To download the Web and put it into an e-book file, you’d need to download some 550 billion documents. You’d need authorization for all of those, since only 25 billion or so are publicly indexed. You’d need at least a 100,000-terabytes of RAM and a specialized meta-search engine that would act as both a table of contents and index.

You’d need a reading system that could support many different content types, beyond even what a standard browser supports. Some kind of e-book operating system would be in order, something that combined all the capabilities of Unix, Windows and Mac OS X into a web browser interface.

And you’d need bandwidth. Anything that took slower than a week to download all 91,000 terabytes would be too slow. Everything would have radically changed by then, and you’d have to re-download it.

Until we get the technology to do that, it’s safe to say that the Web is not an e-book. It’s something different than a book altogether. But I think it’s also safe to say that books are still excellent vehicles for the kind of things we find on the web. Just as a photograph is a record of a moment in time, full of information that probably changed and shifted the split second after it was taken, a book is a container that creates permanent value from impermanent data.

It’s nice to think of the Web as an entity that strengthens its connections over time, builds permanence within its constantly changing layers, takes its own photographs. But the nature of interdependence and constantly shifting connections makes it a poor medium to capture itself.

This is why books, even as they become more dependent on the Web, will benefit from their encapsulation and fixity. Books could become the perfect snapshots of the Web.

A browser’s cache, something like a book, ends up storing a variety of documents that, in total, represent pathways of interest for a particular user: tours of individual sensibility through a vast maze. A digital book platform could be likewise be an editor’s tool for downloading, storing and collecting the Web. An entirely new publishing industry could spring up from culled and curated content bound together into digital editions, each connected to each other, reflective of the Web and connected to it when possible, but not fatally part of it. It would be a web of books capturing the temporal flashes most deserving of permanence.

Perhaps such an ecosystem will become the best alternative to downloading everything. Who will build a tool to make it happen?