We had a long and fruitful discussion but at the end decided we do not have enough information to determine exactly what dependencies or integration points exist within the Hydra architecture. We looked over several documents on the wiki which are good but do not directly address these concerns. Our one take away is that the current attempt to register what versions applications are using is not working and that accomplishing that task could help inform this effort. To that end, we propose replacing the current "manual" software registry with an automated method to register information about each Hydra application effort.
What we want is to be able to document the internal and external interfaces which would allow us:
- more flexibility in being able to leverage each component
- expand choice in alternatives
- ease of integration when an application requires additional technologies in concert with the Hydra components
- ease of maintenance - reduce maintenance costs
Information about applications would allow us to know -
- who is on the treadmill?
- who is taking a rest?
- what are the dependencies for that application?
- what additional technologies are integrated with the core Hydra gems?
- what versions is that application using?
- what institutions are involved?
- discover similarities and opportunities for alignment.