|
|
|
|
|
|
by bayesian_horse
1317 days ago
|
|
|
Depends on the use case. Recursive SQL can be good enough, maybe even faster, for certain use cases. The problem isn't so much the query language but the indexing. Graph engines index the nodes and edges in a particular way so that traversal is fast. Most examples of Recursive SQL I've seen will only involve nodes on exactly one Table an with exactly one kind of a relationship/edge (for example a tree with "parent" edges). Graph DBs allow you to relate multiple different types of nodes using multiple kinds of edges. The edges can have queryable attributes like an intermediary table in a many-to-many relationship. And all of that is still indexed efficiently. |
|
|