Open danthegoodman1 opened 2 weeks ago
Then these can be used as additional filters in graph traversal (will likely kill performance)
Can mark properties as indexed or not (whether we keep in mem too vs. just on disk).
Might also want a memory-only mode where the whole database is kept in memory in addition to writing to disk, like FDB's ssd
and memory
storage engines
Be able to assign arbitrary properties to node (KV), and allow creation of indexes from them (BTreeMap).