Infinite horizontal scrolling in a timeline

Question

It's known that infinite scrolling may be harmful for the usability and that sometimes it's better to set an end and provide a button to show more content.

My question is, in a web application context, what would be the impact of infinite horizontal scrolling on a timeline like this:

http://almende.github.io/chap-links-library/js/timeline/examples/example02_interactive.html

In this example, horizontal scrolling is done by dragging and dropping the timeline, so it is not continuous. Users have to click each time they want to scroll. Isn't that similar to click in a button to show more?

Answer 1

Your example does not feature infinite scrolling in the same way it is discussed in the article you have linked to. Apart from the obvious difference (drag to navigate), the user can scroll up or down to zoom in or out the timeline: that gives the user a sense of where the current moment lies in the bigger story. A user can zoom out to see the entire timeline. In contrast, in the examples mentioned in the article, the user doesn't know their position in comparison with the entire content (I'm now at page 5 of …?).

In your prototype the issue at hand is to visually indicate StartTime and EndTime at the left and right edges of the timeline, so that the user will have a visual indication of the timescale they're currently browsing.

Answer 2

A "show more" button (or action) doesn't fix any of the problems associated with infinite scrolling of the type referenced in the linked article. The problem is not that the user doesn't need to do an action to scroll more, it's that

• infinite scrolling makes it impossible to maintain track of your position relative to the content (scrollbar is useless)
• it's impossible to have any idea how much content there is
• it's impossible to skip to a specific point in the content, both future (because it doesn't exist) and what has already been revealed (because it keeps moving)
• consequently it's impossible to link or bookmark a point in the timeline or leave the page and come back to it later
• browsers don't have good usability for ridiculously long pages
• because all the content already revealed remains the page constantly grows in performance requirements (memory requirements, CPU requirements because of active elements in posts, etc.)

That user has to perform an action (click or drag) to add more content to the page doesn't fix any of those issues.

Your timeline is very different however from the linked example, so the paradigm doesn't really apply. But... On your timeline some of the problems are slightly mitigated (because you can zoom out, and "today" makes a natural signpost, and because you can go a specific month, they are effectively virtual page numbers), but other problems still remain.

For instance, how do I know when the future events stop? How do I know that there isn't an event coming in 2 months' time ahead of my current scrolled position? 6 months? A year's time? Same goes for backwards.