Broadcast date: April 27th, 2022, 19:00 CET
SysML v2 is in its final run. In this episode, we address the question of what will then become of the many SysML v1 models.
The SysML v2 specification also includes a formal definition of the transformation from SysML v1 to SysML v2 models. Since Tim is one of the transformation developers, we didn’t invite an additional guest this time.
We’ll talk about how the transformation works and what the challenges are. We will also look at different use cases of how a transition from SysML v1 to SysML v2 can look like.
You can join us live at the Youtube Livestream on April 27th, 2022, at 19:00 CET or later on Youtube, Spotify, iTunes, Amazon Music, or Google Podcasts.
Podcast: Play in new window | Download
Thank you for this interesting conversation and for sharing your insights.
I have some very basic questions from a layman’s perspective:
1. I am about to start my journey in MBSE (which first involves learning the Language properly) and serious modelling. Since it would take another 1 or 2 years for sysml v2 to be published and for subsequent tool support – how can I ensure minimum rework when v2 is official. I understand that transformation tools would be available even for large models , but what I also gather is that a 100% accurate transformation from v1 to v2 cannot be guaranteed at least in this point of time.
2. In connection to point 1) is there a guideline that I can follow that ensures that I do not feed “garbage in” to my sysml v1 models now and develop them in such a way that allows a transformation to v2 as smoothly as possible.
3. can user-defined stereotypes in v1 be transformed to v2?
Thanks in advance.
We aim for a 100% accurate transformation. Probably we can achieve it on paper. Be aware that the official specification of the transformation is not a running implementation but a PDF. So, it finally depends on the tool vendors.
Another challenge is the SysML v1 models. Depending on the tool, they often do not conform to the SysML v1 standard. The transformation only considers conforming models.
I have two recommendations for your SysML v1 models to be well prepared for a smooth transition:
We plan to transform user-defined stereotypes to v2. v2 also has a language extension mechanism. I will start working on that part of the transformation soon and hope it will work without any issues. In any case, we need a solution for that because every SysML v1 model uses user-defined stereotypes (or at least should use them).