PostgreSQL Grouping Sets: Subtotals With Clear Row Meaning
Use PostgreSQL grouping sets for explicit subtotal levels, grouping markers, correct null interpretation, and aggregates that avoid double-counted totals.
Hackbitez journal
Category: Software
PostgreSQL pg_trgm: Similarity Search With Measured Limits
Use PostgreSQL pg_trgm for measured text similarity and pattern search, with query-matched indexes, tested thresholds, short-input checks, and access filters.
Kubernetes Downward API: Metadata Without API Credentials
Use Kubernetes Downward API for selected Pod metadata and resource fields, with explicit delivery mode, update handling, fallback checks, and privacy limits.
Docker Cp: File Transfer With Explicit Paths and Ownership
Use Docker cp for deliberate file transfer with verified container identity, destination rules, ownership checks, symlink policy, and consistent source data.
HTTP 303 See Other: Separate Submission From Retrieval
Use HTTP 303 See Other for post-submission retrieval, with approved destinations, authorized confirmation pages, retry safety, and tested client behavior.
Python Queue: Backpressure and Honest Task Completion
Use Python Queue with bounded capacity, correct task accounting, visible worker failures, and deliberate shutdown rather than assuming join proves success.
Python Tomllib: Parse Config, Then Validate Its Meaning
Use Python tomllib with binary file reads, explicit schema checks, resource limits, deliberate numeric conversion, and safe configuration activation.
Git Describe: Build Labels Need a Stable Ref Policy
Use Git describe for traceable build labels with explicit tag policy, history coverage, dirty-state handling, and a full commit identity alongside the label.
Git Diff No-Index: Compare Files Without a Repository
Use Git diff no-index for explicit file comparisons, with correct exit-code handling, controlled diff helpers, private output, and clear comparison scope.
PostgreSQL Enums: Stable Labels and Planned Evolution
Use PostgreSQL enums with stable labels, deliberate ordering, compatible clients, and staged schema changes rather than treating values as easy-to-edit text.