Originally Posted by cybmole
out of curiosity - how easy would it be to build a pottermore type solution i.e. don't try to add any DRM but instead add some unique seller info to each copy sold. could that be a simple scripted process with the right tools ?
i would guess that the sort of folks who will honestly purchase a self published book from an author's own site are not the sort of folks who will then upload it to a torrents site or send copies to everyone in their address book so I would question the need for drm anyway - but I suppose it only takes one bad apple...
While better than DRM in my opinion, it's not really possible to do it securely, as Kovid explains in this post: http://www.mobileread.com/forums/sho...98&postcount=6
. Basically any non-randomised watermarking can be easily detected and removed, and absolutely all watermarking will succumb if the user copies the content in some analog way and re-creates it, such as rendering images of the book and OCR-ing them. This will beat even one scheme I saw which encoded information by using different unicode homoglyphs (which has its own set of problems regarding fonts).
There is also the issue that a user can "legitimately" lose his file, for instance I've had my reader stolen on one occasion.