Warning:
This wiki has been archived and is now read-only.
Variant management
Approach USDL Variant Management
General Info
USDL needs extensions and alterations along several dimensions. A user or community might need additional or alternative attributes, classes, relations or even whole modules. Users might call the same entity differently. Further, the information might deviate depending on “context categories,” e.g., geopolitical, industry, or business process.
We conjecture that this will lead to a proliferation of USDL variants for specific countries, industries, etc. much like what has happened in the product world. We need a mechanism to sync or otherwise interrelate these variants with the main (comprehensive/core) USDL.
There exists a disruptive solution consisting of four main pillars
- Canonical Grammar : UN/CEFACT Core Component Technical Specification (CCTS)
- Context Logic: UN/CEFACT Unified Context Methodology Technical Specification
- Tooling: Knowledge Integration Library formerly called Warp10
- Processes: Tool-supported processes for governance
It is our idea to use this solution also for USDL
Useful Links
- The approach was initially presented during 2011-03-09_Approach_to_USDL_variant_management
- The slides can be found here File:2011-03-W3C-USDL-Variant-Mgmt.pdf
- A recording of the session can be found here [1]
- Further information about an older prototype by Gunther Stuhec, Standards Architect, SAP:
Discussion
- First feedback by W3C members was collected during 2011-03-09_Approach_to_USDL_variant_management
- Further threads occurred on the mailinglist
- Main concerns are
- we do not want to lose the software engineering support we currently have from EMF/Ecore
- governance process unclear (how much governance will be involved, how hindering will it be?)
- independent organization to host the solution
- license costs