[kaliqt]

So it's a bit of tradeoffs, we may add a second DB to the compose by the time we reach v1.0, but, if we had to pick the convergence of simplicity and flexibility to start, PostgreSQL is it.

We prototyped SQLite as well. It just wasn't working in the stack like we had needed.

The idea is simple: use as few components as possible to achieve the outcome. Better to have less than more initially, because we can simply add the more and extend the API, but ripping everything out and trying to trim it down is damn near impossible, especially when the platform garners widespread usage.

The decision to do the less then more instead of more then less approach was spurred by us (the wider project) always having way too much bulk and then finding it impossible to turn the ship when we needed to (https://github.com/vircadia/vircadia-native-core).

We need to be agile to reach the milestones we have set out, hence the differed approach using off the shelf parts and only adding what we need, when we need it.