Embracing the Web with JSON and PLV8
summary
Uses Postgres 9.2 JSON plus the PLV8 JavaScript PL to treat rows as documents, build functions (mustache, jsonselect) and create expression indexes that are slow to build but fast to query. A prototype daily-standup app runs on Node/backbone/Postgres, checks hundreds of DBs every 30s, and moves from prototype to production without migrations. PLV8 cuts a Fibonacci demo from about 3 minutes to 300 ms, and server-side PLV8 validations stop bad JSON writes so the approach actually works.
AI-written summary · about
chapters (46)
- 0:00Intro
- 1:50Who is a DBA
- 3:40Schema Agent
- 4:00Why People Love Our Tomorrow
- 4:23Compute the Evidence
- 4:53Nested Data
- 5:22Document Structure
- 5:40Wally
- 6:08Transactions
- 6:26Document Storage
- 7:54Rails
- 8:14Adders
- 8:41PostPost
- 9:12SeePaul
- 10:50Kill
- 11:36JSON conventions
- 12:10JSON is easy
- 12:45JavaScript
- 13:15Chrome
- 14:32Fibonacci
- 16:29Example document
- 17:00GET function
- 18:10INDEX function
- 19:40Jason
- 20:07Janice
- 20:19JavaScript libraries
- 20:39JSON template
- 21:10Jason select
- 21:27Jason select code
- 22:12Startup function
- 23:21Adding date
- 23:38Bad idea
- 30:02Simple things
- 30:31Simple table
- 31:20Asking for help
- 31:53Adding authentication
- 32:46Jasons luck
- 33:01Easy to do
- 33:51Relevance integrations
- 35:28The biggest success
- 35:57Mental overhead of switching languages
- 36:58Benefits of using the same model
- 37:18Validation
- 38:03Checks
- 39:49Metadata
- 40:21Developerfriendly documents
related talks
33mSchemaless SQL The Best of Both Worlds
33mExtending your database server's abilities
51mBuilding a Realtime "Backend in a Box" with PostgreSQL and Vue
31mUsing Databases to pull your applications weight
30mPostgreSQL & NodeJs as Core Feature of Backend for Frontend
46mUse PostgreSQL, not Rails, to make Rails Faster!
24mThe Upsidedown Stack: React on the Server and Postgres in the Browser
34mThe Road to Hype: How NoSQL became Serverless