Why are there so many updates on newly released eBooks?

The short version is, because the conversion process from Open Document Format to EPUB3 is imperfect and takes a lot of manual work. The conversion tries to change all double spaces to single, which means I have to manually change them back.

Yes, I am aware that style guides say that only a single space between sentences is correct. I maintain that the style guides are wrong. Consider the following fragment:

“[...] were you there anyway?” John said, [...]

Now, tell me: Is that all part of one continuing sentence, or is it the end of one and the beginning of the next?

With only a single space, you can’t tell, can you?

It’s not just that a double space after a period looks better on the page and is easier to read. It’s also that a single space is AMBIGUOUS. Replacing double spaces with single THROWS AWAY METADATA that can be important to meaning.

I’m a writer. Details matter.

I auto-re-insert as many as I can. However, there are many possible patterns, and ultimately it’s not feasible to rely solely on search and replace, it REQUIRES manual reading and fixing. And re-reading, and re-reading. Sometimes I have to go through the book two or three times before I’m confident I’ve caught all of them. Some errors escape me for more than a year. I recently spotted — and corrected — an error in A Line In The Stars, released in October 2024, in a line that said ‘destroyers’ where it sould have said ‘cruisers’.

And that is why there are updates. Every time a book updates after publication, it means that I have found and corrected at least one error.

Because you should have an error-free book, and you shouldn’t have to wait a year to get the book.

A quick note to say that I have just seen the almost final draft of the cover art for my upcoming nobel In Flux. This cover artwork is by C.J Evelyn, and it is superb. I'm looking forward to being able to show off the finished version soon.

For reasons (one assumes) of readability (and not, one hopes, simple control-freakishness), Amazon Kindle Desktop Publishing overrides any fonts specified in an eBook, instead displaying all books in whatever preferred reading font face the Kindle user has chosen.

Now there are sound reasons to do that. I get it. Really I do. Some people pick ghastly text fonts. I have a particular hatred for Times New Roman, in my opinion one of the most horrible text faces ever designed, with its primary intention that of remaining more or less readable while jamming as many characters as possible onto a fixed-size page. I have my Kindle reader app set to one of the cleanest available (within Kindle, that is) sans-serif fonts. (I detest Roman serifs. They're ugly.)

But come on, KDP. There are times when a writer wants to set off a small amount of text in a different font for a specific creative effect. Godthief has a preface containing the text of the Prophecy of Tendarrion, and a brief excerpt from a discussion of prophecy by Sage Ryanon of Krent, both in appropriate font faces.

Kindle smashes them both flat.

The fourth chapter of my current work-in-progress contains the text of a brief handwritten letter. I went to a great deal of trouble to find and select a freely-licensed, clean, readable script font that looks like an elegant copperplate hand. I went to the trouble of finding out how to properly embed a font into an EPUB3 eBook, including figuring out a widespread documentation error. (See my previous post.)

To no avail. Kindle smashes them flat.

SUPPOSEDLY, Kindle honors and preserves embedded custom fonts in headings.

No, it doesn't. Not unless there is some undocumented secret to what Kindle will acknowledge as a heading. Certainly using an XHTML heading tag doesn't work.

Readability is important. But so is the artist's, or writer's, intention. Nobody would say that it was acceptable for Kindle to smash all book covers down to a plain white page with the title in black.

And it's not acceptable for Kindle to smash custom fonts used in strictly limited places for specific effect, either.

It's an old béte noire of Kindle publishing that Kindle makes it difficult to specify your own fonts. There are arguably sound reasons for this; it allows readers to make the decision of what (Kindle supported) font they find most readable for books.

Nevertheless, sometimes you want to override a font or force a specific font to produce a specific visual effect.

This is actually possible. I am told that even if done correctly, Kindle will sometimes still strip out custom fonts. But it will mostly work, as long as you do it right.

First, choose the right font type. OpenType (fontname.otf) is best supported, followed by TrueType (fontname.ttf). You probably should not even try to use Adobe Type 1 fonts. And make sure that either you choose a freely licensed font, or you have a valid license to use it.

Second, your font needs to be embedded in your EPUB3 file. Try putting it under OEBPS/Fonts, and don't forget to add it to the manifest (content.opf):

<item id="oldlondon_otf" href="Fonts/oldlondon.otf" media-type="font/otf"/>

And here lies the crucial “gotcha”. I will not say all of the documentation, but at least, all of the reference CSS documentation that I have found and read says that the proper media-type to declare for an OpenType font is application/opentype.

This is outright WRONG. The correct media-type, the one that ACTUALLY WORKS, is font/ttf.

Third, your CSS stylesheet must contain a correctly formatted @font-face record, like this one:

@font-face {
   font-family: "OldLondon";
   src: url(../Fonts/oldlondon.otf);
}

And then finally, when you use the font, you must reference that font-family:

.prophecy {
   font-family: "OldLondon", cursive;
   font-style: italic;
}

And that SHOULD work.

Victory! After being served with a DMCA takedown demand by the intellectuAl property lawyer I retained, TikTok has finally decided they should pay attention to my impersonation-account complaint, found the offending account to be in violation of their impersonation policy, and taken it down.

My thanks go out to both said attorney, and to Coach Brent, the TikTok user who reported the impersonator to me.

A word of thanks to Scottish reader Iain Muir, who entirely correctly pointed out that I incorrectly described the late, great Iain M. Banks in Bearing Gifts as an English SF writer. He is of course a Scot. There is a difference, and details matter.

(Yes, I said is, not was. I maintain that Iain M. Banks did not die, he SUBLIMED. If you don't know what I mean by that, then you should read his books — in particular, The Hydrogen Sonata.)

I have located and corrected the error, and the correction is going live at this very moment, both for the eBook and the print edition.

Lessons learned today:

There are occasionally reasons for saving an ODT document in DOCX. Principally if you need to send it to someone who is unable to, or for other reasons does not, use ODT, or when using pandoc to export to EPUB. (As previously mentioned, it turns out that LibreOffice's EPUB export is horribly broken, because it is built on top of a piece of alpha-quality abandonware called libepubgen. It also transpires that pandoc's ODT reader plugin is also badly broken. The problem can be worked around by using LibreOffice to save the document as DOCX, then using pandoc to convert from DOCX to EPUB, and then running the EPUB through Sigil or Calibre for fix-up.)

However, if you then accidentally try to open that DOCX in LibreOffice for further work, edit it, and try to save back to ODT, the styles in the document will end up HORRIBLY buggered in bizarrely incomprehensible ways, and you'll be going crazy trying to figure out what went wrong.

So maybe don't do that.

Authors are often advised not to read reviews of their own books. While I’m aware of this, I still like to keep a finger on the pulse of my readership, and reviews and reader comments are really the only way I have to do that. (Except for those of you who have taken up the invitation in my books to come and join the discussion.)

This morning I was checking the ratings on some of my books, and happened to notice a review on A Line In The Stars that complained about the fact that the Stardock Trilogy contains commentary embedded in its story that has bearing upon current events and sociocultural issues.

Well, yes. Of course it does. You don’t actually understand WHY we write, do you?

Look, any fool can write trite, vapid fluff entertainment that you can read or watch with your higher critical faculties switched off and not miss anything. Just a glance at probably about half the total volume of TV and movies ever generated by Hollywood should show you that. But that’s not what drives a lot of us. We don’t want to generate more empty fluff. The world has more than enough of that. Present-day mass entertainment, on the whole, is bread and circuses writ large.

The late Iain M. Banks, for whom I hold an incredible degree of respect, once said that science fiction is the only genre of fiction that actually matters, because it is the only genre of fiction that tries to predict and anticipate the potential problems in our future and devise solutions for them, or approaches to solving them, before we actually encounter them. Elizabeth Bear, another writer for whom I have immense respect, has sociocultural issues deeply embedded throughout her magnificent White Space trilogy, centered in large part upon what separates civilization from the howling barbarians outside. Because these issues need to be talked about. They need to be understood.

This is why we write. Not just to entertain you, our readers — but to make you think, or at least to give you things to think about. That is the true purpose of literature. To make you think.

If we do not make you think, or at the very least give you important things to think about, we have failed in our responsibilities.

If we put those important issues in front of you and you refuse to think about them, you have failed in yours.

No, not the business-oriented chat platform. Slack more as in J.R. “Bob” Dobbs.

Since I got Godthief completed and published, I’ve been feeling anxious because I don't have an immediate plan for what to write next. I have a number of story starts sitting in my slushpile, several of which I quite like, including one based on a scenario which I personally consider both quite alarming and a very real threat, but none of those starts has an actual underpinning story attached to them that I’ve figured out yet. And so far I’ve been unable to come up with a follow-on story in my Stardock universe that I consider good enough or complete enough to tell.

A couple of nights ago, I finished reading Folded Sky, the concluding volume of Elizabeth Bear’s magnificent White Space trilogy. I am in awe of Elizabeth Bear; she is a magnificent writer, and one who says a lot of things about society that I'd like to say in my own writing but says them better, and White Space is in many ways reminiscent of the late, great, and dearly missed Iain M. Banks. And of course I went on and re-read her updated author bio, and realized that Elizabeth — of whom I will repeat, I am in awe — has written fifteen books in twenty years.

Then I realized that I have published seven books in the past 18 months, written over the course of about three years total.

You know... maybe I could afford to cut myself some slack...

Godthief is copy-complete and in the hands of my beta readers, going through final proofreading right now. It should be released in eBook format either at the very end of May, or the very beginning of June — most likely June, allowing for time for Amazon to process it and make it available.

This is the first book I have written using pandoc as part of my writing workflow. Using pandoc forces one additional intermediate step of saving a copy of the manuscript in .DOCX format, for the time being, because pandoc's .ODT reader module is… severely buggy. But pandoc — although not perfect — does an IMMENSELY better job of generating usable, editable XHTML than the libepubgen that LibreOffice/OpenOffice uses.

Using LibreOffice's EPUB export, the workflow basically becomes: Export as EPUB, fix up the EPUB in Calibre and/or Sigil, hand-update it for EPUB 3.3 compliance, strip out every last bit of the ghastly formatting, throw away the CSS, then go through the entire book line by line manually re-applying the formatting correctly and hand-writing new CSS.

Using the ODT-to-DOCX-to-pandoc method, I still have to write new CSS, and I still have to massage some formatting by hand, and I still have to go through the manuscript manually re-adding extra blank lines and HR dividers. But all of the text formatting survives, in standard XHTML tags, instead of an almost-illegible sea of <span></span><span></span><span></span><span></span>. (LibreOffice's EPUB export can easily generate XHTML files half of whose total file size is mostly-redundant <span> tags for opaque span classes.)