| [mdt-ocl.dev] OCL 7.0 plans |
HiImprovements to the Pivot OCL have been getting increasingly hard within even a slightly enlightened view of API compatibility. It is therefore time for a major version change so that the many waiting- for-major-version clean-ups can be resolved.
OCL has not had a self-inflicted major version change since OCL 1.0, 17 years ago, and suffers from a diversity of major versions caused by other projects and the start-at-one policy for new plugins. This is pretty crazy. (Over the last few years I have jumped some minor versions so they are now pretty uniform.)
All Pivot plugins and features will therefore move to 7.0 in the next release, quite possibly a year or two away. The Classic OCL plugins and features may well continue as is. 2.x for Ecore, 3.x for UML.
The new version should remain source-text compatible, but will have significant API simplifications and hopefully a decent Environment Facade to avoid the need to tunnel through the numerous managers. For known third party use cases such as Papyrus some effort to retain compatibility via deprecated methods will be made. If you have clear Pivot OCL use cases please identify them on https://bugs.eclipse.org/bugs/show_bug.cgi?id=581716.
https://bugs.eclipse.org/bugs/show_bug.cgi?id=509309 is a planning bug that links to the many issues that deserve attention.
Regards Ed Willink