Concludes the lesson series with hordes, the level formed by gathering clans into sets, completing the layered structure of the Data Algebra.
Intro | Lesson 2 | Lesson 3 | Lesson 4 | Lesson 5 | Lesson 6 | Lesson 7
Transcript
Slide 1: Welcome to Lesson 8. We’ve climbed three rungs: couplet, relation, clan. Today we take the last step and complete the tower. The move is exactly the one you’ve seen three times now, so it will feel almost inevitable. We gather clans, whole tables, together into a set. That set is called a horde, and it is the algebraic name for something you use every single day without a second thought: a database. Once we have it, the tower is finished, and every kind of data you’ll ever meet has a home somewhere on those four rungs.
Slide 2: Let me say the move plainly, because its very repetition is the point. A set of couplets is a relation. A set of relations is a clan, a table. And a set of clans is a horde. That’s it, the same operation, gathering things into a set, performed a fourth and final time. One clan is a single table. Many clans, collected together, are a database, or a schema, or a file system. The tower doesn’t grow more elaborate as it rises; it grows by the same simple rule at every level. And with the horde, we reach the top.
Slide 3: Here is a horde. Three clans, three tables: a Customers table, an Orders table, a Products table, each one a set of relations, each relation a set of couplets. Gather those three clans into a single set and you have a horde. And what is a collection of related tables, sitting together, meant to be used as a whole? It’s a database. The thing your organisation runs its entire operation on, the database, is, algebraically, just a set of clans. No new concept was needed to reach it, only the same gathering move, one more time.
Slide 4: And as at every level, once you have the abstract idea, you start seeing it everywhere. A database is a set of tables, a horde. A file system is a set of files and folders, a horde. A database schema is a set of related tables, a horde. These are not three different kinds of thing needing three different theories. They are one structure, the set of clans, appearing in different corners of computing under different names. The algebra sees through the names to the single shape underneath, which is exactly what lets one engine serve them all.
Slide 5: Two quick but important notes. First, time. You might expect temporal data, history, versions, time-series, to need special machinery. It doesn’t. Give each relation a timestamp as one of its yins, and a point in history becomes just another value in the data. Comparing across time is then an ordinary operation. Second, partitions. Split a clan into subsets and gather them back up, and you’ve deliberately built a horde, useful for organising large data. And remember the tower runs both ways: take the union of a horde’s clans and you descend again. Up and down, as you please.
Slide 6: And there it is, the tower, complete. Couplet at the foot, an element of G cross G. Relation above it, a set of couplets, P of G cross G. Clan above that, a set of relations, P squared. And now the horde at the summit, a set of clans, P cubed of G cross G. Four levels, built by one rule applied four times: each level is simply a set of the things one level below. That single, humble pattern, gathering into sets, is enough to represent every data structure in computing. That was the claim we set out to prove, and here it stands.
Slide 7: So, to summarize. A horde is a set of clans, and the database you rely on is exactly that, a set of tables. Time needs no special apparatus, a timestamp is just a yin, so history is only more data. And the whole tower now stands on one rule: couplet, relation, clan, horde, each a set of the last. That completes the mathematics. From here, the series turns from what the algebra is to why it matters, and what was wrong with what came before. Next time, we look honestly at the old model. In Lesson 8, good math and bad math.
Slide 8: That completes the tower, four levels, one simple rule, every kind of data accounted for. And with this we complete this short introduction to Data Algebra. In the next series of videos we will introduce our flagship database JabrDB.