Relational Databases
4 filmsComing soonOne question read end to end: what a database is beside a spreadsheet, what a table and its key are, what one SELECT makes the engine do, and where the rows physically sit.
Almost everything you do online ends up as a row in a database, and more and more often an AI writes the questions that read and change those rows. One idea at a time, from one question to the engine that answers it, in plain words.
One question read end to end: what a database is beside a spreadsheet, what a table and its key are, what one SELECT makes the engine do, and where the rows physically sit.
How the engine answers harder questions: putting two tables back together, turning millions of rows into a few totals, choosing a plan, using an index, and finding words inside text.
Designing the data: splitting one sheet into tables so that each fact lives once, rules the data cannot break, changing the shape of a database while it runs, and the databases built round documents and keys instead of tables.
What a database promises and how it keeps its promises: the four guarantees of a transaction, two people writing at once, a power cut a moment after 'saved', a deleted database and backups that did not work, copies that lag behind, and the choice forced when a link between buildings breaks.
What the age of AI asks of databases: a store built to read billions of rows for one total, a column that holds meaning and is searched by nearness, the query an AI writes that a person must read, and the database that holds an agent's progress.