Plan to effectively push COB forward

[cmungall]

from slack discussion

could you formulate a battle plan on how we could most effectively push COB forward? Maybe an inside out strategy where we COBbify 10-20 core OBO ontologies and then provide instructions to the community to follow suit? I am looking for something like this:

• Import COB (keep BFO import? Remove?).
• Check that none of your ontology classes fall outside of COB
• If they do, request a new (root) term and push push push until it gets in COB

As I see it there is work that needs to happen on 3 largely independent aspects

1. Technical-logical-release-management
2. Increasing coverage across OBO
3. Solving intractable upper ontology modeling issues

Help is appreciated on all these aspects

1. Technical-logical-release-management

For 1, let us keep our efforts focused on the end goal of enabling individual ontologies to completely use COB as the sole upper ontology, as currently documented here:

https://obofoundry.org/COB/using-cob/

This page highlights some of the challenges, some of which are described in GitHub issues here, e.g. #213 (we could do with some organization of our issues, so that it's easy to see all technical blockers)

2. Increasing coverage across OBO

For 2, @cthoyt suggests doing this in concert with http://purl.obolibrary.org/obo/IAO_0000700 which is a great idea (I think we should actually make this a dashboard requirement - and this makes the ontology much more friendly in OLS even without COB). Note that 2 may depend somewhat on 3:

3. Solving intractable upper ontology modeling issues

We all know which ones these are. Let's keep discussing these - but keep them on the relevant issue, which should be easy to find (again, some organization would help here).

If we want to discuss a general strategy for addressing these issues - e.g. Schulzian conflation - then let's make a separate issue for these.

Let's use this issue here for discussing high level strategy, and formal criteria for what success or a series of small wins looks like.

[StroemPhi]

What I'm currently missing in COB are textual definitions. This violates OBO principle 6.