Persistence Patterns / Heterogeneous Data Sources RRS feed

  • General discussion

  • I'm struggling with how to organize an application with a complex domain model and multiple data sources.

    We have a SQL Database and a database consisting of Paradox tables.

    I hear the repository pattern shines in this scenario.

    What is the best way to separate a repository from the actual persistence code?

    I would like to keep general descriptions of the repositories in one library and the persistence code in a separate library. I hope to eventually replace the Paradox tables completely and I do not want code I intend to keep to even rely on the structure of the Paradox tables.

    Thanks in advance.

    Wednesday, November 11, 2009 4:11 PM