Future of Executable UML


[May 19, 2015:  The following is an excerpt from a post to Executable UML on LinkedIn.  I recommend reading the entire thread.]

Hello Lee, Charles, Ed, George and All,

Coming from the heart of xtUML/BridgePoint-Land, I would like to express my enthusiasm regarding Papyrus-RT and the Future of Executable UML.  The teams at xtuml.org and One Fact have been following the project… albeit mostly through conference presentations and leaked information found on social media.  🙂

The One Fact team plans to participate in the project directly and indirectly.  The xtuml.org community will support with source code, models, tests and standards/proposal reviews.

Technically, BridgePoint became fully open source last October (’14).  Practically, BridgePoint is accessible and buildable/modifiable only in the current 5.0 release (still in beta April ’15).  The tool is built with xtUML models, so the Eclipse/Java model compiler had to be documented and made easier to use.  The team has had its “head down” accomplishing this, but is now coming up for air and rejoining society and participating in the Future of Executable UML.

I envision an incremental, “never broken” strategy that does the following:

  • recognizes the existing body of xtUML models (as valuable IP and valuable for testing future Executable UML syntax/semantics)
  • similarly acknowledges existing model compilers and their usefulness going forward
  • steps xtUML (the language and toolery) forward toward new capability without a gap (“gap” means a tool that neither executes nor translates models)
  • retools onto open standards -based technologies
  • leans on Papyrus heavily always pressing toward sensible convergence
  • keeps the focus on modeling, even modeling the tool itself