HTTP up close: requests, answers and connections

The idea

HTTP requests are lines of text: the client writes a few lines, the server writes some back, over one TCP connection. Here the client is the laptop, and webServer is one nginx server that carries three sites.

Steps

1. A request typed by hand

   Every browser and app speaks the same few lines of text, so you can speak them yourself. nc opens a TCP connection to webServer's Port 80 and sends each line you type. Type a request for the front page in HTTP/1.1, one line at a time, and read what comes back.

   laptop:~# nc webServer.lab 80
   laptop:~# nc webServer.lab 80

2. Status codes: found, missing, moved

   A program reading the answer needs to know at once whether it worked, without reading the page. The number in the status line says so. Ask for a missing page, then a moved one, showing only the status line and headers.

   laptop:~# curl -sI http://webServer.lab/missing
   laptop:~# curl -sI http://webServer.lab/old-page
   laptop:~# curl -siL http://webServer.lab/sale