Neil Gaiman on internet piracy

[Thanas]

Neil Richard Gaiman is an English author of science fiction and fantasy short stories and novels, comic books and graphic novels, audio theatre, and films. His notable works include the comic book series The Sandman and novels Stardust, American Gods, Coraline, and The Graveyard Book. Gaiman's writing has won numerous awards, including the Hugo, Nebula, and Bram Stoker, as well as the 2009 Newbery Medal and 2010 Carnegie Medal in Literature. He is the first author to win both the Newbery and the Carnegie medals for the same work

But remember, it is just like theft.

[Xon]

I raised this point in the "what if copyright went away" thread and everyone ignored it.

Advertising and reaching people that your product is worth buying is incredibly expensive. Sinking a hundred million into advertising a single movie isn't actually very uncommon now.

[Xon]

Final point (tl;dr) is that Neil Garman's view comes from, exclusively, his experience within the Written Medium. Therefore, it cannot be used to defend or even make any strong commentary, and the problem that is "Internet Pirating".

The primary point you miss from Niel Gaiman's talk is it's all about the free advertising and word of mouth. Movies are different because the movie producers literially create a market via advertising. These days, advertising is just as if not more than important the quaility of the work being sold. For non-block buster releases, it costs to bloody much to drum up support for your product, so you are reliant on grass-roots todo it for you.

Yes, the dynamics change when you are the market leader, but that is because everyone already knows about you and then only way to go is down. Even then, companies need to keep advertsing or usage of thier product drops like a stone. And this is for stuff like coke or pepsi which have been advertising for longer than effectively everyone on this board has been alive. Never mind a random movie you've never head of before.

Then you can move onto that the complexity or ease of the initial act of copyright infringement itself is meaningless. After all, for everyone else it's just a matter of triggering the download off.

Copying novels is quite a complex process. It literially entails ripping the book apart so each page can be scanned, then compiling all the pages into a single image, OCRing it (converting from an iamge into text) and then proofreading it to correct the inevitable fuckups where the software has rendered something into text incorrectly. Even vanity published works like Stuart Slade have had this done to them. Popular releases will have this done to them inside of a few days of a retail release. Then in addition, fan translations mean translating an entire novel.

Digital Media from dvds/blu-ray/etc are even easier to copy and require a fraction of the effort. Cracking games is actually very difficult to achieve and requires a very focused and solid understanding of computers.

[General Zod]

Copying novels is quite a complex process. It literially entails ripping the book apart so each page can be scanned, then compiling all the pages into a single image, OCRing it (converting from an iamge into text) and then proofreading it to correct the inevitable fuckups where the software has rendered something into text incorrectly. Even vanity published works like Stuart Slade have had this done to them. Popular releases will have this done to them inside of a few days of a retail release. Then in addition, fan translations mean translating an entire novel.

Scanning a novel isn't nearly as difficult as you might think. And do we honestly need to bring up Stuart in every fucking thread about piracy?

[adam_grif]

This is actually Neil Gaiman talking about internet piracy, despite what the topic title led me to believe.

[MKSheppard]

Scanning a novel isn't nearly as difficult as you might think.

I'm afraid you honestly don't know a damn thing about scanners. Handheld scanners, unless they've made some magical boost in quality will produce images with noticeable "steps" in them as the person's arm twitches etc.

OCRing is very much a case of GIGO. Put a clean scanned image in, and little if any proof reading is needed. Put crap image like from one of those scanners, a lot of proof reading/retyping is needed.

[Patrick Degan]

OCRing is very much a case of GIGO. Put a clean scanned image in, and little if any proof reading is needed. Put crap image like from one of those scanners, a lot of proof reading/retyping is needed.

We've never even bothered with OCR scanning at our office; we'd have to edit and reformat copy anyway with too much of the material we'd use for ad text, so there's just no point to it.

[Phantasee]

And yet there is an abundance of easily available digital copies of popular novels and comic books...

[fgalkin]

Considering that almost everything is now released in ebook format, do people really even bother with OCRing rather than just distributing the ebook?

Have a very nice day.
-fgalkin

[Guardsman Bass]

And yet there is an abundance of easily available digital copies of popular novels and comic books...

It really depends on the fan-base for a work. There were illegal online copies of Robert Jordan's Towers of Midnight available within a week or two of the hardback release.

No, I didn't download one (I own the hard-back).

[Xon]

Considering that almost everything is now released in ebook format, do people really even bother with OCRing rather than just distributing the ebook?

Sure, when they are avaliable. Except the physical book is quite often released sometime before the official ebook. Even then, you still get people buying stuff like Baen's Advanced Reader Copies and sharing them (they are slightly less-proofed copies avaliable upto 2-3 months before the official release). The amusing bit is when the ARC copies then get fan-proofed.

It really depends on the fan-base for a work. There were illegal online copies of Robert Jordan's Towers of Midnight available within a week or two of the hardback release.

That title had an scanned copy out in days, it was a really rough OCR(ugh multi-megabyte pdf), by the end of the first week it was a almost as good as an officially published ebook. I'm got the actual hardcopy somewhere on a shelf but I prefer to read on my computer or on the go. Plus I can't lose an ebook because I put it down and a family member snatched it up.

Scanning a novel isn't nearly as difficult as you might think.

You are still scanning every page, and you still need to OCR the resulting image. That is nothing more than a fancy handheld scanner.

And do we honestly need to bring up Stuart in every fucking thread about piracy?

He is an awesome example of how an author and, apparently, publisher will apparently pull a project because draft posted on a publically readable forum showed up on a torrent site. This wasn't a case of someone breaking NDA, breaking company rules, breach of contract, or commiting fraud to leak a pre-release copy. It's the equivelent of taking a photo of an art's display in public view and then pinning the photo on a community bulletin board(which still actually exist).

The whole point of what Gaiman was talking about, is that wasn't a lost sale but free advertising. Which is capable of real and demonstrable financial gains.

[Edi]

Pebkio's banning split to Parting Shots.

[General Zod]

You are still scanning every page, and you still need to OCR the resulting image. That is nothing more than a fancy handheld scanner.

How does that change the point that it's easier than you claimed? Have you even used any recent OCR software? The technology has improved considerably to the point where unless the document is completely shit it's going to get 95% of the scan correct right out.

He is an awesome example of how an author and, apparently, publisher will apparently pull a project because draft posted on a publically readable forum showed up on a torrent site. This wasn't a case of someone breaking NDA, breaking company rules, breach of contract, or commiting fraud to leak a pre-release copy. It's the equivelent of taking a photo of an art's display in public view and then pinning the photo on a community bulletin board(which still actually exist).

It's hardly a relevant example at all. Nothing was preventing him from vanity-publishing aside from the fact that he's a giant crybaby. Most publishers and movie producers can't afford to pull an project just because someone leaks a copy early because by then they've already invested quite a bit of money into it. The Wolverine Origins movie is the most relevant example I can think of that was leaked early but still put into theaters.

[MKSheppard]

How does that change the point that it's easier than you claimed? Have you even used any recent OCR software? The technology has improved considerably to the point where unless the document is completely shit it's going to get 95% of the scan correct right out.

Have you? I've OCRed a lot of things; and the only way you can get 95% or better accuracy is if you have a 300 or more DPI scan from either a fixed flatbed scanner or from one of the high end camera-based scanners that google uses for Google Books.

Also, something as random as the typeface that the book uses can mess with the OCR program. And this is with a $400 program I own and only one generation behind the latest (version 9 vs 10).

[General Zod]

Have you? I've OCRed a lot of things; and the only way you can get 95% or better accuracy is if you have a 300 or more DPI scan from either a fixed flatbed scanner or from one of the high end camera-based scanners that google uses for Google Books.

Considering clients where I work enjoy sending us PDFs in horrible formats that we have to print out and re-scan, yes.

Also, something as random as the typeface that the book uses can mess with the OCR program. And this is with a $400 program I own and only one generation behind the latest (version 9 vs 10).

Who's going to waste $400 on OCR software for fiction? The stuff we use at work is just a little over $100, and that's for the professional version. If someone was really intent on pirating books they'd likely just get cracked OCR software. Doing a search for typos in Word after it's been scanned takes all of an hour, if they really want to be thorough. Frankly you people are overestimating how difficult it is to scan books.

[Xon]

How does that change the point that it's easier than you claimed? Have you even used any recent OCR software? The technology has improved considerably to the point where unless the document is completely shit it's going to get 95% of the scan correct right out.

I haven't had to OCR anything for a while, but I have had to read OCR'ed stuff and while it isn't nearly as shit as it was it still isn't great. 5% errors can really matter depending on where those errors are and if you manage to hit a systematic weakness.

Also the major point I was initially making is it's the scanning of the entire book which is the hard work, and then proofreading is a significant investment in time. Honestly the difficulty of the OCR (which really is point & click), is really irrelevant. It isn't like ripping a movie or tv series, instead it is computationally expensive. It isn't a significant investment in a person's time actually interacting with the original work.

It's hardly a relevant example at all. Nothing was preventing him from vanity-publishing aside from the fact that he's a giant crybaby.

It's still a good example of an author who see any copyright infringement as theft despite it being capable of being free advertising. And yes, the 'deal' he apparently has with his publisher is the sunny side of stupid since he claimed they wheren't willing to publish as long as there was a single google reference to the torrent.

Since this got split when the Pebkio got banned;

Exactly where is the demonstratable harm for a series of release bots to populate a site followed by a series of a downloader bots automatically grabbing everything, with no demonstrable correlation between the amount of material moved and actually used? Furthermore, the assumption that every act of downloading or publishing torrents actually depriving the original content owners of a sale if really bad logic. Never mind, that even if someone is copying the content, it does not mean they are illegally doing so.

Lets put it this way, how the hell can a person harm a content producer via copyright infringement if said content producer is producing works which will never be made avaliable in that person's region? This isn't some hypothetical arrangement, it happens with TV shows all the damn time. There are shows which have only shown on ad-sponsered TV decades ago and have never been repeated, and never had a DVD release. Exactly how can someone deprive the content owners of a sale when the bloody stuff isn't avalaible to begin with?

I've often been perplexed at content creators who chuck a hissy-fit over foreign copyright infringement when they don't even sell stuff in that country nor plan to. Especially when it isn't in the local language, yes people could import but they simply can not be statistically significant compared to area's which know that the product even exists!

[General Zod]

I haven't had to OCR anything for a while, but I have had to read OCR'ed stuff and while it isn't nearly as shit as it was it still isn't great. 5% errors can really matter depending on where those errors are and if you manage to hit a systematic weakness.

Also the major point I was initially making is it's the scanning of the entire book which is the hard work, and then proofreading is a significant investment in time. Honestly the difficulty of the OCR (which really is point & click), is really irrelevant. It isn't like ripping a movie or tv series, instead it is computationally expensive. It isn't a significant investment in a person's time actually interacting with the original work.

If you haven't had to OCR anything for awhile how can you make any kind of claim as to how much time it takes? It sounds to me like all of this is an assumption on your part.

It's still a good example of an author who see any copyright infringement as theft despite it being capable of being free advertising. And yes, the 'deal' he apparently has with his publisher is the sunny side of stupid since he claimed they wheren't willing to publish as long as there was a single google reference to the torrent.

Wasn't the material publicly available anyway? I have trouble believing that a vanity publisher would give a shit about whether or not it's been torrented.

[Xon]

If you haven't had to OCR anything for awhile how can you make any kind of claim as to how much time it takes? It sounds to me like all of this is an assumption on your part.

I listed the OCRing as part of it, but honestly, even with shitty OCR software physically scanning an entire novel is the slowest part of the process. It's the entire process end-to-end which is slow.

Wasn't the material publicly available anyway? I have trouble believing that a vanity publisher would give a shit about whether or not it's been torrented.

That's what Stuart Slade claimed was the issue with hs current publisher. Right after that "generously" giving some ukranian kid a week to remove the links to the torrent(which said kid posted and then joined the forum and linked to it) from google before threating to take up legal action and aluding that he had already started. But this is starting to get off-topic.

[Uraniun235]

I've never done it myself but I would have thought that someone looking to scan a book would undo the binding so that the pages could be run through a scanner automatically.

[General Zod]

I've never done it myself but I would have thought that someone looking to scan a book would undo the binding so that the pages could be run through a scanner automatically.

If you don't mind destroying the book in the process, sure.

[ThatOneCatC]

OCR software has increased in quality dramatically in the last 10 years. We routinely scan color/ inverted pages (magazines, photo journal captions, advertisements and such) as .jpegs at 200 dpi and are getting a steady 90% and up rate on text capture. On a 300 dpi Tiff we can easily reach 99%. Add in the automated despeckle (which no longer counts punctuation as speckles) and the output of OCR for poor quality text can actually make the text easier to read in the native format. I was given a (as of yet) unreleased software that can OCR ink stamps that are at an angle on the page. We have used it with great success in exporting to different formats. The speed of OCR has also increased to the point where each page can essentially be OCR'd at the time of scan. The larger reference materials we scan require a different hinge on the flatbed to accomodate the larger size as well.

All in all it does make the scanning of books far more tedious than it is worth in order to pirate them. The OCR process is the easiest thing we do.

[Elfdart]

I can't help but laugh when music companies and movie distributors plead poverty. Music companies are the worst. They spend ludicrous amounts paying radio stations to play their music and giving out free concert tickets to promote it because they know that the more people are exposed to the music, the more likely they are to buy it.

I would have thought that a record company would be elated that teenagers are distributing for free what the company spends a fortune in cash, drugs, concert tickets and hookers for DJs and PDs in order to get them to play the latest and shittiest single.

Fuck 'em.

I remember reading where Monty Python didn't get YouTube to yank their videos and their sales increased by a huge margin.

[Molyneux]

An interesting and edifying video - and it illustrates precisely why I feel the need to challenge any assertion that piracy obviously damages the music industry. That simply cannot be taken as a given.

[Xon]

Comercial productivity software "piracy" also has a subtle benefit that most people love to ignore. Highly used products get pirated by those who don't have it and learn to use it. Then when they enter the workforce and thier employeer is paying for the licence they don't need a fraction of the trained. There is a good reason software houses tend to throw software to students who ask.

Training people to use Windows, Microsoft Word or Excel is actually crazy expensive, something I wish I wasn't joking about, and the ability to hire workers who already know how to use those products is a major productivity bonus. It's the classical argument for governments footing the bill for education, but now about software.

When you start having network effects dominating the social utility of something, it becomes in everyone's best interests to ensure as much of the population is onboard as possible.

[Sarevok]

What about small developers like indie game companies though ? They can't afford losses if most of their users are pirating because their sales figures are tiny to begin with.