3 ms·
Two ways I can think of off the top of my head: 1. Total reliance on ORMs, and never actually touching SQL (or not in 5 years since you started using ORMs) 2.
by bobwaycott 10y ago
Two ways I can think of off the top of my head:
1. Total reliance on ORMs, and never actually touching SQL (or not in 5 years since you started using ORMs)
2. Not building complex relational schemas with related items that you use SQL to pull in results of combined data that follows those relationships.
I'd wager #1 is the more common case.
- dagw 10y agoOr they're just using a bunch of SELECT * WHERE X queries and then filtering the results in their favorite programming language
- bobwaycott 10y agoThat, too.
- Thriptic 10y agoThis. With smaller amounts of data it's not even a big deal. Just pull everything into python or R and subset in there.
- golergka 10y agoBut in #1 you're not working with SQL, you're working with a layer above it. And wouldn't even the most basic data, properly normalized, require joins? I mean - learning about database normalization is one of the first things you're supposed to do when you work with relational databases, right?