Few questions about Puppet, Part II

Hello folks.

Few days ago, you guys helped me out with some questions regarding Puppet. I came up with few more questions regarding how you implement it.

• Is puppet used to install from source? It appears to me that sure you can do it with Puppet but it doesn't look like something that Puppet was designed for.

• If you need to have a specific version of a service/application which is not provided by the Repos of your distro, then how do you deal with that predicament? You create your own RPMs?

• When installing via Puppet a database engine, do you also assign Puppet to create the actual database, schemas and upload of data or no?

Thank you :)