This article (http://aeon.co/magazine/technology/why-have-digital-books-stopped-evolving/) isn’t wrong about digital book design stagnating, but it’s not entirely because of the Apple or Amazon ecosystems, it’s mostly because of the lazy way publishers produce digital editions, which most often appears to be a “Export to ePub” from a word processor rather than a page layout program.
And the actual open ePub 2.0 format is horrible. It’s based on XML 1.0 and even placing images or using custom fonts can invalidate the format – though some e-reader apps ignore all that stuff anyway. The ePub 3.0 format is much better, but is still based on XHTML circa 1999. It’s a bit better with fonts, images, and other web-like features, but still sucks, easily fails validation, and many eReader apps still don’t support the enhanced features.
One of the biggest issues with the ePub format compared to a physical book is pagination. Because you can change the font size it shifts the page breaks, it can push images onto pages by themselves, it can cause weird breaks in sentences and paragraphs. With ePub 3.0 you can define custom fonts, but as I noted almost no e-reader apps will actually USE them. Last I checked, Kindle for example, is limited to just 7 fonts. It all basically makes even the best book layout job look like someone threw the pages in a blender.
Apple’s iBooks format is much better and can even be gorgeous. It takes the best features of actual book layout and web features. Amazon, too, has a custom Kindle format that is similarly capable of beautiful, interactive book layouts. However, while both of those formats are based on HTML5 (which would work in ANY modern web browser) they are, of course, instead locked into their respective ecosystems. Not for any technical reason, but for monetization reasons, they have taken an open format (HTML5) and bundled it up in a proprietary package.
PDF is arguably the best and most robust format for digital books. It offers support for all of the same features as other e-book formats (annotations, highlighting, searchable text, bookmarking, interactive functions) but it preserves the EXACT look of the paper edition. Images are exactly where they are meant to be. Fonts are exactly the fonts the book designer intended. Page breaks are where pages end (unless you purposely “Reflow” the view).
Also, if the PDF doesn’t have DRM (because that’s also a supported feature) you can freely share it and open it on any device that has a PDF viewer — which is pretty much every device. Exporting the PDF file from page layout software is easy – in fact it’s a core function of the software because that’s the format in which the book is sent to the printer! PDF is already the industry standard for designing books, and both Kindle and iBooks can open the files, so I really don’t understand why publishers aren’t just using (or insisting on using) that as their default eBook format?
Of course that all still requires you to download files, which must either be stored on your device(s) or in the cloud somewhere. But if your resigned to the idea that the publications will be stored in the cloud somewhere wouldn’t it be better if the system wasn’t inside incompatible ecosystems but, instead, was decentralized like the Web itself? That was one of the notions I had when I started creating my “Ryuzine” publishing platform.
While Ryuzine is more suitable for periodicals than books (but only because of cache limits that tend to crash mobile browser if the publication is “too big”), I had noticed that people I know who self-publish webcomics didn’t have an easy way to get their comics into a digital app format, particularly for devices like the iPad. Licenses to publish for iBooks are not exactly cheap. PDF files can get downloaded and passed around, as can CBZ and other comic reader formats, but then you have no control over redistribution and therefore no metrics on how many people are reading it. Until recently apps like Comixology didn’t have an option for independent creators (and even now that they do the app store cut of the take is, in my opinion, egregious).
Ryuzine is based on HTML5 so it makes sense the “e-reader” app is just your browser on whatever device you’re using. It’s really just a web page that can be hosted anywhere (either live or as a downloadable archive if you want to support reading it offline). No app store, no approval process, no partnership agreements, host it where-ever you like, put it behind a paywall if it isn’t free. Sharing is as easy as sending or posting a URL to wherever the publication or download is hosted.
As for things like annotations, I wasn’t too sure people would be interested in that for a digital periodical format, but since it’s really just a web page there ARE third-party web annotation services that should work with it, as should bookmark sharing services. The app has a bookmark function built into it, but it’s stored in a browser cookie so it’s specific to that browser on that device.
So, in conclusion I’d say that I’m of the opinion that it isn’t the digital book format itself that has “stagnated” and not because it’s inside some locked-down ecosystem. It’s the publishers who are simply not taking advantage of the more robust, feature-filled formats available to them even within the walled-garden ecosystems, and are instead offering whatever they feel is “good enough” with minimal effort.