Science Quote by Ken Thompson

There’s an almost deadpan understatement to Ken Thompson’s line, the kind that signals a builder talking about necessity, not philosophy. “I wanted” is doing quiet work here: it frames a foundational operating-systems idea as a practical craving, like a tool missing from the bench. And the tool isn’t “virtual memory” in the abstract, the textbook concept that makes RAM feel bigger. It’s “virtual memory, at least as it’s coupled with file systems” - the hinge that turns memory into something you can name, persist, share, and recover.

The intent is engineering leverage. Virtual memory by itself is a trick of indirection: map addresses to something else. Coupled with a file system, it becomes a unifying interface between two worlds people used to treat separately: transient computation and durable storage. That coupling implies paging, memory-mapped files, demand loading, swapping - mechanisms that let programs act as if data is simply “there,” regardless of whether it’s on disk or in RAM. It’s not just convenience; it changes what kinds of software are feasible, because it changes the cost and complexity of handling data.

The subtext is a worldview that shaped Unix: keep abstractions small, sharp, and composable, then let them interact until the system feels bigger than its parts. Thompson isn’t selling a grand theory; he’s admitting the motive force behind much of computing progress: impatience with friction. In one line, you can hear the era’s constraints - scarce memory, slow disks - and the ambition to make those constraints disappear behind clean, reliable semantics.