The current trends in web design include several abominations. That is not to say that there have been no positive trends; quite the opposite. Today’s websites are tremendously better than their ancestors. However, with change comes mistakes, and there is one mistake that I wish to call attention to today, the infinite scroll.

With the advent of new technologies, it became possible to dynamically add content to a web page rather than requesting a new one, a relatively bulky operation. Some intrepid designer followed this rabbit and thought, “Hey, when we reach the bottom of the page, we can just automatically load more content. How cool is that?”. Cool it may be, but the novelty quickly wears off, and once it does, the user is left in an awkward position.

The easiest comparison is with content that is paged. The standard interface includes controls for paging forward and backwards and, most importantly for this topic, a display of how many pages of content there are. The user can instantly see if they are on page 1 of 100 or page 1 of 2. This is an important piece of information when browsing. In an infinite-scroll format, this information is hidden. The user has no way of knowing where they are in the flood of content: is it a puddle, is it an ocean, are they even near the shore? The infinite scroll undermines the good work of the scroll bar.

Ah, the scroll bar, an elegant way of displaying where a part fits into the whole. When viewing content that does not fit in a window, the scroll bar clearly indicates how much content there is beyond what is currently on display. The user has a clear reference for where they are and how far they have to go any direction. Now, add an infinitely scrolling page to that, and you have essentially untethered one end of the scroll bar. You take a look and see that you are three-quarters of the way down the page. All seems well, but then, all of a sudden, the page grows, and suddenly you are but halfway down the page. Something you passed at the halfway mark is no longer at the halfway mark, and where the madness ends, you have no way of knowing. This may seem petty, but people understand things spatially (above, below, before, after, half, thirds, proportions galore). Removing the spatial references provided by the scroll bar leaves the user a bit lost, and for what?

The next problem is perhaps even more basic: the infinite scroll interferes with hypertext, one of the fundamental technologies that the web is built on. Suppose you have a page, some hypertext with a link. You click the link which opens a new, more different, page. Later, you decide to return to the original page via the magic of the ‘back’ button in your browser. Life is good. Apply the infinite scroll (I’m looking at you, tumblr) and repeat the process. Everything is fine until that fateful step backwards. Now, you are returned to a page which does not (yet) contain the original link you used. You have to scroll and load and scroll and load until you find your place again. This is inconvenient or, as some would call it, bad design. Sure, you could open links in new tabs or windows to avoid this problem, but that’s just fatalistic pragmatism.

Another flaw of the infinite scroll is that it only goes in one direction. Take an example from the recently updated the Onion website: I was happily reading my favorite feature, grumbling quietly to myself about the infinite scroll, when lo, disaster struck. The page befuddled my poor mobile browser, so it helpfully reloaded the page. On an ordinary infinite scroll, this would just be a tremendous inconvenience as I would have to scroll and load and scroll and load until I found my place again, but this was no ordinary infinite scroll. No, this page was smart enough to record my place for me and adjust the URL accordingly (sort of black magic I presume), so when the page reloaded, I was right where I left off. Practical? Fabulous? Genius? Hubris, I say! Now, I was unable to scroll up! Everything I had passed was lost to me. Popping back up to check something in a previous entry was now impossible. Since I was provided with no navigation tools in the page, the only way to “scroll up” was to return to the original URL and scroll and load and scroll and load. This unspeakably clever webpage seemed bent on preventing me from simply viewing its content. Woe!

Now, paging controls have their pitfalls and pratfalls, which I will eagerly discuss at another time. I do not wish to imply that they represent some sort of Platonic content parceling ideal. However, the infinite scroll falls into a rather more sinister category of missteps: hiding information and removing control in the name of streamlining the interface. The temptation to simplify is strong, but the infinite scroll is a misguided sacrifice on that particular altar. It streamlines the experience, sure, but at the cost of versatility. It simply does not scale to meet the needs of a mature volume of content. This is why we have dates, archives, tags, categories, search, and all manner of wonderful means of organizing and browsing content. Without them, we would be back in 1998, simply scrolling until we find what we are looking for.

Let us call the infinite scroll an experiment; an exercise in testing the capabilities of the modern webpage. The result is clever, perhaps even useful (in the right circumstances), but it is not, by itself, good interface.