#547 Notes that releases should import ontology version IRIs rather than the base ontology IRIs. The problems mentioned in that issue wouldn't exist if all users relied on local copies and the catalog file. But, the oddities of Protege and some user's preference to grab ontologies via the web can lead to problems with resolution of imports. While importing via version IRIs sounds like a good idea, it raises a question about whether the official release of CCO should be just a single merged file vs. 11 ontology files as standalone release products. If the former, then the individual ontology files really just exist for development purses, meant for advanced users, and thus should use base IRIs in import statements. If the later, i.e., each ontology as bonafide release, then we will need to implement better import control.
@alanruttenberg @johnbeve @neilotte @APCox
#547 Notes that releases should import ontology version IRIs rather than the base ontology IRIs. The problems mentioned in that issue wouldn't exist if all users relied on local copies and the catalog file. But, the oddities of Protege and some user's preference to grab ontologies via the web can lead to problems with resolution of imports. While importing via version IRIs sounds like a good idea, it raises a question about whether the official release of CCO should be just a single merged file vs. 11 ontology files as standalone release products. If the former, then the individual ontology files really just exist for development purses, meant for advanced users, and thus should use base IRIs in import statements. If the later, i.e., each ontology as bonafide release, then we will need to implement better import control.
@alanruttenberg @johnbeve @neilotte @APCox