Hacker News new | ask | show | jobs
by math 2231 days ago
At the moment you more or less need to write a DBMS in your app code, but I don't think that's the end state. I think what we're seeing the beginnings of something big - it just might not seem like it yet because it's the v1 / no where near complete version. I think having all your data in a single system (Kafka, KsqlDb, ..) that allows you to work with it in cross paradigm ways will turn out to be very compelling.
2 comments

> At the moment you more or less need to write a DBMS in your app code

Since we're discussing misunderstandings in the community, it should be pointed out that a Database Management System (DBMS) is not merely a database, to say nothing of a data store. Oracle, Postgres, et al are genuine "DBMS". ATM you are very likely putting together a data store in your app code.

You're conflating DBMS with RDBMS.

I interpreted the OP as DBMS=database, which absolutely includes application code that stores and retrieves data in proprietary formats.

Even a linked list mmap'd to disk can be a database, just maybe not a very good one.

A RDBMS is simply a Relational Database Management System. A quick tour of CS history reveals ancient curiosities such as Hierarchical Database Management Systems. And there is more:

https://www.studytonight.com/dbms/database-model.php

There are many kinds of databases. Graph, key/value, relational, hierarchal etc. That doesn't change the fact that any app that writes code to store and retrieve data is creating their own database or using someone else's.
There are many kinds of data models. A DBMS is a DBMS, and typically a specific DBMS supports a specific data model. A log structured file is at most a data store.

To be fair this is a somewhat fuzzy categorization (DBMS, DB, Data Store) and it can cause confusion. An DBMS is a system, just like it says right on the acronym tin. A DBMS is a system that provides capabilities such DDL, DML, auxiliary processes, etc. to manage a database. There are various data models, e.g. a triple store, for databases, but the data model X is a orthogonal to XDBMS.

Meteor did some pretty interesting work with this by tailing mongodb’s oplog. Definitely still room for innovation.