Interesting to see that after having (no longer so basic) support for JSON documents we will right now have graphs in SQL databases as well.
Although there probably are less use cases for graphs that there are for non or semi-structured data in stored in JSON documents, use cases do exist and they often are annoying to model using just tables, joins and standard relationships (SQL-wise).
On the other side, it will make SQL databases (their implementations) even more complex than they already are.
I would be also interested to see how, in the article there's Postgres example, it was implemented under the hood, so it has acceptable performance; I guess similarly to joins?
Although there probably are less use cases for graphs that there are for non or semi-structured data in stored in JSON documents, use cases do exist and they often are annoying to model using just tables, joins and standard relationships (SQL-wise).
On the other side, it will make SQL databases (their implementations) even more complex than they already are.
I would be also interested to see how, in the article there's Postgres example, it was implemented under the hood, so it has acceptable performance; I guess similarly to joins?