9 Comments

Awesome article Zach! I loved how you discussed the growing pains you experienced from initial decisions you made and how you eventually remedied them. You needing to fix them is an awesome sign off growth and all the success you've worked hard to achieve :)

Expand full comment
Feb 14Liked by Zach Wilson

Very interesting read! Thank you Zach! I am curious why you used ARRAY_CONTAINS and not a simple IN?

Expand full comment
Feb 13Liked by Zach Wilson

Why not use database triggers to also create the denormalized tables - instead of refreshing the materialized views . .? Is there a reasoning behind this choice ?

Expand full comment
Feb 11Liked by Zach Wilson

Great article. I love data modeling content.

Expand full comment

Very interesting overview. I didn't know about the CARDINALITY keyword!

I recently started a small project on the side, and had to design an ERD. I completely fell into the trap of too many tables, for 'just in case I'll need that feature'... So ended up with 11 tables, 4 of which I still never used, as I didn't get to building those features yet.

It makes much more sense build on top of something simple, and add columns/tables as you go.

Expand full comment

How report on the data? Worth doing a dimensional model with a star schema for analysis?

Expand full comment