The little server that could
summary
Builds a tiny Ruby web server that actually opens an internet socket, binds and listens, accepts connections and writes HTTP responses via the Rack call convention; shows a real 5 second sleep to prove blocking on external API calls and then adds forked child processes to avoid serial waits; demonstrates why Ruby GC behavior breaks copy on write before Ruby 2 and how SIGINT trap lets the process shut down cleanly; honest about being slow, insecure and not production ready.
AI-written summary · about
chapters (10)
- 0:00Start
- 1:44The Little Server That Could Can't
- 3:24Servers are an Abstraction
- 3:44Servers are not Magic
- 5:311. Interacts with the OS to talk to the outside world 2. Conforms to a specific API
- 7:32Literal "Spec" & Figurative "Spec"
- 12:55Defining our Socket: 1. Addressing Format 2. Kind of socket
- 17:261. The Outside World 2. The Application 3. You, the Developer
- 28:01Benefits of Threading: - Uses less memory -Threads die with process - Much faster to communicate
- 28:27GIL Global Interpreter Lock
related talks
36mThe Little Server That Could
35mLet's build a simple HTTP server with Ruby
37mRuby on Rack: The Magic Between Request and Response
18mThere is no Magic: Write your own HTTP server, and then stop
29mReal Time Rack
26mConcurrency in Ruby: From fork() to Fiber
31mFrom Rails to the Web Server to the Browser
36mBuilding a Ruby web app using the Ruby Standard Library