That’s crazy - I was just about to say that I’ve been rereading some Barbara Hambly and I’m encountering a different weird problem - there will be a section break (you know, which would normally be a few blank lines in a text) only the last paragraph of the previous section is stuck onto the next section, which is very jarring.
I reread It recently via Kindle and was very distracted by errors like italics starting and not ending for a paragraph, for instance. In addition to the regular typos.
When I first started doing the E-reader thing I was using the Google Book app on my Droid. With that it was easy to see what was going on because I could just flip to the actual scanned page and read that. Usually that would make sense of the word, sometimes it would reveal a big smudge or a crease in the page that even I couldn’t figure out. But it was free and I moved on.
When I upgraded to a Kindle, I was surprised to see how many errors were still present and disappointed that I didn’t have access to the scanned page as well.
An idea that I thought of a while back: When you notice an error, I would like a way to highlight the word and tag it as an error. No explanation or possible correction, just that it’s wrong. If a certain amount of people tag the same word it would trigger someone at Amazon to look into it and fix it.
I suggest that there’s a threshold so that they don’t have someone looking into every single thing that every single person thinks is wrong. But if, say, 100 people tag the same word, there’s probably a problem with it. Also, if it’s wrong, they probably don’t need an explanation so I don’t see any use in them having to wade through all the reasons/explanations/fixes/frustrations etc that people would submit.
Also, doing this would give people an outlet for their annoyance rather then posting on the internet or in the Amazon reviews about all the errors in the books.
Amazon should offer a bounty on errors in their ebooks.
The first submission of a particular error would get, say, 10 cents. The second would get 9 cents, and so on. People would have a real (albeit small) incentive to submit errors that they might otherwise just ignore. Payment would occur when at least 10 people have submitted that error. This way, Amazon could easily filter out innocent false submissions. (What are the chances that 10 people would independently submit the same non-error as an error?)
At some point a human would have to confirm that the reported error is truly an error. But this is much easier than searching for errors to begin with.
Finally, the payment could be made in the form of “store credit” on Amazon; possibly only good towards ebooks. If you’ve got, say, $1.50 sitting around that can only be spent on ebooks, and you’re already the type of person who reads ebook, then it’s likely that Amazon will make an additional sale of an ebook to you. I think this would offset the cost of the bounty program.
Because – see my post #14 – they don’t understand the difference. They think the standard way to get text data into a computer is by scanning and don’t understand the resultant errors, not to mention the enormous processing time needed compared to formatting a simple text file.
I often have people write a story, print it out, and give me a printed copy instead of the file. When I ask for the file for publication purposes, they typically say, “Oh, just scan it in.” They think it’s the same thing.
This one was also a good laugh, except to the author.
Not all electronic format is created equal.
To create a printed book the original word processing file from the author is stripped and imported into a page layout program and the resulting PDFs get proofed and edited and mistakes corrected before the book is printed. Those PDFs are crappy for converting to ebooks, so instead the original word processing file is formatted into an ebook or the printed book is scanned.
The editing and proofing steps from the printing process have to happen all over again for the ebook, but they don’t seem to be as rigorous as for printed books, so you have a lot of errors. Or you have a scanned version with errors introduced by the scanning process that get missed when proofed (if the book is proofed at all). Pick your poison.
You know, they could do the job right the first time and use actual typesetting software.
I’d say that most books are available in print format, since most books were written before electronic media. I have a couple of collections of authors on my nook which were definitely written when even the most modern writers could only use typewriters…a collection by Agatha Christie and one by Andre Norton. Christie died in 1976, and Norton died 2005. Norton started getting published in the 1930s, when of course the books wouldn’t be in electronic format. Those old books have been scanned in, which is much, much cheaper than having someone sit down and type them in.
Confuses me, too. I know my book is available as an electronic file, because I still have it that way. I suspect they simp;ly “automatically” scan the books without even asking if an eloectronic version is available.
I’ve been reading e-books since before “e-books” was a thing. I used to load up books, in the form of .txt files, on my Palm V, way back in, well, let’s see…it was pre-2002, I know that much. The capacity of the Palm V was such that most books had to be broken up into two or three parts, each loaded on individually, one at a time.
Interestingly, I’m still seeing pretty much the same level of scanning errors now that I did then.
Not the NYT best-sellers that THIS freelancer copyedits for a major NY publishing house on a regular basis. Please put away the broad brush, folks.
Hmph. I am a copy editor myself–also freelance, as publishers will not cough up the dough to have in-house staff anymore. Years ago, you could have been copy-editing , and gotten benefits and a regular salary and out of it. Now, it’s all underpaid, overworked freelancers.
I always highly recommend reading the reviews before you buy any e-book. Due to this I’ve been able to mostly buy books sans errors, or very few errors. As others have said, I’ve found errors in print books, too. E-books seem to have a higher rate, but people are eager to tell you if the book you’re about to buy has this problem.
I have to say that I would truly be pissed if you didn’t buy my book simply because the e-book incarnation of it has typos.
Oh well. Can’t help it. I can live with one or two typos, but if I find out the book has a bunch of typos…nope, not gonna buy it.
I hate it, though, when the reviews tell you nothing about the edition - when you’re looking for something in the public domain that has a lot of different editions, say a Shakespeare play or a Bronte book, what you need to know is whether THAT edition has weird crap going on with it or how the notes are or whatever. I am not interested in your review of Hamlet!
[quote=“Musicat, post:14, topic:614082”]
I don’t think the general public, and that seems to include publishers, know the difference between original data in a text file and a scanned copy. While OCR is much improved now, I was constantly bombarded with “but can’t you just scan it?” years ago when OCRing was in its infancy. **The concept of a graphics vs. text file still isn’t a well-known technical difference.[/**QUOTE]
This is the key problem.
If the e-book is an editable text format, you can simply correct mistakes like you would on word, if it’s a gif or jpeg the you need to use image edting software for each case instead of “change all **Sinus **to Sirius”.
On a somewhat related note, I hate how Amazon lumps all reviews together. It happens with books, where they lump together the reviews for the hardcover and paperback editions. But at least with books, the content is generally the same.
But it’s really annoying with movies. They’ll lump together all of the reviews for a movie: DVD, Special Edition DVD, VHS, foreign releases. So if you’re trying to find out what extra features are included with a specific release, you have no idea. You’ll read a review about how this movie comes with a bunch of bonus features and then realize they were actually reviewing a different release. Or read a review about how there are technical problems with the release and not know if the reviewer was talking about this release or some foreign bootleg they bought.
The most egregious error I’ve encountered on my Kindle was a book in a series I’ve been reading in which one of the main characters is named Quinton. It quickly became clear that somebody ran a search & replace to make sure his name was capitalized every time it appears … because every single word in the book that starts with the letter Q is capitalized. This resulted in some confusion each time I encountered it, because in this particular genre (fantasy) it’s not unusual for an ordinary word to be used for a mystical concept, and it’s fairly common practice to capitalize the word to indicate it’s the “mystical” version that’s being referred to rather than the “ordinary” version. So when I would come across “… the Quiet …” or “… the Quick …”, I’d subconsciously “perk up” thinking there was some significance there, before realizing it was just another sloppy search & replace error. I e-mailed the author about it, and she told me she was aware of it and had complained to her publisher without results. But she said she’d pass my comment along and hope that perhaps customer complaints would move the publisher to fix it.
OTOH, I’ve been working my way slowly through the Cambridge Medieval History, a huge, five-volume scholarly work originally published around 1910-1920, and it is literally riddled with OCR errors. But I can forgive this because it was provided by an essentially volunteer project and only cost me $2.99.
For certain classes of non-errors, possibly quite high. There are a lot of commonly held misconceptions about language. And how do you handle authors who purposefully use language in a nonstandard way and the plethora of readers who don’t get it?
Plus, with the internet, assuming that crowdsourced content submissions are independent generally isn’t safe. Especially if there’s a monetary reward to be had. How hard is it to code up a botnet that will submit fake errors to get the $0.55 payoff of making it through your filters. How many resources do you need to devote to verification and filtering? Are there actual people who are motivated by an (average) $0.055 bounty, or are you better off not offering money and reducing the incentive to botnets (not eliminating it, just reducing it). Are there vandals who will put dirty words into childrens books because they think it’s funny? Are there political ideologues who will try to distort the words of their opponents? God forbid you attract the attention of the /b/tards.
At first glance, crowdsourcing often seems like it can magically solve all kinds of problems, but the outcome is often not so simple, and possibly is worse than just putting up with some typos.