Skip to main content
Back to blog
•2 min read

Databases Are Weirder Than I Expected

techdatabaseslearning

i started a databases course. turns out storing data is an entire field and it's been around since before i was born.

who knew? (everyone in the industry, apparently.)

sql: first impressions

it's english-ish. like, you can kind of read it and understand what's happening:

SELECT * FROM students WHERE grade > 80

"give me all students where grade is more than 80." makes sense.

but then it gets weird:

SELECT s.name, c.title, e.grade
FROM students s
JOIN enrollments e ON s.id = e.student_id
JOIN courses c ON e.course_id = c.id
WHERE c.department = 'CS'
GROUP BY s.name, c.title
HAVING AVG(e.grade) > 75
ORDER BY s.name

my eyes crossed. multiple times.

joins: the concept

when i first learned about joins i genuinely didn't understand why data would be split across multiple tables. just put it all in one big table, right?

wrong. so wrong.

the short version:

  • redundant data is bad (takes space, causes inconsistencies)
  • splitting data into related tables is good (normalization)
  • joins bring it back together when you need it
  • there are like five types of joins and they all do different things

i drew diagrams. so many diagrams.

the beautiful parts

once it clicks, SQL is actually pretty elegant:

  • aggregations: count things, sum things, average things, in one line
  • subqueries: queries inside queries (query-ception)
  • views: save a query as if it's a table
  • transactions: do multiple things atomically (either all succeed or all fail)

there's a reason this tech has been around for 50+ years.

the frustrating parts

  • debugging is hard (where did the data go? why is it null? why are there duplicates?)
  • performance matters and i don't understand performance yet
  • every database has slightly different syntax
  • accidentally deleting data is terrifyingly easy

i've run DROP TABLE on a test database and still felt a moment of panic.

nosql exists

apparently there's a whole other world of databases that don't use SQL. mongodb, redis, cassandra... they solve different problems.

adding to the "things to learn eventually" pile.

practical realization

every app i want to build needs to store data somewhere. understanding databases isn't optional—it's fundamental.

the course is hard but i get why it's required.

wrote my first stored procedure today. it works. i don't fully understand why. classic.