Just FYI, there is an ongoing discussion about modularization here:
https://github.com/EnvironmentOntology/gaz/issues/21
I just wanted to discuss a few things at a higher level to make sure we're on the same page and to check some basic assumptions.
First, this is a blocker on most other issues since we want to be able to manage GAZ in version control, and GitHub imposes file size limits. We *could* go back to some alternative, either a VCS that allows larger files, or something like DropBox with a lock mechanism (shudder...).
Second, we are going for an instance based representation. How many people have tried editing instance graphs in Protege? It's pretty clunky. But if people are fine doing that, that's OK, I thought it might be worth people experimenting before we go too far down this path.
Part of this depends on what kinds of workflows we see being required. Mass entry of hundreds of new places in Protege would be ineffecient. We can use templates here, but it would be helpful for us to get on the same page about how template editing fits into the overall workflow and pipelines.
Also, I expect many places will already be in Wikidata. It would be easy to have something that mireots a chunk of WD or geonames into GAZ.
In fact one possible route would be to use Wikidata *as the edit interface for GAZ*. Simply enter it in Wikidata, it's available for SPARQLing almost immediately, we can have a script that runs via Jenkins and pulls new entries down. How does that sound to people?