by wjw on May 19, 2015

So I have an ebook promotion coming up, and I’ve been uploading new versions of some of my ebooks to the various sites.  The new versions have clickable links, which means that if you like the book, you can immediately click on another book by me and and download it.

(And that is certainly an advantage etexts have over traditional books, for sure.)

Anyway, I’ve had no problem uploading up the revised files, except on Kobo.  Because Kobo refused a couple of the uploads, telling me “We can’t upload your book,” without saying why.  Which is no damn help at all.

After wasting an hour or so trying brute-force tactics mixed with fiddling with the files to try to guess what was objectionable, I realized that I probably wasn’t the first person to have this problem, and I decided to use magic and invoked Google.  Which gave me the solution.

Now Kobo sells ebooks exclusively in epub format, a format used by all ebook providers but Amazon, which employs mobi.  Kobo wouldn’t accept my epub files for reasons best known to itself, but it will also translate to epub from files in other formats.

Ah.  You see where there is going, yes?

So I used Calibre to translate my epub file to mobi, which is Amazon’s format, and then I uploaded the mobi file to Kobo, which doesn’t sell mobi files.  But Kobo converted my mobi file back to epub, which it then allowed me to place on sale.

Because artificial intelligence, right?  Right?

I feel a little less threatened by the Singularity now.


Jerry May 20, 2015 at 4:33 pm


John Appel May 20, 2015 at 9:50 pm

I’m an information security guy. I’ve spent longer doing infosec than many people’s entire IT careers have lasted. And one reason I’m convinced I’ll always have a job is that software developers are really, really bad at theirs.

Ralf T. Dog May 20, 2015 at 9:51 pm

The problem is, content can shift in translation. I hope all your formatting survived.

wjw May 20, 2015 at 10:40 pm

Errors will inevitably creep in with translation, but I can’t fix it, it’s =their= fault, so I try not to worry about it. I spot-checked the file and it looked okay.

I wish people would stop using Adobe Digital to read ebooks, because it can read files in a very strange way. I can create a file that will read just fine on Calibre, and Sigil, and on an actual ebook reader, but on Adobe will turn into a hot mess, with no way to fix it. And Adobe’s the software that many of the e-publishers will use to spot-check the file. Crap.

James R. Strickland May 22, 2015 at 9:13 am

Calibre has these kinds of problems. It also can’t produce a mobi that Amazon will accept. I use Jutoh these days, usually to clean up an ebook generated in Scrivener.

For all that epub and mobi are standards, it’s damned hard to find software that will create according to the standards, and even harder (for the Windows-impaired) to find ebook readers that actually respect the trouble you put in to making your book look nice.

Or maybe, after some months of getting used to LaTeX for typesetting hardcopy books, my control-freaky tendencies are rearing their ugly heads again. The zen of ebook seems to be: Deliver the text and accept that you do not control the format.

The format is controlled by chimps.


wjw May 23, 2015 at 2:06 pm

Indeed. Though I’ve never had the problem with Calibre that you describe— Amazon will accept mine, and they look okay for the most part. I use Atlantis to create the epub file, and I’ve never had a problem with Atlantis converting one of those.

It’s a measure of how sad the industry is that after a decade there’s still no mobi editor, and no epub editor that we can trust, and no standard way to read epub. It’s just HTML when all is said and done, so what’s the problem?

TRX May 24, 2015 at 6:23 pm

As a reader, I have the opposite problem – while there are some epub and mobi readers available for my ancient tablet, none of them actually work. The PDF viewer will display a file, but only in a microscopic font with pagination that can’t be turned off. The “web browser” fails on 99% of the real web, but it will show a straight HTML file, so I use Calibre to convert “reader” files to HTML and use those. Most of them still take some hand editing or they’ll render as gibberish, though.

Comments on this entry are closed.

Previous post:

Next post:

Contact Us | Terms of User | Trademarks | Privacy Statement

Copyright © 2010 WJW. All Rights Reserved.