Featured article
Declare the Grain Before Writing the Query
Define what one row represents, distinguish entities from revisions and attach that meaning to keys and measures.
SQL, models & data integrity
Original guides and technical studies, written to make the assumptions and trade-offs visible.
Featured article
Define what one row represents, distinguish entities from revisions and attach that meaning to keys and measures.
Give each data rule a scope, a null policy and an action, then keep rejected records visible in the accounting.
Inspect join cardinality, aggregate at the intended grain and preserve missing matches when they matter.
Align population, units and reporting boundaries before comparing totals, and keep missing keys distinct from value differences.
Separate delivery identity from business revision and make repeated input converge on a defined accepted state.
Distinguish instants from local dates and define numeric units and rounding boundaries before aggregation.
From the wider community
Publications from DEV Community. Titles and summaries are attributed to their original authors; the links open the original articles.
Introduction This article takes through the ideas that decide whether a model works as...
The FETCH FIRST ... ROWS ONLY clause arrived in the SQL standard with SQL:2008, and Oracle Database...
“My database is only 50 MB. Why is my API slow?” It's an easy assumption to make. Small database =...
A complete data platform — batch + streaming, single-node + distributed, local + Kubernetes. What I built, what broke, and what I learned.
By default an Oracle client and server negotiate no encryption, so every query and every row crosses the network in readable bytes. A packet-capture lab shows the SQL and a confidential value in plaintext, then four lines of sqlnet.ora (REQ
I have been using OpenSearch for querying and aggregating data at work. Initially, I was just...