Notice: this Wiki will be going read only early in 2024 and edits will no longer be possible. Please see: https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/wikis/Wiki-shutdown-plan for the plan.
CDT/summitecon2013
Contents
CDT Summit at EclipseCon 2013
We have scheduled a one-day CDT summit at EclipseCon 2013 in Boston, MA, on Thursday the 28th of March, 2013. To help plan for this meeting, please sign up and add any agenda items you'd like to see discussed.
EclipseCon information is here: http://www.eclipsecon.org/2013/
Open to all, even if you didn't sign up :).
Time: 10:00 a.m. to 3:30 p.m.
Room: Beacon Hill 2 & 3
Attendees and Interests
- Marc Khouzam - Ericsson
- Debug, Multicore, Visualizer
- Martin Oberhuber - WindRiver
- Quality, Debug (TCF), Usability, Remote Edit-compile-debug
- Sergey Prigogin - Google
- Refactoring, C++11 features, Codan
- Glen Anderson - Analog Devices
- Debug, Usability, Multicore
- Dominique Toupin - Ericsson
- Josh Kruck
- Debug (DSF, MS), Refactoring
- Bill Swanson - Tilera
- Visualization, Multicore, Eclipse/CDT Redistribution
- Doug Schaefer - QNX
- Everything, but currently working on improving Qt support
- Also if we don't get many more people, we can meeting in one of the evenings as well
- Jeff Johnston - RedHat
- Jim Adams - Analog Devices
- Reibert Arbring - Ericsson
- Norman Yee - Analog Devices
- Leo Treggiari - Intel
- Vladimir Prus - Mentor Graphics
- Eric Cloninger - Klocwork
- Static analysis, refactoring, Android
- Mike Wrighton - Mentor Graphics
- Greg Watson - IBM
- Brian Watt - IBM
- Teodor Madan - Freescale
- Jesper Eskilson - IAR Systems
- MBS, TCF, indexer
Agenda
- Eclipse 4.x discussion:
- reactions to 4.x platform
- CDT compatibility and/or upgrade plans
- effect on other projects, Visualizer, etc.
(e.g. if CDT retains 3.x compatibility, related plugins may need to as well)
- Advertising new CDT features
- Dialog popup with "Did you know?" and "Don't show this message again"
- Other?
- Debug
- Quick Launches (Local C/C++ App, Remote C/C++ App)
- UI Junit tests: Dependency injection vs other CDT UI test techniques
- Remote/synchronized projects
- ...
Minutes
- Introductions
- 24 attendees from 14 companies
- E4
- What is the gain to start using E4 APIs?
- Easier to modify the look of the UI -- No need to move CDT to E4 for that
- Uniformity of the look and feel -- No need to move CDT to E4 for that
- QNX feels the UI must be improved -- Does that require E4?
- Once the CDT community agrees we can use E4 APIs, the decision to do so can be left to individual committers. There does not seem to be a reason we'd have to move all of CDT. We can instead take a joint E4 and 3.x approach.
- Is there a reason to avoid all use of E4 APIs?
- We want to keep CDT compiling on 3.x
- We can re-discuss this requirement for the Luna release
- What is the gain to start using E4 APIs?
Decisions
- CDT 8.2 (Kepler) will not use any E4 APIs to keep compatibility with 3.x.
- CDT will be free to use any new Platform APIs as long as it is not an E4 APIs. This will allow others to put CDT and the necessary part of the platform on top of 3.x.
- CDT will not commit to supporting 3.x as a platform.