Jump to: navigation, search


Revision as of 14:53, 3 September 2008 by Recoskie.ca.ibm.com (Talk | contribs) (Agenda)

CDT Summit Fall 2008

The CDT Fall Summit will held Tues-Thurs, Sept 23-25, 2008 at the IBM Software Lab in Toronto (well, Markham really), Canada. This summit, as always, is intended for contributors to the CDT to review their plans and discuss issues with the rest of the CDT community. It is intended for contributors, i.e., people working, or at least planning on working on the CDT. Space is limited and we're happy to meet with users and adopters of the CDT at EclipseCon next March.

To register, please add your name to Attendees list below. Also, feel free to add to the Agenda to let us know what you'd like to talk about.

There is no fee for participation. However, sponsorship from participating vendors would be appreciated and help fund events at the summit.

Driving directions are here.

Hotel availability is here. Most of us will likely be staying at the Hilton which is about a block away from the lab.

Those needing official invitations for visa purposes to get into Canada should contact Chris Recoskie from IBM Toronto.


  • Project Management
    • Improving the CDT development process to improve quality and communication.
    • What version is June 2009? 5.1 or 6.0?
    • How to improve our JUnit coverage. GUI testing?
  • Debug
    • Migrating the Debug Services Framework (DSF) into a component of the CDT
  • Build
    • How to clean up the build system architecture so we don't have to do it again
      • How to add a new type of builder?
        • Right now this is very hard to do. (See Remote Make Builder)
      • Reusing the wizards
        • Right now this is very hard to do. (See New Remote C/C++ Project wizard)
      • Harmonizing Project Templates and custom wizard pages
        • People are confused that there are two ways to extend the flow of the wizard that are completely orthogonal.
      • Scanner Info/Discovery
        • How does it work? How should it work?
      • Should we continue to support generating a makefile?
      • Concurrency Issues
        • Deadlocks, deadlocks, deadlocks
        • What is the reason most of these are happening? How can the architecture be changed to help alleviate this?
      • Scalability Issues
        • XML objects from project file held in memory
        • Build model is always loaded on startup, not truly "on demand".
        • The way modifying/canceling settings is handled is inefficient.
  • Resources
    • Collect input for the e4 flexible resources project
    • Any workaround we want to do for June 2009 (e4 is June 2010)
  • Index/DOM
    • Can we finally conquer templates?
    • How can we make ambiguity resolution fail-safe?
    • Do we have a plan for supporting C++0x?
    • Can we add some fault tolerance? (Missing include search path, ambiguities found by name resolution).
    • Can we provide an Outline View for inactive code (#ifdef ..)?
    • Indexing EFS based projects.
    • Indexing files that are sensitive to the macro-dictionary provided by the context (e.g. stddef.h)


  • Doug Schaefer (Wind River)
  • Sergey Prigogin (Google)
  • Mike Kucera (IBM)
  • Vivian Kong (IBM)
  • Chris Recoskie (IBM)
  • Ken Ryall (Nokia)
  • Elena Laskavaia (QNX)
  • Dominique Toupin (Ericsson)
  • Mikhail Khodjaiants (ARM)
  • Pawel Piech (Wind River)
  • Jeff Johnston (Red Hat)
  • Andrew Ferguson (Symbian)
  • Markus Schorn (Wind River)