A better interface for Kindle
Extreme makeover, ebook edition
I’m late to the ebook thing, perhaps hopping on the movement just before it coasts to an end, but a week ago I inherited my wife’s Kindle Paperwhite.
The e-ink device is, so far, a happier place to read long-form than is my iPad Mini, that much I’ve decided. I’ve also decided that Kindle could be better still. Kindle smartly solves some issues with long-form digital reading, but its UI feels like a gen-1 product, not seven-year-mature design.
(My experience with other e-readers is limited to a handful of iPad and Android apps, none of which do any better.)
The issues as I see them.
The Kindle’s simple UI is designed to closely mimic the function of a paper book. Text is presented and navigated one page at a time. Progression through an ebook goes left to right.
Simple function and familiar form are great methods to introduce a new concept to old users, but can also hold back useful advances in interface. Apple ditched skeumorphism in iOS, not just because felt-covered apps are embarrassing in this contemporary decade, but also because relying on metaphors based on old technology limits what can be done with user interface. (UI in Game Center and Apple’s Podcasts apps was badly cornered by reliance on old metaphors.)
Paper books read left to right, cover to cover, but everything else we use — with rare exception — scrolls top to bottom. Flip pages on the Web? No thank you, and I like my Twitter feed scrolling to infinity (or 2006).
A vertical scroll isn’t just “modern,” I wouldn’t advocate Kindle ditch page flipping for the sake of appearances. But a vertical scroll — combined with other UI improvements — could go a long way in solving Kindle’s bigger problems.
After a week with Kindle, I am continually challenged by position and context.
Position and context
Position and context are the biggest challenges for consuming long-form, digital text. Position is maintaining progress in the text, the ability to walk away from reading and come back text later to the same place. Context is knowing the relation of position in context of the rest of the text, how far through a book, how long before the next chapter, what was written two pages ago, who said what.
Kindle smartly solves position and context by removing them from the user’s mind. I read a book. I close the book and load a new one. I jump to another device. Kindle remembers for me where I left off. Kindle reduces context to page forward and page back, though Amazon usefully complicated the interface slightly to improve contextual positioning when the company introduced an estimated time to the next end of chapter. (That estimated time remaining annoyingly hurries me to read faster, though my wife — a heavy Kindle user — swears the impulse wears off.)
It’s smart, but incomplete. Position and context are still issues on Kindle.
I often need to flip back a page or two to re-read a detail that later becomes important in the book. And while a countdown to the end of a chapter is helpful, I don’t need a chapter end to stop reading, just the end of a thought or excerpt, a useful break in the text, and finding one requires flipping forward pages.
Either case results in lost position.
I’ve wanted to jump to the start of a book to pull a detail from the front matter (year published), but neglected the urge because maintaining position in an ebook feels a bit like riding a river on a life raft. As long as my progress is forward, I can hold onto my position in the raft, but wading backward or forward in the book gives me the anxiety that comes with temporarily abandoning the raft. I’m confident I can find my way back, but I better not swim far.
Kindle’s tendency to reflow text makes finding a lost position especially difficult — the page I left isn’t guaranteed to look the same when I find it again.
This plays into context, too. Did I flip backward four pages or five? Though Kindle displays content in pages, it actually knows content in terms of location values. The location value is useful for the device, but not very useful for a human. At first blush, location values seem a modern substitute for page numbers. In practice, the values don’t offer nearly the same context.
I’m currently at Loc 4734 in the book I’m reading. It’s not an especially long book, but that number is large enough to lose much meaning. I have no concept of how 4734 fits into the overall structure of the book; how many Locs are in this thing? Worse still, moving to the next page jumps to me Loc 4741, which isn’t a predictable count, and if I navigate backward or forward and Kindle reflows the text there’s a chance I’ll never be able to see “Loc 4734” again. As a reader, the location value tells me little and I won’t remember it.
And so while Kindle would have me not worry about position and context, in reality I often do. And the current interface doesn’t do much to help me.
How I’d solve these problems.
1. Progress bar
Forget page numbers and location values. The former are meaningless for digital, and the latter are unusable for humans. I want context for my progress in a book, my positional relation to chapter beginnings and ends, in a way that makes sense for the medium and makes sense for me.
A subtle progress bar serves both requirements. In this mock, light dots mark measured positions in the book (percentages, or whatever makes sense for the space). Hollow circles mark chapter beginnings. A dark circle marks my current location, simply contextualizing my progress against the overall structure of the text.
(Older Kindle e-ink devices have a horizontal version of the progress bar, but without the functionality I’ll describe later.)
My mock shows a progress bar that admittedly consumes a significant portion of the screen, but consider these points: (1) This is a mock to illustrate a concept, the position and size of the bar aren’t critical to the concept; (2) The progress bar could disappear and reappear based on touch inputs if desired; (3) I have further interface needs for the bar as illustrated. Stay with me.
2. Review / preview lines
Why doesn’t every line of an ebook start on a new page? Because moving from one line to the next is easier with context, and jumping from one page to the next costs lost context.
But with digital text, that context needn’t be lost. By showing the user a grayed-out review of the last lines of the previous page, the newly-loaded text comes with familiar context, easing the transition to a new view. Further, the UI can provide a preview of the text from the next page, reducing the jarring mental jump required to flip pages in a traditional paper book.
I earlier promised more functionality from the progress bar, and Snapscan is the payoff. Snapscan is an interface for thumbing through pages of an ebook without abandoning position.
Touch and hold over the current position of the progress bar, and slide upward or downward to scan through adjacent pages which preview in full on the screen. Release touch from the progress bar and the interface snaps back to the last-read position.
This could work for any position on the progress bar, allowing the user to easily grab a detail from the front matter or pull a definition from an appendix and then simply lift a finger to return to reading.
(Gestures allow a more durable navigation with Snapscan—i.e. navigation that outlasts the touch input. A user can, instead of releasing touch over the progress bar, slide the finger to the right into the text area, and then release to more permanently navigate.)
Other minor UI improvements. Just getting greedy now.
When I read text on a computer screen or smartphone, I don’t read one page scroll at a time. I constantly pull text into my view, rather than stretch my eyes to the corners of the screen, often one line at a time. I do this partially to preserve visual context for my position — the same reason I recommend the review / preview lines — and partially because my eyes are most comfortable at about yea-high.
Scrolling text one line at a time on an e-ink display is likely not the answer (e-ink refreshes are harsh) but I can see use for scrolling blocks of text smaller than a whole page refresh.
Half page scrolls, single paragraph scrolls … With the grayed-out review / preview lines illustrated earlier, shorter scrolls could play a lot like iA Writer’s Focus Mode, which usefully limits the reader’s focus to a couple of lines at a time. (This is especially helpful for keeping place moving from one line to the next in large blocks of text.)
I’d previously tried reading ebooks on tablets like Nexus 7 and iPad, but found the devices too sensitive, too easy to accidentally trigger a page transition or, even worse, close the app. The Kindle’s touch screen is less sensitive to accidental input, but it’s still a bit delicate, largely because no portion of the screen is “dead” to input. That is, touching anywhere on the screen does something.
I’d add a dead zone between page-forward and page-backward touch zones. And if we’re assuming the new vertical interface illustrated above, I’d like touch zones to be top and bottom, not left and right, with a big fat dead zone between. (The dead zone can still respond to swipes, while remaining oblivious to accidental tap inputs.) The larger bezel of the lower part of the Kindle Paperwhite makes the bottom of the screen less prone to accidental taps from resting thumbs. And whether the device is held left- or right-handed, the reach to the page-forward input is the same.
Fixed flow text
This is a minor point, but nevertheless an annoyance that makes position and context on Kindle more challenging than necessary. The flow of text should be fixed and consistent, assuming the user sticks to one font size. If the user is flipping pages at a time, keep the word construction of the pages consistent. Page flow can be consistently calculated based on chosen font size, and doing so makes a subtle but positive improvement to ebook usability.
Any one of these improvements could work on its own, and as with any UI work it’ll take a few prototypes to work out everything. If you’re a developer interested in hacking Kindle’s UI, reach out to me on Twitter. E-ink hardware may or may not have a future in today’s market of cheap color tablets, but advances like review lines and Snapscan will usefully improve the experience of reading long-form digital text. Kindle’s gen-1 solutions are overdue a contemporary makeover.