One Bright Star: An Introduction to the Star Schema
why an astronomer charts the night sky around one bright, central star rather than scattering every point of light with equal weight, and how a star schema organizes data the same deliberate way.
Part 1
Part 2
What the Star Actually Measures: Fact Tables
why the bright star at the center of an astronomer's chart is defined by what can actually be measured about it, and what genuinely belongs — and doesn't belong — inside a fact table.
Part 3
The Planets That Describe It: Dimension Tables
why the points of light surrounding an astronomer's central star each carry a specific kind of descriptive detail, and what makes a dimension table genuinely useful rather than just a pile of attributes.
Part 4
When a Planet Grows Its Own Moons: The Snowflake Schema
why an astronomer sometimes needs to break one point of light into a smaller system of its own, and how a snowflake schema normalizes dimension tables into layered, related detail.
Part 5
Star or Snowflake: Choosing Your Sky
why an observatory director has to choose deliberately between a simple, flat chart and a deeper, layered one, and how to make the star-versus-snowflake decision on genuine evidence rather than habit.
Part 6
One Point of Light: Grain and Why It Decides Everything
why an astronomer must decide, before logging a single observation, exactly what one entry in the log actually represents, and why a fact table's grain is the single most consequential decision in its design.
Part 7
A Calendar Carved Into Every Chart: The Date Dimension
why every single chart in the observatory's archive references the exact same calendar, carved once and reused everywhere, and why the date dimension is dimensional modeling's most universal building block.
Part 8
The Same Star From Two Observatories: Conformed Dimensions
why two separate observatories charting the same star need to agree on exactly what counts as 'the same star' before their charts can ever be compared, and what a conformed dimension actually guarantees.
Part 9
When the Chart Goes Out of Date: Slowly Changing Dimensions, Type 1
why an astronomer sometimes simply corrects an outdated entry on the chart and moves on, overwriting the old information entirely, and what that choice deliberately gives up.
Part 10
Keeping the Old Chart on File: Slowly Changing Dimensions, Type 2
why an astronomer sometimes files a brand-new chart entry instead of correcting the old one, deliberately preserving exactly what was true at the time, and how Type 2 slowly changing dimensions track real history.
Part 11
Sweeping Up the Loose Stars: Junk Dimensions
why an astronomer gathers a handful of small, unrelated markers into one shared reference card instead of scattering them individually across the chart, and what a junk dimension actually solves.
Part 12
A Star With Nothing to Measure: Factless Fact Tables
why an astronomer sometimes logs an entry with no reading attached at all, simply recording that an observation happened, and why a factless fact table is still a genuinely useful kind of fact table.
Part 13
When Two Stars Share a Planet: Bridge Tables
why an astronomer sometimes finds a single point of light that genuinely belongs to more than one star system at once, and how a bridge table handles this many-to-many relationship correctly.
Part 14
Not Every Measurement Adds Up: Additive, Semi-Additive, and Non-Additive Facts
why an astronomer can sum up total observation time across every night of the month but can never sensibly sum up a star's recorded temperature, and why not every measure in a fact table behaves the same way.
Part 15
Charting a Whole Constellation: Fact Constellations and Galaxy Schemas
why a full observatory archive eventually needs more than one central star to make sense of everything being tracked, and how a galaxy schema connects multiple fact tables through shared dimensions.
Part 16
The Number That Doesn't Need Its Own Planet: Degenerate Dimensions
why an observation's session ID sits directly on the chart's central star rather than earning its own descriptive point of light, and what a degenerate dimension actually is.
Part 17
Pre-Drawing the Busy Skies: Aggregate Tables and OLAP Cubes
why a busy observatory pre-draws the most commonly requested summary charts in advance rather than recalculating them from raw observations every single time, and how aggregate tables and OLAP cubes speed up real queries.
Part 18
Teaching the Telescope to Draw Its Own Chart: AI-Assisted Schema Design
why an observatory that once relied entirely on senior astronomers to hand-draw every new chart now has a telescope that can propose a first draft on its own, and what genuinely changes when schema design gets AI assistance.
Part 19
When the Telescope Points Itself: AI Agents Querying the Star Schema
why a telescope that can decide on its own which star to observe next needs a genuinely well-organized chart to work from, and what changes when AI agents query a dimensional model directly.
Part 20
One Sky, Every Kind of Question
the star and its planets, the moons and the constellations, the grain and the calendar, every article's lesson reassembled into one coherent sky capable of answering every kind of question asked of it.