Saturday, August 29, 2015

New Kindle Sample Files

Two new Kindle sample files have been added to the free downloads over on my main website. All of the sample and template files have been collected together on the Tutorial Index & Resources page for easier access, and I've made all of them free now, rather than requiring a download code.

The two new files are based on the Simple and Advanced templates, but both been slightly revised with new content and notes to augment and illustrate the features available in that particular format.

The 8-page "Basic Layout" sample uses the "comic" book-type with the auto-orientation function enabled to showcase the fact that it includes both single and two-page spreads, with and without an inner margin between them. It has no active text, no hyperlinks, and no text or image pop-ups, which cannot be added if you want these two-page layouts. It also illustrates how the Virtual Panels feature works to zoom a page, or page spread, in equally divided sections that can be swiped in sequence.

The 16-page "Zoom Features" sample is based on the Advanced template, but with the orientation locked to portrait, since in order to have region magnification you cannot have page spreads, so there's little point in making landscape available (unless the pages themselves are landscape, which these aren't). This sample showcases all of the more complex features that are currently available in Kindle fixed layouts with no book-type value added. These include both text and image hyperlinks, a linked Table of Contents page (as well a menu TOC), live text with full functionality, several methods of switching text and images for "interactive" animations, a new bilingual text example, comic panel zoom with live text in both the default and magnified regions, and "lightbox" shading both within and outside zoom areas, as well as others. Note that the interactive features do not work well on the Kindle for PC app due to its incorrect positioning of active links, so this sample must be read on an actual Kindle device.

While the companion templates have previously included both the epub source code file as well as a pre-converted mobi for reference, they will henceforth only contain the template epub file itself. Since it can be converted easily into a mobi file itself there's little reason to waste the bandwidth. And while I make these available for free, I still have to pay for file hosting and delivery, which is why I had previously charged a minor fee to those who have not purchased my formatting manual. Splitting the epub and mobi files into two separate downloads will allow me to better track what is being used and what not to waste my time updating further.

My ulterior rationale for doing this is to make these sample files more readily available for potential ebook formatting clients, to whom I send the file if they're unsure what features they want included in their project. It's often just easier to show than to explain how something works.

Finally, for those who have purchased my Kindle formatting book, there will be a new revised edition released shortly, with several updates due to the changes caused by the latest Kindle firmware upgrades, as well as some editorial revision for clarification and the elimination of a few sections that are no longer relevant. I will post when the new edition is available, and send out an email to those who have purchased though my website. Those of you who have bought the book on Amazon will have to deal with their tedious file replacement process, which has not gotten better. But more on that when the time comes.

Sunday, August 23, 2015

Kindle Fixed Layout Functionality Chart Update 2015.2

For the second time this year the Kindle Fixed Layout Functionality Chart has been updated due to changes in the various Kindle reading systems. I've just completed another round of testing to determine how the latest system updates have affected functionality in fixed layouts, and some of those changes are significant, whether for good or ill.

I've also added a Change Log just above the chart to make it easier to see what's happened since the last updates, as well as adding a row denoting the latest OS version tested for each device or app.

These tests involve loading a minimum of nine different Kindle test files (more if some new variable needs testing) - each of which contain over a dozen pages that have been created specifically for this purpose - onto the seven different Kindle apps and devices I currently own. This creates a series of 72 iterations of a Kindle fixed layout file on a Kindle reading system, requiring a total of 1044 page loads (if each page is only viewed once on each system), and countless orientation changes of each device (which for the Paperwhite is a pain, to say the least), each page of which must be run through a battery of tests to determine what is working and what is not, pursuant to the ten items listed on the chart (only 9 of which I now test), and carefully noting any anomalies that occur. This tends to get somewhat confusing if one is not quite awake, and thus requires a lot of coffee (donations gratefully accepted).

The primary variables involved are the three "book-type" values (comic, children, or none), and the inclusion or absence of region magnification code in a given file. In addition, the orientation-lock variables have generally been included in the testing, with respect to page-spread functionality, which in any case has made no difference, but still has to be tested to determine that this is still true. This, of course, as with the book-type, requires a change in the metadata value, and thus the creation and conversion of a separate Kindle file for just that instance (delineated by titles such as Kindle-FXL-Comic-NoRegMag-Landscape.mobi). Region Magnification itself is dependent on its inclusion or absence within the publication itself rather than the metadata value entered, which is entirely irrelevant.

As mentioned in the notes, and obvious by the data included, I am still only testing these seven Kindle reading systems, due to lack of resources to procure more, or time to deal with it in any event. I plan eventually to get at least one HDX device, but at present just can't justify the investment, not only due to the financial expense, but as much because it frustrates me to no end how poorly the software is designed, and I'd almost rather not know what features are unsupported - or flat out broken - on yet another version of the Kindle OS. If just one feature worked the same on all devices it would make things so much easier.

UPDATE: The 'layout-blank' property has now been tested thoroughly, and the results noted. It's still essentially useless.

Saturday, July 11, 2015

Kindle Publishing Guidelines 2015.2


Just a few quick notes on this latest update to the Kindle Publishing Guidelines. Most of it covers minor details, but there are a couple of interesting items. All the sections containing changes are listed in the Revision History shown above (highlights mine). Those three highlighted sections are all deletions relating to Adobe products, and along with the removal of InDesign from the list of means by which "publishers can create Kindle books in-house...using Kindle Publisher tools" in Section 2.2, expunges all prior references to Adobe in the Guidelines, save for one mention of InDesign in Section 2.3 Third-Party Conversion Services, where it's listed as a "typical input format" for outside conversions (i.e. non-Amazon related).

2.2.1 Kindle Plugin for Adobe InDesign

The deletion of this section removes any mention of the Kindle Plugin for InDesign in the Guidelines, along with the link to its web page. That page still exists, and is linked to from the side menu of the other Kindle Publishing Tools pages, but is still marked as "Beta." The removal of it here from the section on "Paths to Getting Your Content on Kindle" implies that Amazon are giving up on support for this plugin, and no longer consider it a valid path to publication.

This is not wholly surprising, given that they haven't updated the plugin (or the corresponding Kindle Plugin for Adobe InDesign Publishing Guidelines) since version 0.973 in September of 2013. In fact, they've never updated the plugin for InDesign CC, and still list CS6 as the latest version supported. But then, they still list Windows 7 as the newest PC operating system too. Until now I just presumed they were being negligent, but the intentional removal of this section says otherwise. Were they preparing to release an update they would not have done this.

2.2.1.3 Using KindleGen

Removed the "C:>" root element from the command line example.

Added a note that "zip formats are supported for XMDF and FB2 sources" and "directory formats are supported for XMDF sources."

Changed/added the following -locale element references:
  • zh: Chinese (from ch:Chinese)
  • pt: Portuguese (from bp: Brazilian Portuguese)
  • ru: Russian (added)
3.1.1 Text Guideline #1: Body Text Must Use Defaults

A new paragraph has been inserted to the third bullet item (detailing the range of gray-scale colors acceptable for text elements with a forced color), giving specific directions on how to determine what the gray value is for a given color. See the Guidelines for details.

3.1.4 Text Guideline #4: Other Encodings Are Supported

Some editorial clarification here simplifies a previously convoluted and unnecessarily redundant sentence structure, while a second method for specifying HTML encoding by an XML declaration has been added:
<?xml version="1.0" encoding="UTF-8"?>
In addition, the first method (using the <meta> tag in the head element) has had its character set reference altered from the previous "iso-8859-1" to the more standard "UTF-8".

3.1.5 Text Guideline #5: Use Supported Characters and Spaces

Previously titled "Spaces and Unicode Characters," this section adds two new paragraphs stating that plain text UTF-8 should be used to represent characters (hence the change in the preceding section), except where XML entities are required. The second paragraph specifies the three instances where XML entities are strictly required, these being:
  • < (&lt;)
  • > (&gt;)
  • & (&amp;)
where the values in parentheses are required in order to avoid misinterpretation of the code by the reading system. For example, if ampersands are used in the opf metadata it will result in a failed conversion.

The example is provided here that the © symbol should be used rather than the &copy; named reference.

3.6.5 Image Guideline #5: Use GIF or PNG for Line-Art and Text

Removed the previous reference to the 127 kb image file size restriction. Now it just says "...fit the image size limit."

3.7.3 Table Guideline #3: Create Simple HTML Tables

Removed all the references to specific Kindle iterations, including the Kindle 1 with its particular issue in dealing with tables. Now it just says that <table> tags "can be displayed on Kindle devices and applications." Period. Not sure if this means the ancient K1 can now handle simple tables, or that they no longer care. The only person I knew who had one got rid of it years ago.

3.8.3 Styling Guideline #3: Design for a Good eBook Experience

A new section that adds a lengthy paragraph that is essentially useless, and includes this utterly unenforceable guideline:
Using fixed-layout format just to replicate print layout is not allowed in Kindle books because customers report this as a bad user experience.
As if they could know what your intention was. Maybe you intended to replicate the print experience, results be damned. Or maybe you just suck at formatting. Hard to say. But I guess Amazon intends to try. Does this mean they plan to start pulling poorly formatting fixed layouts? One can only hope. On that note, however, see the next section.

4.1 Metadata Fields Supporting Fixed-Layout Books

Okay, so now we come to the crux of the matter. Amazon has made a crucial change to the Region Magnification metadata description which, oddly, not only does not remove it (even though it's non-functional, and thus, irrelevant), but essentially makes it mandatory. Not the element itself, mind you, but Region Magnification itself. Here is how it now describes this element:
An optional tag for enabling the Kindle Panel View and Kindle Text Pop-Up features that are required for comics and children's books (italics mine).
I employ the italics to point out the seeming discrepancy in the statement. In essence, the tag itself is option only because it doesn't work; i.e. it makes no difference if you add it or what value to assign to it. As I have discussed both in my Kindle tutorial series and my formatting manual, KindleGen will add the RegionMagnification entity itself (and the correct value of true) if Region Mag is, in fact, present in the publication, regardless of whether you put it in the metadata or not, and will remove it otherwise, even if you enter it with a value of false. Thus, there is no point in adding it at all. The tag is truly optional.

However, according to the latter half of this statement - which is the portion added in this update - Panel View and Text Pop-Ups are now required for Kindle comics and children's books (respectively, one must presume, since Panel View is only available in comics). This is not a critical distinction for comics, however, since the Virtual Panels feature is present by default where custom panels are not provided by the content creator. Still, this statement technically makes is a requirement to include your own "Region Magnification" elements, a term the Guidelines uses interchangeably for Panel View (i.e. in the heading for section 5.4).

The real dilemma here, though, is with regard to children's books. Many Kindle ebooks of both types consist of full page images with the text included in the image. This is a poor design decision, as I have often stated, due to the loss of many crucial features, but one that is unavoidable for many ebook creators due to the cost and complexity of adding pop-up elements. According to this new addition, that is no longer an option.

Now, granted, this has technically been the case all along. Section 4.2.2 which follows this has always listed as "Requirement #2" the use of Region Magnification in children's books, rather than adding it to the later "Recommendations" section. But it has not been strongly enforced.

Does the addition of this clause imply that Amazon intends to begin enforcing this rule more diligently, and start refusing children's books with no Text Pop-Ups? Or is it just a strongly suggestive guideline that has been added in order to be more consistent with the section that follows?

Curiously, where this metadata element chart is repeated in Section 5.1 for comics the change has not been carried over. Editorial oversight or intentional omission?

8.1 Media Query Guidelines (and subsections 8.1.1—8.1.5)

Incidentally, I did not do a post when the 2015.1 version of the Guidelinse was released, as it included only this one change, although it was quite useful. This new "Media Query Guidelines" section was added which is fairly extensive and contains a lot of very helpful notes for handling backwards compatibility. This is, of course, primarily of use for reflowing ebooks, and so not as much of interest for my purposes, as fixed layouts are another beast entirely.

Monday, June 15, 2015

Amazon Changes Kindle Unlimited Payout Method

Amazon has just announced a critically important change to the way authors are paid for titles borrowed through both the Kindle Unlimited and Kindle Online Lending Library programs. The former is Amazon's successful $9.99/month ebook subscription service, while the latter is the one-book-per-month benefit Prime subscription members receive, but both reward the authors of the borrowed books with royalties paid on a monthly basis.

In either case, the authors of the borrowed titles have previously been paid a percentage of a "global" fund each month, divided equally between the total number of titles borrowed, regardless of their length, or whether those books were ever read. This is a crucially important distinction, as it has given and enormous financial advantage to authors of shorter works, since the payout has always been the same for, say, a 20-page children's ebook as it is for a 600-page historical novel: they each count as one borrow, and thus impart the same payment.

This, along with the general downward push of ebook pricing by consumers who feel ebooks should be cheap due to their lack of physical substance (as if the content does not matter), has lent great emphasis to the proliferation of shorter works which can be produced (and consumed) much faster. The continuing trend of children's ebooks to top the sales charts in percent of growth year-on-year bears witness to this trend. And while I am not an advocate of exclusivity in general, there has been very little incentive for authors of longer works to enroll their titles in Kindle Select, Amazon's exclusive distribution program that is required for all titles to be included in the lending programs.

However, today Amazon announced that beginning on July 1st, they will radically change the way that payouts are distributed to authors of borrowed content. Rather than dividing the funds on a per borrow basis, the payments will now be allotted on a per page read calculation. That is, the total global fund for each month will be divided by the total number of pages read of each author's work during that period, and doled out accordingly.

This has significant implications, both with regard to the benefit of enrolling in the lending programs, as well as to the creation of new content. It will now be vastly more advantageous to add pages to your work (as far as Amazon's program is concerned, at any rate), rather than releasing shorter pieces in order to get more titles borrowed. And while we all would like to think that book content is produced solely by authors who care only about the work itself and not the monetary benefits, one truth I've learned during my years of successes and failures as an author is that writing books is a business, and the authors who are most successful approach it that way, producing content calculated to bring the greatest return on their investment of time and energy. The Kindle lending programs have become a pivotal part of that return for many, dispensing tens of millions of dollars to authors every year - many of whom have been utterly ignored by the traditional publishing machine.

The key factor in this change is that a single ebook that is borrowed now returns an investment more in line with the amount of work it took to create it. A short novella of 100 pages will be paid a rate one third of that received by the author of a full length 300-page novel, and 1/6th the royalty allotted to the author of a 600 page book that likely took an equally longer time to write - assuming, of course, the whole book is read from start to finish.

And that is the other crucial component of this change.

Just borrowing and sampling a book is now no longer enough to trigger a full payment to the author of that work; instead, Amazon's "Big Brother" ability to monitor the status of all content on the Whispersync web (created ostensibly to allow a user's account to sync a title across all of their devices and reading apps) gives them the data required to make this change, since they know exactly how many pages have been read in every ebook ever bought or borrowed from them (assuming it has not been hacked and side-loaded by the reader, which at best is a small percentage of the total number). This is one case in which this admittedly creepy surveillance capacity turns out to work in the content creator's favor. If a reader borrows five books, say, but only reads 10% of each (the minimum required to trigger the payout previously), while another reader borrows just one book but reads the whole thing, the authors of those works will receive due compensation equal to what the readers actually consumed.

And as far as I can see, that is as it should be.

One factor that I will be highly interested to see as a result of this is how it affects the per-unit payout. That is, given an equal number of total ebook borrows month to month, with at least some of these now receiving less compensation due not being read completely, logically speaking the remaining titles that are read clear through will receive a higher payout overall than they would have otherwise, since it is based on a total pool of funds that is established by Amazon each month (and drawn at least in part from subscription fees for Prime and Unlimited membership).

This amount has always varied month-to-month, due to seasonal trends in reading, but has ranged anywhere between $1-10 million, give or take, generally averaging something like $2-3 million (this month's being $10.8 million due to a huge $7.8 million bonus). With the establishment of Kindle Unlimited in July of 2014, the per-unit payouts have rapidly declined - dropping from $2.20 per borrow in June of last year (just prior to the KU rollout) to a low of $1.33 in October - due to the increasing number of borrows among which the pool must be shared. My guess is that this is Amazon's attempt to rectify that problem, as well as the dilemma presented by the steadily increasing number of subscription members overall.

The ultimate question is: Will Amazon increase or decrease the monthly payout based on these factors? And more importantly, will authors feel they are being compensated more fairly as a result, or less, and how will they respond?