What actually happens before a web page reaches you
A page feels like a document sitting on a shelf, waiting to be picked up. Almost none of them are. Most are built to order, in the moment you ask for them, by a machine you will never see.
Open a site and something arrives almost immediately. The speed makes it feel like a file was simply handed over, the way a book is handed across a counter. Occasionally that is true. Far more often the page did not exist a second earlier. Your request travelled to a machine, that machine worked out who you might be, gathered the pieces it needed, arranged them into a finished page, and sent the result back. Then it forgot the whole episode and moved on to the next visitor. The page you are reading is the residue of a short burst of work that has already ended.
This is the split that gives our name its joke, and we will only make it once: a child element sits inside a parent, and the work described here sits inside the work you can see. Everything on this site concerns the inside half. Not the marks that describe a document, not the rules that colour it, not the behaviour that runs once it has landed in front of you. Our subject is the interval before any of that: what a machine does with your request while you are still waiting.
Continue reading