|
@Freeshadow: Read-only zip files won't help - nothing to prevent extracting the files and rezipping them in a new zip archive. Only way it would work is if only certain kinds of ereading software had read privileges - and then you have a proprietary DRM system
@SeeUsmile: Apologies for butting in, hope replies from other than the redoubtable Jellby can be of interest too ;-)
The only way I can think of offhand is to put a repeated floating text box/image on the left- or righthand side, sufficiently frequently to ensure that it appears on every page on all viewers.
Not pretty...
In epub3 you'll probably be able to have a fixed header/footer, maybe even a background image?
@ldardini: Watermarking is a subject I find fascinating. As has been commented, it is probably impossible to make a rock solid watermarking solution for epubs. A few points to bear in mind:
• Epub watermarking, unlike encryption, depends on the method being secret.
• The social aspect of watermarking vs. DRM must be very visibly conveyed. Some kind of preamble that says "Be nice, don't share, 'cause we've been nice and not used DRM. This file has been watermarked visibly and invisibly as your personal copy. We trust you not to try and remove these marks." (It's my conviction piracy has become more rampant because media companies treat their customers with suspicion of criminal intent, which customers then feel compelled to justify;-)
• The watermarking must instil a well-founded fear in a would-be pirate that he or she can never be sure of having removed all watermarks from the file.
• From this follows that several watermarking methodologies must be used concurrently, and that they must be updated or changed frequently as they are discovered and revealed by pirates.
• Except for deliberate visible watermarking, e.g. an "ex libris" or "this book lent to John Doe", the visual appearance of two ebooks with different watermarks must be the same.
• Hiding information in images is well-studied.
• Hiding information in the zip-file itself is likely to be ineffective, as a would-be pirate will normally find it most convenient to create a new zip archive for their modified copy.
• Hiding information in a xhtml/opf file is not so well-studied. I believe there are some commercial solutions. There is a wealth of possibilities in producing quite different files that map to the same visual output. Spacing, line breaks, id tags, even character encoding can all be different in the xhtml encoding and still be equivalent. Unfortunately, this hidden information can be efficiently removed by converting the xhtml to text, and then reencoding the text as xhtml. This method can be made less appealing by closely integrating text and images, by using complex layouts like tables etc.
@Jellby: At first sight it doesn't seem too difficult to automatically insert some kind of ex libris into an epub file.
• unzip
• Create the ex libris as a standalone xhtml file.
• insert references to it in the manifest and spine in the opf file
• rezip
Isn't the opf-file fairly strictly defined, with a not too complex xml dtd? Then you should be able to do something like this and be fairly confident of a valid result, shouldn't you?
|