Browse files

Add "why_everything_is_relational"

  • Loading branch information...
alexpdp7 committed Jul 27, 2018
1 parent c5a1b6d commit 1e819ecfc0f638f79c5196d0a51c954be86c7a39
Showing with 15 additions and 0 deletions.
  1. +1 −0
  2. +14 −0
@@ -19,6 +19,7 @@ We believe that writing such software should be a solved problem, but it is too
This is an index of the topics we believe are relevant to the discussion. Feel free to send us pull requests to expand the index or fill in topics.

* Data first systems design
* [Why everything here seems to talk about relational databases?](
* Database design
* Some information is not relational
* Principles
@@ -0,0 +1,14 @@
# Why everything here seems to talk about relational databases?

We don't think that a relational database is the best home for all data out there.

But we think that in most scenarios, it should be your default option, especially if:

* Incorrect data is a significant problem. Relational databases excel at constraining data and keeping invalid data out of the database
* Ad-hoc reporting is needed. Normalized relational databases are very easy to query and to extract data from in most scenarios

Relational databases also benefit from being one of the most mature data storage technologies, so we have reliable, featureful implementations with huge ecosystems behind them.

In short, we think that unless you have a well-argumented reason for choosing another data store, your best initial option is a relational database.

We will discuss reasons to look into other data stores in [TODO].

0 comments on commit 1e819ec

Please sign in to comment.