HTTP

HTTP is the language a browser or a program uses to ask a machine on the network for something, and the shape the answer comes back in.

In more detail

A call says three things: what it wants done, what it wants it done to, and a handful of short lines about itself — who is asking, what kind of thing is being sent. Under those lines comes the thing itself, when there is one.

An answer has the same shape the other way round: a number saying how it went, its own short lines, and the thing asked for. Nothing is remembered between one call and the next, which is why every call carries whatever is needed to make sense of it.

What it does on a real document

Asking for a document is one call that carries something: the description, and any file it draws on. The answer carries the PDF itself, so what your program receives is the finished document and not a link to go and fetch it.

Two of those short lines in the answer say how long reading the description took and how long the drawing took, which is what lets you watch a document's cost without measuring anything yourself.

Where it comes from, and why

It was written to move documents between machines, in the plainest form anybody could agree on, and it stayed plain enough that the same calls a browser makes are the calls a program makes.

Later versions changed how the words travel down the wire — packed tighter, several conversations at once — without changing what a call says. A program written against the old shape still reads.

See it on a finished document

The rules behind this word, and the texts they are read from

Back to the whole glossary