JK Rowling Announcement

[mikaelalind]

Originally Posted by anamardoll

10 to 1, it'll be the buyer's credit card number. Always a good thing to embed in data that will be streamed over unsecured WiFi networks and 3G.

I doubt it will be credit cardnumber. Elib uses watermarking, and I doubt that Swedish law would allow them to use creditcardnumbers. Besides, I don't think there is time to add the CC number. But maybe UK laws are different.

[anamardoll]

PLUS, I do not want my credit card data living in a book that streams over the WiFi/3G to my device. That's just basic d'uh.

What are they going to embed to identify users? About the only thing they can verify from an online purchase is the CC#.

And god help us if they decide to keep a database where 1 = Ana = CC#; 2 = Diap = CC#; etc. That's just BEGGING to be hacked and stolen.

[EowynCarter]

An other watermark problem : company buys pBook. pdf download, watermarked, comes with it.
Makes sense that everyone in the company can get the pdf, but, it also means having to trust everyone not to distribute the pdf....

[anamardoll]

Originally Posted by mikaelalind

I doubt it will be credit cardnumber. Elib uses watermarking, and I doubt that Swedish law would allow them to use creditcardnumbers. Besides, I don't think there is time to add the CC number. But maybe UK laws are different.

How do they maintain a trace to, say, Ana Mardoll? I'm curious. Are they maintaining a trace database of book number #109796327846237486 = Ana Mardoll? They can't tie it to the user name, though, since a user name is meaningless. They'd either have to keep the CC# or the user address, and the CC# is the only thing easily verifiable and relatively unique to each user.

What's going to stop that DB from being hacked for all those juicy CC#s? I'd actually PREFER the CC# be embedded in the book (one download) than in a database that will always be vulnerable.

[ilovejedd]

Originally Posted by anamardoll

Seems like you could at the very least replace it with an exact sized image that is blank.

You'd just save a 348 px by 653 px blank image as HPWATERMARK.jpg or whatever and zip it all back up.

That only works if you know exactly which image contains the watermark. It's possible to add extraneous data to images and still have it displayed properly. Besides, I don't think they'll be stupid enough to name the watermark so obviously. You're going to have to strip out all images and check all html and xml files to make sure there are no hashes that include the digital watermark. If they do use CC# as watermark, it will no doubt be encrypted or stored as hash. How difficult it would be to break the encryption, we don't know.

Although, no doubt, someone will figure out how to remove the watermark within a week of release.

[anamardoll]

Originally Posted by ilovejedd

That only works if you know exactly which image contains the watermark. It's possible to add extraneous data to images and still have it displayed properly. Besides, I don't think they'll be stupid enough to name the watermark so obviously. You're going to have to strip out all images and check all html and xml files to make sure there are no hashes that include the digital watermark. If they do use CC# as watermark, it will no doubt be encrypted or stored as a hash. How difficult it would be to break the encryption, we don't know.

Although, no doubt, someone will figure out how to remove the watermark within a week of release.

Precisely. *We* don't need to know what they named the file if it's an image -- I was being facetious with my name (although how many images are embedded in a typical epub, really? The HP books aren't illustrated.) There's a big ol' internet out there and Google is a good friend.

I actually find a watermark more "invasive" than DRM. I find this whole thing....disappointing. To say the least.

[ilovejedd]

Originally Posted by anamardoll

The HP books aren't illustrated.

There's the cover and I'm pretty sure I've seen a few illustrations in my books at the beginning of some of the chapters (US release, Mary GrandPre?). Granted, it's been a while since I last read HP. After the convenience of ebooks, it's just a pain to go back to reading heavy pbooks with actual pages.

[anamardoll]

Originally Posted by ilovejedd

There's the cover and I'm pretty sure I've seen a few illustrations in my books at the beginning of some of the chapters (US release, Mary GrandPre?). Granted, it's been a while since I last read HP. After the convenience of ebooks, it's just a pain to go back to reading heavy pbooks with actual pages.

So, being really generous, maybe 50 image files in the epub?

I doubt it will be a traditional watermark image, however. I'm very interested in what data they plan to embed or store to "identify users".

[CleverClothe]

Originally Posted by DiapDealer

Amazon hasn't used Device specific PIDs for Kindle books for quite some time now, but your ultimate premise is still correct.

BTW, I'm not bitching about this... I'm genuinely curious as to how this will be achieved.

Trying not to derail the discussion, but I hope your just trolling. Kindle DRM could not work without device specific PIDs. It is key to how they limit the number of devices. It is also how the DeDRM tools work, by discovering the key for your device.


OT:

It is easy to find a watermark like that. Just have two people buy the book and compare the files. The watermark will stand out like a sore thumb. Then you can focus on the surgery.

If it is in an image, you can chop it out from a header. Or if it is throughout the image data but not visible, you can screenshot the rendered version and then clean it up. No more watermark.

[ilovejedd]

Originally Posted by CleverClothe

It is easy to find a watermark like that. Just have two people buy the book and compare the files. The watermark will stand out like a sore thumb. Then you can focus on the surgery.

If it is in an image, you can chop it out from a header. Or if it is throughout the image data but not visible, you can screenshot the rendered version and then clean it up. No more watermark.

See, the ebooks aren't out yet and there are already methods for finding the watermark. Even if you don't buy two copies, it's possible to remove the watermark through inspection of the files. Granted, it'll be quite time-consuming.

No doubt, we'll see tools to remove the watermark shortly after release.

[Dulin's Books]

Originally Posted by EowynCarter

I don't want watermark to spoil my books. Now what

it wouldn't be one you can see while reading. but it will be locked in your copy so if you make a hundred copies and pass them out those copies can be traced to you.

[kalex]

and its going to be partnered with Sony we all know how secure that will be

[ilovejedd]

Originally Posted by kalex

and its going to be partnered with Sony we all know how secure that will be

Lol!

[ApK]

Originally Posted by CleverClothe

Trying not to derail the discussion, but I hope your just trolling. Kindle DRM could not work without device specific PIDs. It is key to how they limit the number of devices. It is also how the DeDRM tools work, by discovering the key for your device.

Have you seen Alf's blog under the section "Q: Who wrote these scripts?"

[HarryT]

Originally Posted by CleverClothe

Trying not to derail the discussion, but I hope your just trolling. Kindle DRM could not work without device specific PIDs. It is key to how they limit the number of devices. It is also how the DeDRM tools work, by discovering the key for your device.

You do realise that Diap knows more about DRM than you could ever hope to in your wildest dreams? I think you can be confident that he's very well aware of how Kindle DRM works .