On the path for a better select
The current indexing implementation in my tyrdbs project is based on a hybrid of prefix and b-trees. It's messy, has significant space overhead and lacks some features I'd like to see so I went looking for a suitable replacement. Last time around, I ...
