France requires non-proprietary format for ebooks

[xdrudis at tinet.cat]

Thank you very much for your explanations.

On Tue, Jan 27, 2015 at 08:06:59PM +0100, Jil Larner wrote:
> worked on the law, we would know if “non-proprietary” meant no DRM, or
> just no non-proprietary DRM.
>
[...]
> However, the case of DRM is a bit more complicated. As of today, there
> is no open specification of DRM for e-book [someone, please double check
> this], which forces publishers to provide an ePub without DRM (to match
> the “non-proprietary” obligation), but the coming months will show if
> publishers play the game of full interoperability, or if they release
> non-proprietary DRM in order to comply with their new obligation.
> 

Possibly redundant here but there is no possible working DRM in
effectively free software, because the goal of DRM is to remove some
freedoms from the user. I think the closest is some tivo-like scheme
in which the software would be free but unrunable when modified.  You
can have open formats for DRM as long as reading the material depends
on some private key the user doesn't have, and any software that uses
it either has to be propietary or signed and prevented to run with the
private key if modified.  It's not something we still don't have, it's
something that can't happen because requirements are
contradictory. Software freedom requires the ability to decode and
copy the content, as part of freedom 0, and DRM needs to prevent it.
Free software needs control to be granted to the user and DRM needs
control granted to the rights holder.

> publishers have made mandatory for the e-book to be “available in a
> usable technical format, taking into account the market's usual formats
> and their evolution, and in at least one non-proprietary format.” By
> decree, the Minister of Culture has made this code of practises
> applicable to all authors and publishers in France.
>

I wonder if the publisher may work around the arret by publishing a
work in two versions, one proprietary market usual format and one
non-proprietary obscure or DRMized format that noone can really
use. They would expect profit from the first version and compliance
from the second. Imagine for example one book published in both Amazon
format and GPG encrypted ePub without publishing the private key
needed for decrypting it (would that count as publication?  I hope
not) . You could even have readers that can read the encrypted ePub if
unmodified software is run which can then be given the private key by
some TPM module (à la secure boot) or internet server (à la remote
attestation). I'm oversimplifying but I hope you see what I mean.

> Stakes for the publishers are high, as the obligation for
> non-proprietary e-book is at the heart of the contract by which they
> obtain exploitation rights from authors. That is, should the publisher
> not comply, the author can withdraw digital rights for his book and give
> it to another publisher.
>

I think it's still good news, compared with places where we don't even
have this requirement, and it may be an excuse for publishers to leave
a DRM nightmare that as happened already in music is not profitable,
and turn to interoperable, open format and wider-market ebooks that
people buy even if they could copy because they have the right price
and convenience.

I don't have a clear definition of "proprietary" though (or
"non-proprietary"). I might be overpessimistic.

> Anyhow, this code of practises begin written by authors and publishers
> altogether, it seems France is turning down the transformation in which
> a book you owned became an e-book a company controls for you. Let's hope
> France will continue to walk toward open litterature.
>

I hope so, because I haven't looked into any details, but I'd like
this debate to have included the readers, the public, not only right
holders. I guess they're represented by government, not even
Parliament in this case, but the authors and publishers are
represented both by government and their associations, so it's still
unbalanced.