Dear,
We have a Value Chain with its development in macroprocesos and processes. The processes modeling occurs distributed by various areas of enterprises, these areas have the autonomy to make the details of their processsos.
We have the aim, as a central objective of this work, to promote integration among these developments. For example, an object created by a process of a manager process X must be the same object that will be reused in the development of another case of a manager process Y.
In our architecture the use of Aris have 2 spaces: one for production and one for development.
We found that the following problem may occur: The process manager is modifying the process X while the process manager is Y modifying the process which makes use of object J that was generated by the X. Since the process X withdrew the object of base J, we started having consistent problems.
Someone already have this problem? How did you solve?
Best Regard,
Francisco
Hi Francisco,
We have a similar problem here with re-use of "non-validated" content. We've instigated security controls so modelers only have access to particular areas where they are working and a library of validated content. The three parts we have in place are:
- Segregated work-areas, so the modelers can't re-use "non-validated" content from other work-areas.
- Release Cycle Management to validate content for re-use and promote to a Validated Content Library.
- A Validated Content Library (that is accessible by all modelers) to enable the re-use of validated content in each of the separate work-areas.
However it's a lot easier to say than do, and we still have issues when modelers are working across multiple work-areas concurrently. Hopefully that helps and gives you a starting point.
Good luck!
Regards,
Jonathan
In reply to Hi Stian, This is possible by by Freuta
Hi,
Thanks, we are experiencing the same kinds of problems and are considering a version of what you suggest here. Our content used to be stored in a single repository but now we're looking at having multiple(dev and prod).
My thinking that is that when a work in process project requires a change to validated content, then it becomes part of the project. Then when a second project also requires changes to that same content they also get a version in dev. Whoever "commits" first, similair to source code versioning, does so easily, and the second project has to verify with the first.
Does this make sense to you? A problem I'm thinking about is that sometime there may be a contradicting change or one that requires both projects to talk in the deisgn phase so that there isnt too much rework, but its difficult to pick up on those details during the modelling process.
