28-29 January 2003, Lexington, MA, USA

Group Membership:

Company	Attendees	Location	Notes
IBM	Kris Kobylinski	Toronto
	Richard Duggan	Toronto
	Marcelo Paternostro	Toronto
	Harm Sluiman	Toronto
	Valentina Popescu	Toronto
	Clay Williams	Watson
Rational	Sam Guckenheimer	Boston
	Joe Toomey	Boston
	Will Goddin	Boston
	Serge Lucio	Boston
	Julien Canches	Toulouse
	Sylvain Duguet	Toulouse
	Dominique Guilbaud	Toulouse
	Kim Coleman	Cupertino
	Allan Pratt	Cupertino
	Dave Scott	Raleigh
	Kent Siefkes	Raleigh
Scapa Technologies	Mike Norman	Edinburgh	Project Lead
	Ian Hunter	Edinburgh
	Neil Sanderson	Edinburgh
Parasoft	Gary Brunell	US	Apologies Received
Telelogic	Eric Samuelsson	Sweden	Apologies Received

Overview of meeting

28^th January

· Introduction

· Hyades Launch material

· Vendor Perspectives

· Test Profile discussion, based upon current draft

· Differences from ComponentTest FW

· Agree work-packages, confirm short-term objectives within work-packages and allocate individuals to work-packages

· Work packages are:

· Model

· Reference Perspective and UIs

· Data Capture/ Communication

· Execution Environment Control / Mgmt

· Other services (Real-Time Model Export)

· Discuss Model work package

29^th January

· Discuss Model work package

· Split to discuss

· Reference Perspective and UIs

· Data Capture/ Communication

· Execution Environment Control / Mgmt

· Other services (Real-Time Model Export)

· Practical issues: repository, builds, testing, coding standards

Work Packages

The following work packages were defined. Reports are due from each by Feb 12 or before.

Work package	Elements	Membership /Leader	Objectives
Model	· Test Model · Trace Model · Statistical Model · Loaders	Harm Sluiman Mike, Allan, Will, Joe, Julien, Marcelo, Clay, Dave, Ian, Neil, Kent, Valentina, Sylvain, Serge	· Respond with a concrete set of responses to OMG Test Profile (version to be released Jan 24). · Make model complete enough to avoid extensions · Requirements for EMF group (team, CM, persistence, scalability, efficiency, behavior, etc.) · Qualities of service
Reference Perspective and UIs	· GUI Model & extension points · GUI Widgets (?)	Valentina Popescu Mike, Will, Joe, Marcelo, Dave, Neil, Kris, Serge, Sam, Dominique	· Use cases · Extension points (or replacement) and sufficiency · Tool coexistence · Common widgets · Requirements for Eclipse group
Data Capture/ Communication	· Data Capture · System Resource Monitor · JVMPI/JSR163 · Trace Collector · Log Collection Framework · Data Injection Framework · Trace Correlation Service · BCI/OCI plug-ins · Data communication · Data Collection Architecture · Standard communication APIs (Data collection) · Standard communication APIs (Master agent)	Richard Duggan, Allan, Kim, Ian, Dominique, Kent, Richard, Sylvain	· Assign ownership for each collector · Define plug-in mechanisms · Define services to provide (e.g. security?) · Prioritise target environments
Execution Environment Control / Mgmt	· Master Control Interface · Agent Control Interface · Testability Interface (?)	Kent Siefkes Joe , Neil, Kris	· Define and Agent Control Interface · Requirements on data communication architecture · Testability i/f requirements / services
Other services	· Real-time model export	Harm Sluiman, Dave, Mike	· Define real time model export functionality (push and pull)

Process issues

· Rationale

· Trans-corporate, trans-national software projects don’t just hatch themselves, you know

· Schedules, status, and milestones

· Via Eclipse web site

· Weekly management teleconference

· Management, not technical

· Conduit to marcomms

· Handles process issues either directly or through delegation to subgroup

· Minutes published to web, public newsgroups

· Two business day review/vetting period of minutes

· Calls are open to the group

· Up to one hour, Monday, 12n Eastern (5p Scotland, 6p Sweden & France, 9a Pacific)

· Technical discussions

· Technical discussion in public newsgroups

· Phone discussions open to the group or wider if appropriate

· Summarize phone discussions to news groups if appropriate

· Emphasis on open-ness, even if thought processes are not complete

· Mgmt meeting attendees gather status from tech groups, bring them in personally as appropriate.

· Development process

· A range of issues were discussed and the relationship with existing Eclipse practice is to be clarified. Additional postings will be made to the private and public newsgroups.

· Next steps

· Draw together notes from breakout sessions and the rest of this meeting

· Each workgroup make a “plan for a plan.”

· Workgroup discussions become plans by Feb 12th

· Turn it into a project plan for 1.0.

· Interim release of parts of the trace model anticipated in advance of 1.0