Glean 拾遗
中 EN
Submit a link
Daily · timeline

A few picks a day.

Fri, Aug 28, 2026 3picks
← 08-27
Calendar ▾
2026 · 08
MoTuWeThFrSaSu ·····12345678910111213141516171819202122232425262728293031
has picks today
08-30 →
06:01

PostgreSQL for Everything

一个数据库打天下？PostgreSQL 取代搜索、队列、图库的实践与边界

Drawing on years as CTO and interim manager, the author argues that PostgreSQL can replace most of your stack: full-text search, document storage, queues, time-series, vector search, cache, even graph databases. He cites Contentful, Instacart and The Guardian as real-world examples, then walks through the mechanisms: tsvector/tsquery, GIN indexes, SELECT ... SKIP LOCKED for queues, UNLOGGED tables for cache-like speed, TimescaleDB for analytics, pgvector for AI retrieval, and Apache AGE for openCypher. The post is a practical starting point for engineers who want to simplify operations, but it is an opinion essay rather than a benchmark-driven study, so the performance claims need validation before adoption.

www.raphaelbauer.com · 13 min · Database · Full-Text Search · PostgreSQL · Vector-Db
06:00

SQLite for Everything: Cut Your Stack Down to One File

SQLite 万能论：它凭什么把搜索、队列、缓存和向量库都收进一个文件

In this reply to 'PostgreSQL for Everything', the author makes the case that SQLite can collapse an entire stack into one file and one function call. It covers reusable patterns: using BEGIN IMMEDIATE plus RETURNING to turn a table into a durable work queue; indexing JSON via generated columns so schemaless writes get indexed reads; treating FTS5 as a same-transaction search index; and using sqlite-vec so embeddings, documents, metadata and full-text live in the same file, making hybrid search a JOIN. The article cites SQLite's own '35% Faster Than The Filesystem' benchmark, and claims warm-cache point lookups land around 1 microsecond versus roughly 100 microseconds for a localhost Redis GET. It also names real limits: a single writer, serialized concurrent consumers, and the need to move to PostgreSQL or specialized systems once you hit them. Useful for backend and infrastructure engineers who want concrete arguments for cutting components out of their stack before adding new ones.

joecode.com · 16 min · Cache · Database · Full-Text Search · Queue · SQLite · Vector Database
06:00

An agent skill that turns codebases into verifiable, interactive system maps

可验证的架构图生成 Agent Skill，把代码库变成交互式系统地图

Archify is an agent skill that turns a system description or an entire repository into an interactive technical map. It supports five diagram types—architecture, workflow, sequence, data-flow, and lifecycle—and produces a single self-contained HTML file with built-in navigation, dark/light themes, finite motion, and exports to PNG, SVG, WebM, and 1200×630 share cards. Under the hood it uses a typed JSON IR as the source of truth, runs validation before delivery, and returns machine-readable repair receipts on failure. It also supports Before/Delta/After snapshots for architecture review. Designed for Raven, Cursor, Claude Code, Codex CLI, and OpenCode, it suits engineers and doc writers who want verifiable, shareable visualizations without hand-drawing.

github.com · 15 min · Agent Skills · AI Engineering · CLI · Developer Tools · Diagramming