the creation and
development of San Francisco common and domain-specific frameworks. The
frameworks allow each ISV to focus on the specific needs of its customers, 
resulting in products with a robust object-oriented architecture and for a 
networked environment. 

This issue contains an introductory paper on the San Francisco
architecture and frameworks, four papers by ISV developers on their use of 
the frameworks for application development, and a Technical Note describing
a supporting tool. There are also three papers on other subjects: software 
requirements, reverse engineering, and code restructuring. We are indebted 
to V. M. Johnson of the AS/400 Division in Rochester, Minnesota, for his 
foresight in conceiving, and his considerable efforts in developing, the San 
Francisco theme for this issue. 

Bohrer presents an overview of the San Francisco project and its results in 
a paper on the origins of the project, its objectives, the collaboration with 
numerous ISVs, the methodology used to carry out the project, and the 
development of the architecture and frameworks. The key objectives were ease 
of ISV entry into object-oriented techniques, capability for ISVs to develop 
significant business applications, and open frameworks allowing for ISV 
choices and trade-offs. 

The ability to expand the power of accounting systems into the more 
important arena of business enterprise models is the focus of a paper by 
Inman. The San Francisco frameworks are shown to be an effective path 
toward this expansion, allowing for more flexible and capable financial 
systems, serving management needs and supporting business analysis. The 
paper discusses general ledgers as an example in which San Francisco 
frameworks succeed. 

Some ISVs and some of their customers face significant challenges in moving 
their software skills and products to object-oriented technologies, such as 
those underlying the San Francisco architecture and frameworks. Henders 
shows how the use of such technologies was introduced into an existing 
software mix, how the challenges were overcome and the difficulties avoided, 
and how the San Francisco frameworks support an easier transition to the new
object-oriented development methods and technologies. 

Knowledge workers, knowledge applications, and the challenge of making
those applications serve broader business purposes are explored in a paper 
by Retallick and Sanchez. Their claim is that managing knowledge data and 
organizing business activities are rarely amenable to real change, but that 
the San Francisco frameworks offer such an opportunity for significant 
progress toward rapidly developed and integrated knowledge-based systems. 

Van der Salm discusses the issues faced by an ISV when it considers the 
prospect of moving from tried-and-true software methods and tools to new 
ones that offer the promise of faster and more effective software development 
and product delivery. It is particularly important for small ISVs to find a 
path to newer and better technologies for software development, despite the 
risks and costs often assumed with change in small companies. The experiences 
the author relates illustrate the potential of San Francisco frameworks for
ISVs moving to object technologies. 

The last contribution on the San Francisco frameworks appears as a Technical 
Note. The author, Polan, describes how to combine the power of the VisualAge* 
for Java** integrated software development environment, with the structure and 
support of the San Francisco object-oriented, Java-based frameworks. This Note 
is intended to provide technical details for using these two systems together. 

The development of software requirements is a challenging and complex task, 
which is both essential to the value of the system that results and 
fundamentally human in its process and progress. Crowston and Kammerer report 
on a study of how requirements analysts can manage their efforts to overcome 
the challenges and produce better results. The authors used coordination theory 
to understand the management of dependencies and collective mind theory to 
understand how individuals affect the work of a team. 

Aiken builds on the general understanding of reverse engineering of programs 
to describe the newer area of reverse engineering of data, which combines 
structured data analysis and data management principles. An activity model and 
a data model are developed to reach the goal of integration of enterprise-wide 
data. The author presents four brief scenarios in which data reverse engineering 
was used to solve or resolve enterprise data issues, and proposes the use of 
these techniques for the Year 2000 transition. 

Dynamic profiling tools have been used to support compiler-based optimization 
of applications by exposing their behavior during execution and particularly 
their uses of execution paths and data. Schmidt et al. thought that these same 
techniques could be applied to code reordering of operating systems software with 
similar beneficial effects on execution behavior. Their results were significant 
and argue well for reordering operating system code using profile data. 

The next issue of the Journal will be a special issue on the Java language, 
application development in Java, and IBM's support for Java. 

Gene F. Hoffnagle 
Editor

*Trademark or registered trademark of International Business Machines Corporation. 

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