I think we should provide some more details about what the custom server assemblies are. If we already provide those details, then point to that section in the documentation.
We should establish a more solid starting point for this tutorial or the rest of the sections/step won't be fully understood.
We also need to improve the Geronimo Plugin Architecture documentation so instructions like "...To assemble a server from existing instance of geronimo we need to include geronimo-boilerplate-minimal plugin which includes several files used to start the server..." have more meaning.
It is not clear what we should achieve after completing section "Identifying the functional components required by our server".
Looking at step 6 under "Assembling a custom server" I see "default/StatelessSessionEJB/1.0/car" and "default/ApplicationClient/1.0/car" but it is not clear how/when these should be selected. I think we should indicate first the components for the application we want to create a custom server assembly for. Then indicate what are the minimum server components needed to get a functional custom server assembly. Last indicate the optional componenets like Administration Console, etc. It is not too clear to me how, if I include the Administration Console plugin the DB wizard won't be included and still have to explicitly add it. (Geronimo Plugin Architecture documentation should probably address this)
Could a plugin (or custom server assembly) contain information to create a DB and populate some data if such DB is on the embedded Derby?