Andy Berner wrote on 02/26/2010 09:15:34 AM: > It will always be the case that an organization's tool infrastructure > requires one tool to "know about" multiple providers of the same OSLC > resource types. Do we have any thinking in general of how that should be > handled?
This is the norm today. A consumer can register multiple providers and manage as it pleases, for example a quality management tool. The user experience at runtime of when the tester selects one of providers is tool-dependent. A tester could be prompted as to which one it wants to use or a team lead could associate a different provider with different contexts in the test project: based on test suite, components, etc. We don't mandate any particular experience. If a QM tool wanted to, it could limited it to one provider. The links to these different resources are absolute URLs that resolve to the appropriate provider. Are there particular issues that you see with consuming multiple providers? Most specs are built with the assumption that a consumer would use at least one provider but not limited to one. Thanks, Steve Speicher | IBM Rational Software | (919) 254-0645
