[Tex2002ans]

Originally Posted by xristy

I suppose it's nice that some people in India get some work out of laboriously generating these ePub and mobi files.

Sure, sure, but I see crap like the InDesign -> PDF -> OCR -> EPUB as a TON of waste manpower. Makes me want to pull my hair out! (As a programmer, I like to remove cruft).

Originally Posted by xristy

My concern is that the market is being flooded with garbage.

Part of the reason why I hopped in on all these conversions. I was sick and tired of the typos/crappy spaghetti code.

And the problem will only be worse when the next book formats come out, a lot of the previous spaghetti code will just be ASKING for trouble (and there will be so many people paying to have their ebooks REDONE).

Like that image problem, the more that the Kindle HDX and higher resolution devices spread, the more people will start complaining about tiny thumbnail images.

Originally Posted by xristy

It's really a shame that PDF as a format takes such a bad rap for eBooks when in fact it is actually far superior for technical materials (I realize that ePub / mobi are great for mass market fiction and lots of non-fiction which I'm sure is like 95% of the sales.) PDFs work great on an iPad with, for example, Goodreader and I expect there are excellent PDF readers on Android tablets as well. There's certainly no problem with PDF access on a PC.

Toxaris hits the nail on the head. PDF is built for PRINT. It was designed for a certain page size, and it was built to be a completely fixed format and LOOK THE SAME NO MATTER THE DEVICE. This means that the device you want to read on "MUST BE THIS LARGE". So a PDF designed for a book size, trying to be read on a smaller device (like an ereader or a phone), get a horrible experience.

Since your medium is FIXED (you know the page size, you know the margins, you know the font-size, you know the fonts being used, ...), you can get away with doing more complex typography. But once you start changing page sizes/changing things around, you will have to redo a lot of those typographical tweaks.

PDFs are also horrible for vision impaired users (Large Print Edition of books, IF the company decides to make them (which most don't)), and pretty poor for readers who are blind.

eBooks are a lot like HTML, it works on any size screen, reflows, any font-size, any font, any colors, any margins, etc. etc. It is about how the READER is most comfortable reading.

There is the whole "fixed-format EPUB" thing (which will get you closer to what you want, like with PDF), but they are a GIANT pain in the butt to make, and are really just geared towards ONE DEVICE only. So have fun moving that Kindle Fire fixed-format book to another Kindle. (Also, fixed-format EPUBs/MOBIs are very hard or impossible to get in the stores.).

Originally Posted by xristy

I only use PDFs which limits availability since there is a mind set to dump mobi / ePubs that are poorly prepared and that apparently really can't use appropriate technologies such as SVG or MathML.

Indeed indeed... you need to pay for quality conversion, and many of these publishers/authors don't. They go with the cheapest.

And avoiding SVG/MathML for now is probably a good idea.... you would be cutting the potential readership down to a few % of the readers.

Good for future-proofing though (although knowing these damn math-books, they just come out with worthless new editions every year).

Originally Posted by xristy

Creating good quality OCR'd PDFs via scanning of books for which source is not available is tractable for me with Acrobat X so I assume there are better tools available to commercial outfits.

I posted an outline of my PDF -> OCR -> EPUB method a while back:

https://www.mobileread.com/forums/sho...9&postcount=10

I use Finereader (one of the most accurate OCR programs), and this is pretty much what you have to do.

It requires just a lot of human labor to fix it up. For example, you can go take a look at archive.org and see how books look "just through OCR" (I laugh when people say "hey look, there is an EPUB of this book already on archive.org!").

Toxaris has his Word Tools which also might help speed up a lot of the book conversion (if you use Microsoft Word). But pretty much all of us just have to OCR -> many hours of human eyes/formatting -> high quality output.

As to the conversion market:

• You have volunteers doing this stuff on MobileRead/Project Gutenberg/elsewhere
  • Mostly just tackling Public Domain works
• You get people who's job it is to convert (like me)
  • In my case, mostly Public Domain, but even all of our new books are CC3.0 (as close to Public Domain as possible)
  • Luckily, I was able to convince them with reason/quality... A/B crap/amazing comparisons are fantastic for this.
• You have companies who get paid to convert quality
  • Like Hitch's company, who is a VERY minor portion of the conversion market
• Companies who get paid to convert crap
  • A much larger portion of the conversion market
• Large/Small publishers
  • (Some of which who go through the crap converters, or have in-house converters)
  • In many of these cases though, they just have their typographer export directly from InDesign, and of course, this output isn't the greatest
    • They probably took an InDesign course on "how to export to 'EPUB' (iBooks)!!!" So everything that is non-iBooks gets garbage.
    • A lot of these old-timey typographers are probably so used to the "print world" too, and not used to HTML/CSS/reflowability.
• Then you have the flood of self-published stuff
  • (Run through Calibre, SmashWords, or some other automatic conversion), which probably dwarfs all the rest combined.

Originally Posted by xristy

Is there no way to promote the availability of PDFs for technical materials? Aren't PDFs routinely created as part of the workflow for physical printing?

Nowhere to really sell the PDFs though besides your own site. I believe PDFs for technical documents can be sold through something like Nook Study:

http://www.barnesandnoble.com/nookstudy/index.asp

But that is probably only open to huge publishers.

So for now, you get physical print + EPUB (every other book store) + MOBI (Amazon). Everything else is a free-for-all on individual sites.

Originally Posted by xristy

Are the sales of poorly formatted mobi / ePub technical books really sufficient to warrant the effort expended in producing them?

Yes, the sales are massive (and if they pay those Indian conversion companies, hmmm, lets say it can be anywhere from $.50-$5 per page). This is WAY less than what they are paying a typographer/editor/everyone else to actually typeset the book.

As Hitch said, the converters get PENNIES compared to the total cost of producing the book.

So they have sunk in tens of thousands of dollars in a book, they can sink in a few hundred dollars for some crappy conversion, and get a HUGE boost in sales for the Kindle Editions.

Originally Posted by xristy

Wouldn't it be cheaper by far to forego producing mobi / ePub versions and just offer much lower cost PDFs for sale?

People want to read these books on their multitude of different devices, and not be limited to PC/large tablets.

I assume most of these larger publishers also do sell the digital PDFs on their own stores (for example, all of these complex Math/Engineering books probably do sell PDFs right on the publisher's site).

But again, this is talking about the market for NEW books.

This is completely ignoring the entire hideous market of Scanned+Reprinted books. Those are another horrible beast altogether!

And scanned/reprinted MATH BOOKS... the horrors!