Team for Capella 1.3.0 (2019-01-25)
Compatibility: Capella 1.3.0
What's new
The Team for Capella 1.3.0 major release brings several major changes (compared to Team for Capella 1.2.1):
- Compatibility with Capella 1.3.0.
- The lazy loading has been activated for shared representations: the project opening is faster as no representation data are loaded. This loading is done only when the application requires it: open/copy/export/delete representation, ...
- The change management has been improved to allow a fine-grained change tracking and a better integration with Mylyn Tasks:
- The Commit History View has been improved to filter the commits list in relation to the selected elements and to show the change set of each commit.
- The Server is now configured in Auditing mode: the history of every change is kept.
- See how it works in the Team for Capella section of the What's new in Capella 1.3 ? webinar.
- The Diff/Merge limitation (Team for Capella 1.2.x) is no more effective: while comparing a local project to a connected project or between two connected projects, no differences will be shown between representations if they are identical.
- The Importer job and the import wizard do not require a valid licence nor a token to be allowed to import data from the server anymore.
- Move from CDO 4.6 to CDO 4.7.
More details can be found in the integrated documentation.
Important note
When updating from 1.1.x or 1.2.x to this new version, it is recommended to use a new workspace: durable locks have been disabled by default but older versions used to activate this feature as a user choice. It is also possible to clean the old workspace by removing the PREF_ENABLE_DURABLE_LOCKING=true preference value from the workspace metadata located in: .metadata\.plugins\ org.eclipse.core.runtime\.settings\fr.obeo.dsl.viewpoint.collab.prefs file of the workspace.
Fixed Issues
This version also introduces several fixes:
- The "Discard remote changes" action now releases the corresponding locks.
- Compatibility checks have been improved at connection.
- Permissions can now work with models whose path contains spaces.
- The "override existing resource" capability checks done by the Export wizard have been corrected to support servers configured with user profiles.
- User Management and Durable Locks Management views only display the elements of the selected repositories.
- A few other minor fixes around performance and stability of collaborative features or the compatibility between Capella features and the collaborative add-on.
- Documentation minor fixes and improvements
Focus on the lazy loading of shared representations
This new version dramatically reduces the time spent to open a Connected Project containing many representations.
Now only the semantic model (containing the data) is loaded when the project opens. A representation (diagram, table or tree) will not be loaded until the user opens it.
As shown on the figure, opening a shared Connected Project is now far quicker than opening a local one and independent from the number of representations. This figure is based on a Capella model containing 24 000 elements and 64 diagrams containing 300 graphical elements each.
Instead of waiting for the loading of all the representations at once, the time is distributed on each representation's first opening. This small additional time depends on the number of graphical elements.
The gain is also noticeable on the memory footprint. Opening the Connected Project will only consume the memory corresponding to the semantic model, while the memory corresponding to the representations is consumed only on their first opening.
This new mechanism is activated by default, but it can be deactivated if you prefer to use the previous loading strategy.