I have reviewed some threads concerning how best to share models and repositories, but I tend to agree with some previous comments that the documentation provides many options and caveats without really providing a best approach for a given scenario.
Therefore, I would like to solicit comments on how best to set up EA for use within a small project team with relatively low concurrency (especially within packages - use cases would likely be the responsibility of one person, data models another, and class/sequence models yet another person).
Setting up multiple database repositories to support multiple models (please correct me if my reading of the docs in this is wrong) seems overkill, since we work on smaller projects that have little to do with each other and so should not be lumped into the same model/repository.
The XMI Import/Export seems to facilitate team members working in isolation and then merging their work in a manner amenable to version control systems, but this seems inconvenient for business analyst types who just want to open a project, do some work, and save it someplace.
The SCC/CVS version control integration would seem to suffice in a small project team environment with low concurrency in terms of work going on within a given package. It seems as though retrieving a project from source control, doing some work, and checking it back in is relatively painless from EA?
So, to avoid the trial and error as much as possible, any feedback on what would work most efficiently to support a small project team on small software development projects from end to end (requirements, analysis, planning, developing, stabilizing, deploying...) in terms of setting up EA with a repository and/or version control and/or shared/replicas/etc. would be most appreciated.
DR
