I dont understand why this page is necessary.
Previous versions of JB had all the datasources ready to be enabled - just modiify the xml. i.e. the worked case examples were there, almost ready to use. The XML could be provided by development teams to test teams, and only required the copying of the file to the allocated place. No existing scripts needed to be manually altered in the installation process. Migration between dev, test, production and DR was simplified and could be easily automated.
Now I sit, well into the 3rd millenium, with configuration that is as tricky and technical as mainframe JCL configurations in the late 1970s .
All my hard earned and paid for previous experience and scripting of JB goes out the window.
There is a semi functional JB admin consol. I should be able use it to download and install the drivers, connect to my datasources and check they are working. This should be easy to use. I could not care about the syntax of the JB internal stores and configuration files.
I should also be able to get from the admin console a XML fragment to give to programmers in order to configure the persistence unit connection.
I should not have to waste time fiddling and faddling through countless wikis and altering scripts.
JB should utilise maven to better manage version control during installation and production.