Social DRM

[odedta]

Hello,

Does anyone know of a free open source Social DRM for ePubs? Please spare me the lectures about why I should or shouldn't use it, thanks.

[PeterT]

Does this help?

exlibris - add a nice ex libris to your EPUB eBook!

[Freeshadow]

Not something I heard of before but, you could script one like this:

1) Take your "social_identifier_string" of choice
2) make a QR code sized 1 pixel per dot of it.
3) convert the QR image to SVG
4) insert the SVG downsized to 1 pixel into your cover or chapter separator / header / deco... SVG

As long as it isn't known where exactly it is people wouldn't even find it by stringsearch being converted to small dots in svg-lese inside another svg. Eyeball-search of every SVG in the epub under extremely high magnification because every dot of the image could be the QR code sized down.

Having never heard about such attempt of implementing social DRM I assume in good faith that there is no preceding case of it and hereby declare this idea open source

[DiapDealer]

I only know of one, but it's not lecture-free.

[Toxaris]

Originally Posted by DiapDealer

I only know of one, but it's not lecture-free.

That is funny!

[odedta]

YES! Those two ideas were exactly what I was looking for!

@Freeshadow: that is a great idea and I will implement it.

@DD

Thanks a lot!

[skreutzer]

Are you serious, implementing any kind of DRM? Here's a way to circumwent “soft-DRM”, “watermarks” in e-books: two independent people obtain the file from its source and then diff them against each other. If there is no difference, obviously the Social DRM isn't able to trace a file found later in the wild back to the original purchaser, and if they are different, just remove the differences and create a plain file without the useless extra garbage which shouldn't have been there in the first place, and wouldn't need to with an ethical, good approach or business model.

[DiapDealer]

You clearly did not read the instructions.

[skreutzer]

Oh, sorry, I just missed that this was solely a fun question without real-world relevance, and that suggestions are supposed to provide a scenario for a programming exercise, because the correct answer, if the question would be serious, would be that there is no solution that actually can work. Maybe that's why it is called “soft-DRM” – not because it doesn't get as much into the way as “hard-DRM” does but because it is easy to remove. Additionally, probably the identifier of another person could be faked, which then would be blamed for unauthorized distribution.

In this case, I would suggest to investigate if time stamps of the files in an bundled e-book file package can be used, because nobody would check differences in minutes and seconds, which also could be legitimate information as result of the e-book creation process.

[pdurrant]

Originally Posted by skreutzer

In this case, I would suggest to investigate if time stamps of the files in an bundled e-book file package can be used, because nobody would check differences in minutes and seconds, which also could be legitimate information as result of the e-book creation process.

That's an interesting idea. Probably wiped by any conversion or editing, but interesting.

[Freeshadow]

Wiped by every touch (yes pun intended)

[skreutzer]

It's unfair to point out this disadvantage of the method I suggested, because every other method has the very same disadvantage! Every “soft-DRM” attempt might be easily removed by editing or conversion, and even if removal doesn't happen (by accident or with intention), one might still replace an identifier with an invalid one or with a valid one of another individual, so you can't savely assume that the “identifier” identifies an “legitimate” user, which in the worst case will be blamed for stuff he didn't actually do. But having “soft-DRM” in place obviously implicates that it is intended to cause consequences for somebody, while it won't work for a version where the soft-DRM was removed, so only ordinary users get into trouble who indeed already have paid the author and without doubt can't carry out mass distribution by themselves or would be able to prevent/influence it at all.

[odedta]

How do you feel about the notion that when you cannot have security, at least try to detter the thief as much as possible?

[Freeshadow]

@ skreuzer The keyword here is "social".
If it's non intrusive a fair user has no interest in removal.
An unfair one needs nothing but two copies from varying users, diff in a shell and some braincells.
Might even have own interest in keeping unique markers on own stuff.
Think: name on clothes/personal logos/wallpapers on devices etc.
A QR encoded but not first-sight in plain view embedded identifier is subtle, machine readable and thanks to the rising spread of smartphones yet easily accessible to humans.
embedding it as svg code inside existing SVGs makes it at least partly resistant to non intended damage, math games with timestamps are the direct counterexample of resistance. A wee typo correction or inclusion of a custom meta data field is highly likely to screw the identifier up.

[pdurrant]

Personally, I'd like it if books I bought had a nice graphical title page with my name, date and place of purchase, purchase price, and a facsimile of the author's signature.

That's the kind of social DRM it would be nice to have.