Panels

 
P1Product Line Scoping in Practice
P2Agile and Product Line Engineering
P3Catalysts and Inhibitors for Momentum in the Software Product Line Industry




P1: Product Line Scoping in Practice


Scoping is a key activity in Software Product Lines but organizations coming from a one-of-a-kind software development background typically have little understanding of how it is done in practice and why it is so important.

In this session our veteran product line panellists are asked to consider what scoping is, how it relates to other product line activities and most importantly how they perform scoping in practice.

The panel discussion will be motivated by a small number of examples but the audience is strongly encouraged to bring their own questions and examples to the session so that the discussion takes the direction most relevant to their interests and needs.

Panelists

Danilo Beuche works for pure-systems GmbH, a specialist provider of tools and services for the application of Software Product Line technology. As well as managing the company Danilo also consults extensively on Product Line Engineering, mainly for clients in embedded industries. Danilo has been a tutorial presenter, speaker, workshop organizer and panellist at conferences such as AOSD, ISORC, SPLC and OOPSLA. He is also author of many articles in scientific journals and software development magazines. Danilo has a PhD from the University of Magdeburg for his research applying Software Product Line Engineering to embedded operating systems development.

Isabel John is a researcher and project leader at Fraunhofer IESE. She is responsible for product line introduction projects in industrial contexts and performs product line technology transfer to software developing companies. For almost 10 years, she has been working in industrial and research projects in the context of software product lines, scoping, and product line requirements engineering and has performed scoping in several different industrial and academic contexts.

Prof. Dr. Klaus Schmid holds a professorship for software engineering at the University of Hildesheim. He is active in the areas of Requirements Engineering and Product Line Engineering since the mid-nineties. Over time he has been involved in numerous research and industrial projects. His specific interests are in value-based product line engineering and the engineering of adaptive systems. He has authored numerous refereed papers on these subjects and has been responsible for several workshops and tutorials on international conferences in this area. Previously, he was also department head for requirements engineering and usability engineering at the Fraunhofer Institute for Experimental Software Engineering. He received a diploma degree and a Ph.D. degree in computer science from the University of Kaiserslautern.

Christa Schwanninger works as a Senior Research Scientist at Siemens AG Corporate Technology in Germany where she consults for Siemens business units in software architecture and product line engineering. In the past 10 years she has been actively involved in the development of systems for the telecommunication, automotive, automation and medical devices domains.

P2: Agile and Product Line Engineering



Agile software development and software product line engineering are both successful approaches to the developmentof software-intensive products, but can the planning drivenapproach of product lines be effectively integrated with theopportunistic approach of agile techniques? This is a difficultquestion to answer since each broad strategy refers to a familyof possible methods. It is a sufficiently interesting question to have been the subject of a workshop at SPLC 2006 and ajournal special issue [1].

In order to begin exploring this question as a community, a working session will be held during the technical programof SPLC 2008. In a working session, the facilitator stimulatesdiscussion by posing initial questions and assertions to whichthe audience reacts. The interests and experiences of thegroup will guide the discussion from there. In general, weare interested in generating discussion on:

  • your experiences in using agile techniques in actual product line operation
  • reasons why you rejected the idea of using agile techniques
  • what adjustments you have made to software product linepractices to make an agile product line successful

The discussion will continue after SPLC 2008 via a wiki forumand a follow-up session at SPLC 2009 will be proposed tocontinue discussions.

The agile approach can be characterized by the four valuesspecified in the Agile Manifesto [2].

  1. Individuals and interactions over processes and tools.
  2. Working software over comprehensive documentation.
  3. Customer collaboration over contract negotiation.
  4. Responding to change over following a plan.

Additionally, Hanssen and Fægri use the twelve agile principles, which are also a part of the manifesto, to provide acomparison with software product line principles[3].

The software product line approach can be characterizedbroadly by examining elements of the SEIs Framework for Product Line Practice [4]. By choosing different specific practices in each practice area within the framework, very differentapproaches result. Fundamental product line principles include

  • the separation of the roles of core asset development andproduct development, and
  • the interaction of both roles with management to ensurecontinuity across projects.

A software product line organization works to achieve strategic levels of reuse through the identification of common behavior  and the constraint of variable behaviors.

The agile culture would appear to emphasize the autonomyof the individual while the product line culture would appearto rely on the common vision developed from centralized planning. To balance these, one approach, is to consider planning as a strategic process and software development a tactical process. This allows for both stability and flexibility,each to a limited extent.

If we consider agility as rapid response, the above approachfalls short. If the core assets are taken as a strategic stockpileof stable plans and flexible software, changing the stable assets may take too long and changes only to the flexible assets, overthe longer life of a product line, will lead to inconsistent assets. A true integration of the two approaches may be requiredfor a totally satisfactory solution. Such an integration wouldstill operate at both the strategic and tactical levels but wouldcrosscut the core asset, product development, and managementareas.

Model-driven development often is seen as strategic since it requires a long term view and tactical since changes canbe accomplished quickly by changing the existing model andregenerating the asset. Model-driven development is often seen as limited to software development, but in a productline context, model-driven techniques, such as template-drivengeneration, can be applied to plans and documents as well as software. The initial creation of models is strategic because itrequires a longer term view that extends over all the productsin the product line. Their use and subsequent maintenanceis more tactical in that they are exercised over the limitedscope of a single product. Both customization changes andmaintenance changes can be made to core assets for individualproducts more rapidly because the model-driven approach typically has more automated support for consistency checking,automatic (re-)generation, and testing.

Join us to comment on these approaches and to suggest yourown. This is an opportunity to be a participant in the SPLC2008 program, not just a spectator.

REFERENCES

[1] K. Cooper and X. Franch, “Agile product line engineering,” Journal of Systems and Software, vol. 81, no. 6, 2008.

[2] agilemanifesto. [Online]. Available: www.agilemanifesto.org

[3] G. K. Hanssen and T. E. Faegri, “Process fusion: An industrial case study on agile software product line engineering,” Journal of Systems and Software, vol. 81, no. 6, 2008.

[4] P. Clements and L. M. Northrop. Framework for product line practice. [Online]. Available: www.sei.cmu.edu/productlines.


P3: Catalysts and Inhibitors for Momentum in the Software Product Line Industry

Panel Introduction

For Software Product Line Engineering (SPLE) to become a standard and accepted practice in the industry, it must be widely embraced by large corporations, major software tool and service vendors, industry analysts, and technology press. We have seen steady but slow progress in all of these areas over the past few years. What can we do to better control or influence the future trends?

Technology adoption at the industry scale is about building and sustaining momentum.  Through a feedback loop comprising successes and failures, credible validation and invalidation, credible communication, leaps of faith, competitive opportunities, perceptions about perceptions, and so forth, momentum is gained or lost. Catalysts and inhibitors in the loop ultimately determine whether momentum will accelerate or be dampened.

In this panel, we bring together an esteemed group of industry leaders from the key areas that impact momentum of new technology adoption, all of whom have been involved in the early stages of the industry’s SPLE momentum. In this panel we will search for insights across a broad industry spectrum into the current catalysts and inhibitors, as well as opportunities for the Software Product Line Conference (SPLC) community to positively contribute to the momentum.

Discussion Threads

Some of the questions addressed by the panel include:

  • What is the current state of SPLE acceptance and adoption by the broad system and software engineering industry?
  • How quickly can SPLE become a standard practice in the industry?
  • What are the past, present and future catalysts that facilitate and accelerate the momentum?
  • What are the past, present and future inhibitors that dampen or stall the momentum?
  • How do the different roles in the industry – large corporations, major tool and service vendors, industry analysts, technology media – impact this and what is their relative importance?
  • To what extent are each of these roles having positive impact?
  • What can members of the SPLC community – both researchers and practitioners – do to help catalyze the momentum?

Insights Offered

For the SPLC community to fully impact the future of Software Product Line Engineering in the industry, we must now understand and address many factors that have traditionally been outside of the SPLC focus.

Expected insights for practitioners:  Perceptions, trends and opportunities that can impact the technical and business SPLE efforts within their companies.

Expected insights for researchers. Perceptions and trends that can impact research agendas.

Panelists

Charles Krueger (moderator), Founder and CEO, BigLever Software, the industry leading provider of SPLE framework, tools, methodology and services.

John Carrillo, Senior Director of Corporate and Product Strategy, Telelogic, an IBM company.  Leading the strategic initiative to bring SPLE practice into the tools and practices offered by Telelogic and IBM Rational.

Bola Rotibi, Principal Analyst, Macehiter Ward-Dutton. Focused on issues and products concerning the software development and delivery processes and technologies, including a focus on SPLE.

James Cezo, Principal Member Engineering Staff, Lockheed Martin. Leading the adoption of System and Software Product Line Engineering practice for Lockheed Martin Maritime Systems and Sensors.

RSS icon
View a printer-friendly version of this page
Page last updated: 13/09/2008