Presentation Abstract

The rise of Internet usage, and in particular the explosive growth of electronic commerce is raising many problems for software researchers and practitioners. The amount of Internet commerce has been literally rising exponentially (doubling about every 9 months since 1995), and the number of participants, sites, domains, etc. Are also doubling rapidly. We are thus led to contradictory requirements: many companies have been changing their e-business strategies every 6 months, their firms are increasingly dependent on proper functioning of these business processes and systems, yet building complex highly reliable software still takes many months (if you are lucky) or years.

It turns out that e-commerce pushes limits of the state of the art in many areas, including:

• Distributed systems (with millions of servers, son to be billions of participants)
• Transactions (long lived changing multiple participants)
• Distributed massive content management
• Interface paradigms (how to manage mobile devices, speech, etc. in a world of legacy data)
• Performance and Reliability management (guarantees and expectations in a dynamic world)
• Security, Authentication
• Search (for processes and activities as well as complex content)
• New application models and structure
• Software engineering

On the last point, there are many challenges facing the software engineer. Classic processes and experience apply to centralized design, control and management. There is less support for a world of independent creation, ownership, and redefinition. Business models (which will themselves be software objects) will be defined and implemented in terms of other services and applications, each of which might come from several sources and be administered independently.

Yet it is necessary to have expectations of performance, security, reliability and availability. We must manage the specification, design, creation, testing, and dynamic reconfiguration of key processes in such a fluid world.

This talk will address these challenges to the state of the art and practice, offering questions and some answers.

About the Speaker

Feldman did his academic work (AB, Princeton and PhD, MIT) in astrophysics and mathematics. He is a Fellow of the IEEE and a Fellow of the ACM. He has been a member of the Board of the Computing Research Association and chair of ACM SIGPLAN and is founding chair of the ACM SIG on E-Commerce.

He was a computer science researcher at Bell Labs and a research manager at Bellcore before joining IBM in mid-1995. He has published research in software engineering (and was the creator of Make), programming languages, scientific computing and other areas of computer science. He was also architect for a large new line of software products at Bellcore.

At IBM's T. J. Watson Research Center, Feldman leads a department doing research in a wide variety of network-related technologies and application enablers, including electronic commerce, parallel databases, anti-virus technology, and advanced internet multimedia. He is also the Director of IBM's Institute for Advanced Commerce, an organization created to increase IBM's intellectual leadership in e-commerce, and to forge better connections to the outside research world as well as to accelerate creation of new technologies for support of e-Business.