Re: [boost] [review] [sqlite]

I've finally got to understand multi_index.cpp and adapt it to my data structure. I understand the use case that this example is trying to cover is exposing an in-memory data structure as a sqlite interface. Say I've got a server process holding such a multi-index. I've got a working module for the multi-index. I can load it and call it from the server process. I'd now like to somehow use the module from outside the process (e.g. from the sqlite3 command line tool) to inspect my data. Is such a use possible? What would be the recommended architecture? Or is it best to keep such modules scoped to the host process? Thanks, Ruben.