Week 19

We’re halfway through the course, and the last four weeks have helped me connect the dots. I learned how to organize messy data into cleaner tables so we don’t repeat information, and how to write queries that combine tables while still showing items that don’t have a match. I also got the idea of indexes—they’re like a book’s table of contents that make lookups faster—and I saw how data is stored behind the scenes in simple blocks/rows with a map of which spots are free. Finally, I built a small ordered index and practiced keeping it updated when rows are added or removed, then used it to speed up lookups instead of scanning everything.

I still have a few questions. When is it worth creating an index, and when is it overkill? How do I pick the best column (or columns) to index for real-world queries? How does the database decide whether to use an index or just scan the table? And what are some easy rules of thumb for designing tables that will work well for reports we haven’t thought of yet?

Another big takeaway for me is the value of slowing down and planning before I touch the keyboard. Sketching a quick diagram or writing the query in plain English first has saved me from a lot of trial-and-error. When I did rush, I usually ended up with confusing results or repeated data, so I’m trying to be more intentional: define the goal, check the columns I really need, then write the query.

Looking ahead to the rest of the course, I want to get faster at reading my own results and spotting mistakes early. I’ll keep practicing small, focused exercises—like turning a messy table into a clean set of tables, or timing a simple query with and without an index—to build confidence. My goal is to leave the class with a solid routine: plan the data model, write clear queries, and use indexes only when they actually help.