HTTP Headers
Tung Nguyen
10 min read
Between the first line of an HTTP message and its body sit the headers. These are the Key: Value lines you have seen in previous examples. Think of the request line as the action and the body as the payload. The headers are the metadata. They are the extra notes both sides attach to explain how the message should be handled. There are hundreds of defined headers. That might sound intimidating, but they actually fall into just a few simple groups.
Headers are just key and value
Mechanically, a header is very simple. It is a name, a colon, and a value, like Content-Type: application/json. Header names are case-insensitive. This means Content-Type, content-type, and CONTENT-TYPE are all treated as the exact same header. You will see different capitalizations in the wild, but they all mean the same thing.
Anyone can invent a new header. You might see a name starting with X-, such as X-Request-Id. This is an older convention for custom, non-standard headers. The X- prefix is more of a habit than a strict rule today. If you spot a header you do not recognize, it is often one that a specific application created for its own internal use.
Four mental buckets
The easiest way to make sense of headers is to sort them into four main buckets.
- describe the request or the client making it. specifies which website you want to reach. identifies your browser. tells the server which file formats you prefer.
