e-book formats will not be the way of the future.

11 posts / 0 new
Last post

I maintain that epub and other e-book formats have got it wrong, IMHO.

It is not necessary or helpful to import so many book metaphors into the online environment. Specifically, most non-chapter page breaks are arbitrary and unwelcome intrusions into the flow of the text. They are a necessary evil when printing a book, but wholly unnecessary online. They add nothing to the comprehension of the text, but rather tend to get in the way,

It is a quite entertaining novelty for a while, in an ephemeral way, to simulate a book, with dogeared page corners, and animated turning pages, etc. But very quickly logic will prevail, and people will come to see that one thing they don't need is a 'book'.

A lot of the enthusiasm in the past for print formats online, such as white papers, books, download, epub, pdf, and the Times New Roman font, was that content apparently intended for print had been much more in depth than the typical web page. This explained why such content was welcomed and indeed sought after. But increasingly web pages have high quality content: for example, there is wikipedia.

Eventually people will realise that there is nothing desirable or special about a 'book'.

However, serious fiction and non-fiction content intended to be read online needs to reflect the realities of the medium eg use a sufficiently large sans serif font, and not cramping all the readable content into a small area, while making sensible, mature, well structured use of hyperlinking and anchors, etc. Particularly for non-fiction, the online environment can be so much more powerful and usable, because the content can be layered, with plenty of references and citations inline, rather than being a simple, long sequence, maybe with footnotes and endnotes.

It is fundamental to modern online use to employ document windows, and the content should still display sympathetically when a window is resized, which is often done when comparing and contrasting with some other text, or making notes, or even checking one's mail or calendar.

These are the things I mean when I say that the online experience, considered correctly, has fundamentally different requirements.

One oddity: currently printed pages have margins, whereas web pages have indents. Is this distinction real? It tends to lead some inexperienced people to produce online content with no lareral blank space, which can seem very heavy and indigestible. One should make good use of screen real estate, as said before, but not to this extreme.

While I respect your opinion, if you reread your rant you'll find it has less to do with the format than reading systems and the state of publishing generally.

One of your three main contentions is that reading systems do a suboptimal job of rendering content in virtual pages, but that has nothing to do with epub as a format. You can get a reading system that lets you scroll documents if you prefer, and then an epub will render more like a web page. Readium has the option to toggle between paginated and scrolled views, for example. IBIS reader, Safari Books and many other cloud readers also exist. EPUBs are web content, so what you say is optimal about the web can be done with an epub. EPUB is built on the web stack, so in arguing for the web you're not making an argument against epub, but supporting the direction the format has taken.

Your second contention is that publishers still have lots of print artefacts in their ebooks. Again, this isn't an argument against epub. EPUB defines the format; it doesn't dictate how to use it. Publishers would likely be doing exactly the same thing if reading had evolved on the web proper. It oversimplifies the reality of a print-driven publishing industry to say print should just be abadoned. For better or worse, much content is formatted for print and then channeled to digital. Your dream of a better digital experience is one many people share, and will no doubt come as the market shifts more in a digital direction. Nothing in epub has to change to allow what you're suggesting, though.

Finally, your third contention is that people don't like to gain knowledge through books. I hope you can agree this is just opinion, and again nothing to do with epub. Wikipedia may be okay for looking up simple facts, but it's not many people's first choice for learning meaningful information. I think the book has a long future, whether from here on they're called ebooks or some other nomenclature evolves.

EPUB is also called epub by design, so as not to lock it into a "book" mindset. Any document you can publish on the web can be rendered as an epub. I believe we are at a pivotal point in the history of publishing, but adapting from a physical to digital medium is going to take some time. Don't be too hard on epub for the lag, though!


Epub is a layer laid on top of HTML, so must do things that are already included in the HTML set of productions, but maybe arguably done in an easier, nicer way. One would not want the overhead of another layer, and doubtless also losing some of the facilities present in HTML, unless there were some very cogent and persuasive advantages or convenience in using it. Most people think epub is mainly about producing book format, something I find a questionable advantage. Are you saying there are other valuable things that epub format can provide? If so, what are they? Even if epub passes through all the facilities in the underlying HTML, that would still not be a reason to use it, unless it was adding something, even just in ease of production? It seems odd in the extreme to defend a format in principle, if one cannot point to examples of valuable outputs or facilities that are provided by it, That seems to be abstraction utterly running riot. But I am encouraged that you appear to concede that paged display online may not be the answer for everyone. Is epub mainly a convenient way to place an existing book file online? I thought the virtues of the format were held to be far wider than that, as a new and important paradigm of communication? (To clarify: I am very interested in the fascinating issues in user interface and the psychology of perception thrown up by all this. I am not interested in just trying to score points on the internet, which seems to me a puerile thing for anybody to want to do.)

That there is no universal reading preference or modality was never in question, that I recall. EPUB by design accommodates multiple modalities, more that just how you prefer to scroll content. Whether text/audio playback through media overlays or enhanced text-to-speech or being able to read the text through refreshable braille displays. A visual fixation on reading neglects many of the strengths of the format. I find scrolling a severe limitation to reading, and doubt very much that I'm alone, since it displaces your train of thought. Being able to continuously tap and load content with predictable rendering to your top/bottom, left/right language expectations was no small advancement over desktop browser reading.

The scroll had its day, around thousand and some years ago. HTML was a return to the scroll, and it's not surprising that readers are generally not up in arms with current reading systems taking a paginated approach. If anything, you could argue that reading systems solved a major limitation of the web and have a good case to back you up in reader adoption.

As far as reading system trying to represent a print page goes, and whether that is a good or bad thing, you're taking a personal preference and projecting it onto what you believe readers should want. But who are you to say that a reading system should present a digital page and readers are wrong for wanting the familiarity of a book? How do you explain the sudden explosion of ebook reading around 2007 as reading systems and content became available in specialized ebook formats? If you're going to make suggestions that everything is wrong about ebooks, you need better evidence than wikipedia, which itself is in decline last I heard. I'm not saying what we have now will last the next 2000 years, but to get from here to the future is a progression that right now is working its way through ebook formats. If something better displaces the model, so be it. I'm just not convinced it's the web as it is.

What I think you might be too-readily throwing under the bus, as well, is the packaging and standardization that the format provides over something as loose and wild as the web. It does provide ease of production and distribution. An epub is a self-contained online/offline reading package. That alone trumps the online experience for readers who want to read where they want when they want. Naturally that entails some limitations from the full web experience, but readers have already shown those limitations are not always critical to their consumption and enjoyment of the content.

But to do away with epub, what is your proposal for its replacement? How does an author create and distribute their content? Does everyone who wants to write have to now buy domains to host their books on? Bookstores would apparently be gone since you couldn't sell your web site through a store, so does every author also need to buy an online payment system for their site, plus develop a secure interface for cloud reading? If someone is going to host all these books, don't you need an agreed on format and conventions for upload?

These are some of the reasons why publishing is enamoured of formats. It simplifies production and removes it from distribution and consumption. Content creators can work on content, vendors can work at selling and reading system developers can focus on the user experience. Remove a common format that can work from one end of the chain to the other in a reliable and predictable way and what do you have left? To facilitate another model, I expect you wind up with a framework remarkably similar in functionality to epub.

But I'm not here to defend a format for the sake of defending a format. From an accessible content production perspective, which is my background, epub 3 has enormous potential to get accessible production back to the source, and get readers who have been left out from print access to content as soon as everyone else. That's why I don't disagree with you that epub reading systems can present usability challenges or that it is possible to make bad content. I just don't think it's fair to equate it with the format alone, as it's a complex issue that crosses production and consumption. And the web has failed in this regard so many times, too. No one would call the web an accessible place. I'm willing to give epub a shot and see if we can't do better by subsetting what the web offers.

One quick point: I apologize that I had not realized – or, more likely, had forgotten – a crucial difference between .pdf and .docx on the one hand, and .epub on the other, which is that .pdf and .docx documents are nearly always paginated for some notional paper stock, even if they are never printed, making the page boundaries very unnatural and annoying online. Whereas .epub is paginated for the actual device and window it is displayed in, which makes a whole lot more sense as a display rendering. I can understand how scrolling might be considered unnatural and annoying, especially when trying to read quickly. I agree that this kind of pagination is useful, and it is amazing that the common web browsers do not (as far as I know) do it.

I suppose a third possibility might be to have a display that you can easily and naturally start and stop scrolling at a reading speed that you set, somewhat like a tv autocue. Something that doubtless assumes an experienced user, but could be very useful for reading out loud blind in a good way, with full expression, when making speeches, or sales pitches, etc.

I just want to put forward some arguments in favor of non-chapter breaks. My biggest problem with buying Kindle books (beyond those that are poorly adapted to the format) is the LACK of page breaks. Yes, I love the variable font size BUT, with many of the books I read I want to go back to and re-find something I read. This is MUCH easier if I can use a visual memory of what the layout of the page, what nearby pages looked like, where the sentence was on the page, and which side of the book it was on. IF I can remember exact wording, I can do a search but often I don't remember the exact wording.

This level of memory organization can not be done without something like a page format. While it may not be important for most leisure reading, it is crucial (or at least very beneficial) for informational reading.

Other related benefits to matching the printed page format (for books that are in print) is that one can pick up a printed version and find something fairly easily (at least if it is converted from that particular print version). It also allows referencing by page. While this is not a precise as using a long number that indexes a particular word or some other fine division, a page number is easier to remember when trying to look up a reference -- and its backward compatible for existing books.

I would like to see the continuous format that allows me to have a dynamic page layout that is determined by font size, where I "jump" to in the book, etc. but also keep the traditional pagination information so that "classical" page view is possible

Perhaps the breaks need not be the classical page breaks, but some organizational detail that is useful to human memory systems is important, IMHO.

I am currently in the process of converting the (paper) books that I publish into digital form, with EPUB being the probable choice format. However, looking at the actual nitty-gritty of the the format, I can but only shake my head in disbelief. Most of my misgivings have been mentioned in this thread, and I shall briefly elaborate.

"EPUBs are web content ..." "Any document you can publish on the web can be rendered as an epub." "Epub is a layer laid on top of HTML"

In my opinion, the current specification of HTML5 is poorly suited for the structural markup of the content found in a book. Sure, we can use HTML+CSS to render the document to look just right to a human, but the structure for use by an UA just isn't there. The solution that is used by the EPUB standard of custom methods (epub:...) is a baling-wire and spit solution, because although it fixes what is broken in HTML, it also causes EPUB'lications to be invalid HTML markup for use on the web. So, although EPUB is HTML, HTML is not EPUB.
What I mean is that EPUB uses a fundamentally flawed markup language and that these flaws are now fully embedded in the EPUB language as well. It would have been so much better if EPUB had rather defined their own XML that specifically took into account the special requirements of digitally publishing content. A prime example of this is the whole h1-h6 debacle. What we gain by using EPUB is negated by EPUB's use of HTML.

I also just read this article: http://epubzone.org/news/building-an-epub-3-template

The solutions discussed are, again, fundamentally flawed because they are founded on a "web" mentality and not based on a publishing mindset. For example, the "solution" for the <aside> tag. Yes, that is how one would tackle the problem when coding a website, but definitely not a digital book. The reason being that the former is loaded into a browser by way of a request made to a server. Rarely is the source file stored off-line and thus it is very easy for the author to change the code as UA's evolve. (You know what I mean.) The same is not true for a digital book, which is bought, downloaded and stored. Once a digital book has been downloaded, the author/publisher cannot change the code and thus all the kludges and fudges are fixed in those copies forever. And again, the use of the HTML specification by EPUB is the culprit.

Well, I tried intensively to convince the W3C HTML5 working group to provide more elements with semantical meaning concerning literature to HTML5. Unfortunately there was no interest in semantics. They cared mainly about structures for short living web pages, in most cases not very relevant for possibly long living literature. And even if one really tries to use meaningful new element like section - without CSS the default presentation is meaningless - this is no motivation for normal authors to use such new elements at all - semantics in HTML5 failed by design.
At the same time RDFa was developed. To obfuscate this, the former editor of HTML5 'Hixie' tired to obfuscate even this with 'microformats' - fortunately this has ended as a note, not a recommendation, therefore nothing one should ever use.
EPUB 2 had DAISY/DTB as an alternative to XHTML as a more relevant semantical alternative. But obviously this was never implemented in common viewers. Now we stick with these nasty HTML5 drafts, which might really become a recommendation this or next year. Well, concerning long living text and literature, it failed.
But at least EPUB3 allows the usage of RDFa, therefore one can point to more relevant formats with semantical meaning like DAISY, docbook or LML (I specified for this purpose) - together with CSS one has both semantical meaning and proper presentation.
There is as well TEI relevant for text information. But the usual suspects of viewers do not understand the semantics of such XML formats. There is no need to invent yet another format, there are already several proper formats much better than (X)HTML to markup text. But there are no widely used viewers to interpret them natively. Effectively one has to use XSLT to convert it to XHTML again ;o)
Obviously there was no need at all to require authors of EPUB to care about always changing HTML5 drafts and the confusion between the promoted tag soup notation and the XML notation relevant for EPUB or to care about the stupidities introduced to HTML5 in a backwards incompatible way compared to XHTML 1.1 or even HTML4 strict about the allowed structures how to nest elements in each other. This was not perfect in XHTML 1.1 either, but HTML5 structure is tag soup again (and what EPUB3 requires for the XHTML navigation document as well).
There are already recommendations for XHTML+RDFa to cover such needs as a better choice.

To resume - the current solutions are far away from perfect, but there are recommended methods available both for webpages and for EPUB to provide content with good markup and semantics. Authors just do not get it automatically following recommendations, they have to work hard and learn something to get it right and no HTML5 tag soup ;o)
And EPUB is a good method, to provide accessible content especially for offline reading and to provide content in an archive format. Several things I provide online, I cannot provide in an EPUB archive, but there is already a meaningful subset, that I can offer for offline reading already now. Obviously most EPUB viewers need a lot of improvement to get a good presentation for content, that is already possible now with EPUB.

Yes, It's true that e-book formats will not be the direction of the future because when I write my book many people told me that please increase the format quality and text mode.



i need a html book

If developers can't be bothered to build an EPUB 3 reading system that is at least approximately full featured, this is certainly true. EPUB 3 will die a quiet death in the quicksands of time, which is a shame because I personally think the EPUB 3 format is superb.

Secondary menu