This Tuesday I had the privilege to attend the OMG Systems Engineering DSIG technical meeting at which they spoke a lot about OSLC. While I believe there are some official notes still to come, I thought I’d share a little bit about what happened and some of my observations.
There was good attendance by vendor and end user organizations, by folks who’ve been active in the OSLC community and those who’ve just been watching. Some of the organizations represented include: Atego (co-founders of the OSLC UG for Embedded Systems), Boeing (by OSLC Steering Committee member Mark Schulte), Continental Automotive, Engisis, IBM (by me), Koneksys, Lockheed Martin, Mentor Graphics, MITRE, NASA JPL, NIST, No Magic, and Raytheon. Many of these organizations had some experience with using and/or implementing OSLC, and everyone was keen to learn more. As I highlighted before the meeting, there were three main presentations about OSLC, and I think that Parham, Axel, and Mike all did great jobs educating the DSIG about OSLC, sharing some of their perspectives and experiences, and generating good discussion.
A couple interesting items from the discussion included:
- It is very typical for end users organizations to also be implementers as they work to OSLC-enable their homegrown lifecycle tools.
- They are also having good experiences implementing OSLC. There were several comments along the lines of “we’re finding it easy … after overcoming the initial learning curve”.
- As end user organizations get further along with their OSLC enablement, they are all making an effort to engage their vendors who don’t yet support OSLC.
- There has been some success getting vendors to implement OSLC (it sounds like our Software page may be missing quite a number of vendor implementations).
The OSLC4MBSE discussion portion of the meeting was quite lively. It sounds like the first step for this group will be to develop RDF representations for all the elements of the SysML standard. Beyond that, there are many options, but I suspect there will be some requests from this group to extend certain OSLC resources to cover concepts from SysML, while other concepts will be standardized at OMG as an extension of OSLC for SysML. We’ll have to see how this develops, but it’s not like that will take much more effort as many participants in the OSLC4MBSE working group are also active participants in the “core” OSLC community.
There was quite a bit of side discussion about how to manage a product configuration across all the disparate systems you get with OSLC-based integration where everything is live data and not a copy. I think some of the work of the Configuration Management workgroup is relevant here, but I also hope that some of the folks raising those concerns will contribute some scenarios to detail their concerns. This seems to be the case of an old problem manifesting itself in a new way: everyone knows how difficult it was to deal with before and want to confirm that there is a way to handle it in the “brave new OSLC world”.
Without qualification my first trip to an OMG meeting was interesting and a success. I will periodically check in on the OSLC4MBSE working group to see how they’re progressing. I also hope to collaborate with many of the people I met on Tuesday in the days ahead.