Obsolete "system"

[ddooley]

Discussed on May 3rd call. History: past BFO work foresaw a "system" class under which systems could be modelled in more domain specific ontologies. RO thus introduced a "system" class and "connects" and "connected to", and "determined by" and "determined by part of" object properties that reference system in domain or range.

However, the "system" class and its connection/causal relations doesn't seem to provide any descriptive power beyond what RO can do with other mereotopology relations? Shall we obsolete it?

The idea would be to invite some group to provide a more comprehensive model of system for those who need it, and to have this outside RO if/when it arises.

Are the connects / determined by relations being used or can they be deprecated in favour of some other RO replacement relation? Or do we keep them but simply drop their system range and domain constraints?

[ddooley]

Note, there are two system terms from CARO and ENVO ontologies currently under RO system:

Perhaps COB wants "system" and then CARO and ENVO can respectively import that, and place their own system classes under it?

[ddooley]

So we do need a decision from COB about taking on "system". I've created an issue there: OBOFoundry/COB#206

After RO obsoletes "system", then ENVO will need a new parent term for "environmental system". RO uses environmental system in a few other object relations.

[ddooley]

At moment RO "other_imports.ofn" has:

Class: obo:ENVO_01000254 (environmental system)

AnnotationAssertion(rdfs:label obo:ENVO_01000254 "environmental system"^^xsd:string)
SubClassOf(obo:ENVO_01000254 obo:RO_0002577)

Do I manually edit this once we have a new parent for "environmental system"