ICSE2000 home

ezine home


Grady Booch's "Future of SW"

Editor's Day 3 Message

Bricks and Dots with Chris Horn

Talking with Grady

Workshop 9: CBSE

Workshop 11: SEotI

Doctoral Workshop

Workshop 10: WSEF

Workshop 14: WISE 3

ICSE 2001 Preview

CNET.COMS Polls

Ten Visionaries of the 90's

Success Stories

Trends

Unsung Heroes

Student Volunteers

ICSE 2002 Pre-Preview

Poll Results

WOW Memories, Thanks

Even More Real Action from ICSE2000!

Window on the World

Issue 3


Back to WOW home

Workshop 10: Exchange Formats

By Susan Elliott Sim, Rainer Koschke, Ric Holt

Researchers in computer-aided software engineering (CASE) tools and computer-aided reverse engineering (CARE) tools have recognized a standard exchange format as a means for improving the state of the art. This workshop was held as part of an on-going effort to develop one. The reasons for having a standard exchange format are evident in the reasons that the participants gave for attending the workshop: 1) they wanted to be able to make complementary tools work together more smoothly; and 2) they were tired of writing parsers/analyzers and wanted to avoid writing another one.

For the first time, this workshop brought together members from disparate communities to share experiences and work together. The workshop summarized previous work in this field and re-visited XML, XMI, UML, and CDIF. We discussed concrete schemas for high-level information, such as class diagrams or architectural information, and for low-level information, such as abstract syntax trees. We also debated the concepts and the mechanisms to specify meta schemas.

During the last two hours of the workshop, we departed from the planned presentations and discussions. For one hour, we broke into small groups, each focusing on a single topic.

We discussed high-level schemas, C++ schemas and APIs, and notation for exchanging schemas. The groups identified requirements, made prescriptions for progress, and wrote WILL-do lists. Afterwards, each group presented their results to the whole workshop. At the end of the day, virtually all participants committed to refining GXL, Graph eXchange Language, a format in-progress. GXL uses XML syntax and is used to encode attributed graphs. The groups who committed to working together further were University of Stuttgart, Bell Canada, IBM Canada Ltd., Mahindra-British Telecom, University of Waterloo, University of Koblenz, UBW, Philips Research, University of Victoria, and Nokia.

For more information, visit the workshop homepage at: http://www.cs.utoronto.ca/~simsuz/wosef or talk to one of the co-chairs.

 

 

Designer and Webmistress
Caroline Kehoe