Typos in ebooks

[Logseman]

OCR= Optical Character Recognition. I.E we're talking of how they convert printed text into an electronic document. In an OCR process errors are deemed to occur, what people are angry about is that ebook versions are being released without contempt of what the final result is.

OCD= Obsessive Compulsive Disorder. rakulos was just stating s/he can't bear reading badly formatted books and, due to to that syndrome, can't get any enjoyment of the book until s/he corrects them.

[JSWolf]

What we could do is start a thread that lists the books and what errors there are along with the corrections.

[Tonycole]

Originally Posted by AlexBell

I too have been angry and frustrated by the shoddy formatting found in many commercial ebooks. I had a frank exchange of views a week or so ago with HarperCollins Australia about 'Flashman' by George MacDonald Fraser. I won't go into the details now, but I told them that I would be ashamed to put out something as poorly produced over my signature, and that I certainly would not have bought the book if I had known how badly produced it was. I did get a response, but only that the book was produced by HarperCollins UK, and they would send the email to them.

I have a suggestion, but it depends on whether commercial publishers read MobileRead Forums, and whether they give a damn anyway.

Suppose each time any of us finds a particularly badly produced book we notify each other through this forum to warn others not to buy that book - something like a 'Hall of Shame' - and copy the details to the publisher, and mention that the comment has been posted here.

We might even need a special subforum for the Hall of Shame. How do people feel about the idea? Would the moderators permit this?

Regards, Alex

Hi, I think your idea of a sort of "hall of shame" is a very good one. I would be very happy to have a sort of continually growing post on my eReader blog with this, as like so many others, I have been palmed off with full price books from apparently reputable publishers, that were a mass of typos.

If anyone feels like helping with .this,, please send me your "Black list" (name of book and Publisher) at tony@ebookanoid.com, and I shall start the post. I shall also hand on all such emails to a friend who runs a blog for Apple owners, and who is equally angry about this matter, and will certainly join in such a campaign

[sassanik]

Originally Posted by Michael J Hunt

I'm trying to understand all the ins and outs of this discussion, but the use of initials confuses me - perhaps it's obvious to everyone else here, but, what do OCR and OCD mean? Thanks in advance.

MJ

OCR refers to the process of scanning a paper book, or pages of a book, into a computer. I believe that OCR stands for Optical Character Recognition?

OCD= obsessive compulsive disorder



I have noted that a lot of times books have trouble with quotation marks. I find that they are missing very frequently when reading resulting in me reading the line a couple of times to figure out where the character is stopping talking. I am guessing some of this is caused by reflow?

I also have just had some where things are spelled wrong, and have funky punctuation marks from major publishers as well as some of the smaller ones.

Personally I find that I catch different errors in my writing when I proofread a hardcopy instead of a digital one. Which could account for why some of the errors are getting through the proofreading stage.

Amy

[Logseman]

The Hall of Shame has been created, in case you'd like to start bashing the books which deserve it ^^

https://www.mobileread.com/forums/sho...d.php?p=860710

[Sunspark]

What a publisher needs to do, is after putting out a book, advertise that they will pay $1 for every error/typo/problem spotted by a reader who reports it to them. And the corrected versions to be free downloads for people who have paid for the crappy versions.

If just 1 publisher does this, the others will be forced to follow suit because then they will have the rep as the publishers that are no good.

[Goshzilla]

Originally Posted by raac

Hi,

I don't know if this has already been discussed, but aren't others also frustrated at the numerous typos in retail e-books? I have seen numerous mistakes in books from both Sony and Amazon. This doesn't happen with paper books and so I feel somewhat ripped off when it happens with ebooks.

How do others feel? Any tips on getting the publishers or sellers to take notice? Maybe it's time to mount a mass complaint?

These typos happen all the time with paper books. The math and sciences are filled with them. I'm just saying this isn't unique to electronic books.

[WillAdams]

Well, Dr. Donald Knuth used to issue physical checks for errors in his books and programs (I got one for an error and an improvement in his _Digital Typography_) --- now he does virtual monies in the ``The Bank of San Serriffe'':

http://www-cs-faculty.stanford.edu/~uno/boss.html

It would be great if other authors or publishing houses would adopt similar practices.

William

Errors are not limited to e-books. I recall way back in college a Professor required use of a text book he had written for a course in Catalytic Chemistry; first edition published by Wiley. So the first day of class he hands out, I kid you not, 30 pages typed of errata. Some minor stuff yes, but also chemical equations that did not balance and significant missing content; as in “as illustrated in example 5.1 above.” No, not even there. I was forewarned by students who had taken the class in previous years about this so it was not like it had just been discovered in a newly published work. I am certain that the publisher was waiting to sell out the first run before republishing a corrected edition.

Anyway to get back to the subject. If the original book was only available in paper form and scanning with OCR was used to get the electronic version errors are bound to happen. I have been a volunteer proof reader for Gutenberg Project and I know how much proofing and review goes on there before books are released as final. Even so errors can still be found in texts downloaded from there. I don't know how much proofing Google Books does for the books they have scanned, the public domain ones free for complete download, but these tend to have a lot of errors.

[alecE]

FWIW, a quick summary from the book I am reading at the moment - Meta Maths by Gregory Chaitin, published by Atlantic Books.

I'll draw a veil over the difficulty the epub format has with mathematical equations. However I've encountered about ten instances of words which are hard hyphenated so that the hyphen appears in the middle of a word, in the middle of a line. The typo which amused me most though was the failure to distinguish between minus two and root two, so that -2 was displayed instead of √2 (just in case the symbol does not display properly here it represents the square root symbol) - just a little bit critical :-)

Needless to say there are also a couple of ordinary misspellings as well.

In my own not very humble opinion, I cannot see any reason whatsoever why such shoddy work should be permitted. I will be notifying the publisher, whom I am sure will embrace me with tears of gratitude for showing him/her/it these errors.

Now this is a title that, because of its maths content, demands closer scrutiny than a 'normal' text. Thankfully I've only found one instance of a maths error, but this surely implies that the production/proofing processes are sub-standard?

[Ankh]

Originally Posted by HarryT

That sounds like an issue with "soft hyphens". Unfortunately it's all too common.

And easy to fix (if the book is DRM-free). Use Calibre to convert the book to, say, Microsoft ".lit", delete original epub, then convert it from lit to epub.

[calvin-c]

Originally Posted by raac

Hi,

I don't know if this has already been discussed, but aren't others also frustrated at the numerous typos in retail e-books? I have seen numerous mistakes in books from both Sony and Amazon. This doesn't happen with paper books and so I feel somewhat ripped off when it happens with ebooks.

How do others feel? Any tips on getting the publishers or sellers to take notice? Maybe it's time to mount a mass complaint?

Strange-I see quite a few typos in paper books. Probably not as many as I see in ebooks, but I do see quite a few. Maybe it's the books I read. (I don't keep track so this is more an impression than a statement of fact. Maybe I just notice them more in ebooks? I don't think so, but I guess it's possible.)

[WillAdams]

The problem is, ebooks are normally derived from the files which were used to create the paper book which often have formatting applied to them which is specific to the physical books text block size, so things which were correct (manually inserted hyphens, forced line breaks) become incorrect (hyphen in the middle of a word, spurious paragraph break), in addition to glitches from poor conversion methodologies.

The answer of course is to tag the book in a rich, unambiguous markup format up-front such as TEI, then convert to other formats (and when errors are found, always correcting the source), but this requires an extra step which is hard to justify financially.

William

[WarnerYoung]

Originally Posted by Stitchawl

I'm beginning to feel as if I'm the only one who doesn't mind the typos!
I wonder if there is something wrong with me?!?

When I read a book on my reader, it's for the pleasure of reading a story, not to correct student's term papers. If there is a mistake I just read past it and pay it no mind.

Stitchawl

It depends on the ebook, too, of course. If the errors are minor (e.g. a few missing periods, maybe an extra space somewhere), you're not as likely to notice them.

When the errors are literally legion, it's hard not to notice. I recently compiled the list of ebook errors (obviously OCR typos) in By the Sword by Mercedes Lackey, I ended up with a list of 250 typos. And some of the items in my list were "search for this word, it's wrong everywhere".

I sent the list into the publisher, who were actually fairly nice. They replied and said they would look into it. Of course, now that the ebook has been removed from ereader.com, who knows if a corrected version will ever appear again?

[Worldwalker]

Originally Posted by Strether

Having heard about Steven Saylor's mysteries set in ancient Rome, I recently bought the first in the series, Roman Blood, from Amazon. Alas, it turned out to be a Topaz-formatted book. Very ugly and with slow-turning pages, but obviously had been scanned and the errors were legion. Especially hyphenated words that had come at the end of a line in the printed text, but were inappropriate if they came anywhere else, which they did in the ebook. I could only read it for 3 or 4 chapters, then stopped and asked Customer Service to refund my money, which they obligingly did.

Normally, I'd be on the lookout for Topaz books, but this one snuck in on me. Looking at the rest of the series, I see they're all formatted the same way. I wonder if the author knows/cares that they're so poorly produced?

Steven Saylor seems to be happy to exchange email with his fans. He's the one who told me about the UK mass market edition of "The Triumph of Caesar", which is the only way to buy it in anything close to US MM format now that his US publisher is going for maximum revenue per unit and releasing the books only as trade paperbacks. I doubt if he'll know what a botch they've made of his books unless someone tells him. His contact info is on his website. Also, he's a really cool guy.

As for the comparison between free public-domain ebooks (Project Gutenberg, etc.) and commercially produced ebooks: I only buy ebooks from Baen, and I haven't found significant problems, but that's probably because Baen, being a SF specialty house founded by one forward-looking man, has probably had good electronic copies of those books from the get-go, and no need to scan/OCR them. From what I'm reading in this thread, some of what's turned out by the major publishing houses is barely spell-checked and not proofread at all. Project Gutenberg (and sites like Feedbooks and Manybooks, which scrape PG) have the benefit of the Distributed Proofreading project (have you proofread a page today?). Via PGDP, several separate human beings proofread every page of every scan, in several rounds of proofreading, and while some errors undoubtedly creep through (and there are more of them in books created prior to the formation of PGDP), by and large the quality is much higher than the "OCR it and ship it" books from the major publishers.

Y'know, if publishers were smart, they'd so something like that themselves: a system like PGDP, where you could register as a proofreader and get credit based on your accuracy in proofreading (as determined by a comparison of some significant number of people proofing each page) and number of pages you did, which would be applied towards the purchase price of that book once it is finished and released. There's probably some complicated problem involving minimum wage laws that prohibits it, but it still seems like a good idea. At least that way, ebooks would suck less.

Or, if they paid a bunch of college students (or even random people on the Web) ten bucks an hour to proofread OCR'd scans, they could proofread a 200-page book for under $500. It wouldn't take much of an increase in sales once word got around (hey, did you hear BigPubCo's ebooks are better than everyone else's?) to cover that.