Static, but not merely static

Divya Sasidharan has been unpacking the Jamstack throughout January in a series of posts tagged #JAMuary. One entry, "What makes a site Jamstack?", takes on the question that dogs the term for many developers: given how closely static sites and Jamstack resemble each other, where does the distinction lie?

"You could say that a Jamstack site is a static site but a static site is not necessarily a Jamstack one."

That asymmetry is the core of the argument. The label describes something narrower than the broad category of pre-rendered sites.

Version control and deploys

Sasidharan contrasts Git-based workflows with the older FTP methods, describing how changes are made continuously in small increments and how those changes can be reverted. The case for this model rests on what it removes from the equation: mutable deploys. In her framing, mutable deploys come with a range of problems—server environments that shift unpredictably and caches that go invalid—and Jamstack sites sidestep these by taking advantage of the ability of modern CDNs to invalidate a cache completely.

Mutable deploys present a host of challenges. These include ever-changing and unpredictable server environments and issues with invalid caches. JAMstack sites do away with this by taking advantage of the ability for modern CDNs to invalidate your cache completely.

A response to complexity

The generate-deploy cycle is the other structural piece Sasidharan covers. Read in that light, Jamstack is in large part a reaction to the web's mounting complexity, and its move away from servers eliminates a substantial share of that complexity.

The full series is available on dev.to.