1.0 Participants

Faculty:

• Gregory Abowd
• Ashok Goel
• Mike McCracken
• Melody Moore
• Colin Potts
• Spencer Rugaber
• Linda Wills

Research Assistants:

• Christian Ertmann-Christiansen
• Dean Jerding
• Bill Murdock

2.0 Ongoing Research & Development

2.1 Task 1 Methodology Descriptions

The MORALE approach to software evolution is the synthesis of several methodologies supporting individual aspects of the evolutionary life cycle. To support this synthesis and to enable effective transfer of our methods, we are writing guidebooks for each of the individual methodologies.

• Subtask 1.1 Inquiry Cycle Methodology Guidebook
  Mission oriented requirements analysis; under development.
• Subtask 1.2 Software Architecture Analysis Method Methodology Guidebook
  Scenario based assessment of software architectures
• Subtask 1.3 Synchronized Refinement Methodology Guidebook
  Reverse engineering and software understanding
• Subtask 1.4 SBF Tutorial and Reference Manual
  Structure, Behavior, Function language for describing software systems and their architectures.
  Initial draft is complete; case study details being added.
• Subtask 1.5 MORPH Methodology Guidebook
  Model Oriented Reengineering Process of Human Computer Interfaces

2.2 Task 2 Tool Development

While MORALE is primarily a methodology project, we are developing several tools to support analysts using the methodologies.

• Subtask 2.1 - ISVis (Interactive Scenario Visualizer)
  A program understanding and visualization tool for viewing and analyzing event logs of programs executing specific scenarios. The tool is now in "demo-able" form and is being used on our Mosaic case study.
• Subtask 2.2 - ScenICView (nee GBRAT/GBRAT) (Hierarchical Goal Representation Editor/Viewer)
  A tool to support an analyst constructing mission oriented requirements. Several tools from previous projects are being integrated and extended to form the first version of ScenIC View. These include the file version of GBRAT, the MSQL version of GBRAT, and a goal hierarchy editor. A conceptual schema for ScenIC has been completed and the first version of ScenIC View will be based on this.
• Subtask 2.3 - SIRRINE (Self-Improving/Redesigning Reasoner Integrating Noteworthy Experience)
  A tool which uses SBF models to perform intelligent redesign of software agents.
  We are converting an existing tool into SIRRINE, and testing the new tool on an example (meeting scheduling).
• Subtask 2.4 - MORPH (Model Oriented Reengineering Process for HCI)
  A tool for reengineering user interfaces.
  The MORPH tool architecture has been further refined; the knowledge base implementation has been completed for user interface abstractions including Java AWT and tcl/tk; testing of MORPH inferencing for transformation is underway; and the rule base implementation is underway.
• Subtask 2.5 - SAAMtool (Software Architecture Analysis Method Tool)
  SAAMtool is a UNIX-based utility that supports the scenario-based impact analysis of SAAM. This release of the tool allows for the generation and annotation of a simple architectural language based on the work done by Abowd and colleagues at the Software Engineering Institute.

2.3 Task 3 Case Study

The primary validation mechanism for MORALE is the case study. The two case studies we are currently anticipating comprise the analysis of an evolving system and an assessment of SBF as a suitable architectural description language (ADL).

• Subtask 3.1 Case Study Schema
  We are constructing a schema by which case studies can be performed. We are also undertaking a pilot study of web browsers to instantiate the schema. The pilot study has converged to be the addition of user configurable viewers to Mosaic version 2.4. In particular, we have used this situation with ScenIC, SAAM, SBF, and architectural extraction. A paper describing our experience is being written.
• Subtask 3.2 SBF Assessment
  SBF can be used as an architectural description language. EDCS includes several other projects developing ADLs. MORALE requires a means for representing architectures so that analysis can be performed. This task assesses the adequacy of SBF for this role.

3.0 Notable Accomplishments & Technology Transition

• We have been interacting with TACOM as a potential customer for MORALE. They have sent us documentation for a specific domain in which they are working, software loader/verifiers for mobile vehicles.
• Rugaber and Moore have made several MORALE presentations to groups at the Georgia Tech Research Institute. One involved participants from the JMASS group at Wright Paterson Air Force Base and the other involved representatives from Northrop-Grumman.

4.0 Progress on Inter/Intra Cluster Collaboration

• Participated in Architecture cluster meeting in Napa Valley, California (Abowd).
• Helping with organization of Atlanta Rationale Capture cluster meeting.
• Balboa tool installed and tested on Mosaic case study. We are looking at how to integrate it into ISVis.
• Began review of Xinotech documentation
• Investigating the integration of multiple stakeholder viewpoints in ScenIC with the negotiation model of Win-Win.
• Investigating use of SAAMtool to support other ADL descriptions, specifically ACME.

5.0 Publications

• Melody Moore. "Rule-Based Detection for Reverse Engineering User Interfaces." Proceedings of the Third Working Conference on Reverse Engineering, Monterey, CA., IEEE Computer Society Press, November 8-10, 1996.
• Spencer Rugaber and Linda M. Wills. "Creating a Research Infrastructure for Reengineering." 3rd Working Conference on Reverse Engineering, Monterey, California, IEEE Computer Society Press, November 8-10, 1996.

6.0 Travel

• Attended the Dallas DARPA PI Workshop Oct 7-8, and served as Working Group Leader for the Legacy Systems and Reengineering group (Moore).
• Moore, Rugaber, and Wills attended ICSM'96 and the 3rd Working Conference on Reverse Engineering, presenting the papers listed above. Rugaber and Wills have initiated an effort to support tighter collaboration among researchers in the program understanding community.
• Abowd met with the Software Architecture Technology Initiative in Austin, Texas, in December. This group is the initiator of the SAAM.
• Attendd SIGSOFT'96/FSE-4 conference, October 14-18, San Francisco (Abowd).

7.0 Miscellaneous/Administrative/Problematic Issue

Our progress on guidebook writing has had mixed results. On the down side, delivery of individual guidebooks for use by third parties has been slow. The reason for this has been our substantial progress with the Mosaic case study which has led us to a deeper understanding of each other's work and which enables tighter integration of our individual methods. On the positive side, we have spoken with several publishers who have expressed interest in the collected guidebooks. I assume that there will be no legal/administrative difficulties with publication, but I will check into the matter further as negotiations proceed.

8.0 Plans For The Next Reporting Period

• Subtask 1.1 completed
• Subtask 1.2 completed
• Subtask 1.3 completed
• Subtask 1.4 refinements
• Subtask 1.5 completed
• Subtask 2.1 completed
• Subtask 2.5 extend SAAMtool to be able to accept architectural decisions in other ADLs.
• Subtask 3.2 assess the SBF theory and the SIRRINE tool for the examples of Mosiac and meeting scheduling.
• Subtask 3.1 completed
• Subtask 3.2 completed
• Participation in Rationale Capture Workshop in Atlanta, March 5-7. Boehm, et al. from ISI will spend
  an additional day with us on the fourth. Johnson from ISI will spend additional time with us on the
  afternoon of the seventh.