[DE | EN FR | JA | ES | KO | PT]

Welcome to a new issue of the Brave GNU World -- this time with a focus on Free Documentation, publishers for Free Books and the larger picture in this area.

Free Documentation

Quite a bit about the importance of Free Software has been written in this column already. But transfer of knowledge can not be done by software alone, since the implementation of concepts and knowledge in software is a lossy process. When programming, some information is always lost that cannot be regained from the source code alone.

Depending on the complexity of the program or problem, studying the source code alone is an inefficient or hopeless attempt at understanding the function of or knowing how to use a program. Both are essential freedoms of Free Software, however.

Technical documentation is a much more efficient way of transmitting that information with much more background that is usually lost when transferred into software. So it is logical that the concepts of Free Software and Free Documentation are tightly linked. [5]

In fact, Free Documentation is essential for Free Software. Although it is true that much of the knowledge on how to use a program or how it works can also be conveyed by means of proprietary documentation, but only at the cost of limiting the freedom to distribute modifications: software can only be distributed without documentation -- with the aforementioned disadvantages to the recipient.

The freedom to modify a program is also effectively limited, because non-free documentation does not allow keeping the documentation up-to-date with the software. This leads to misunderstandings in using for future users and also puts limits on further improvement, as future authors first have to find out about the difference between a piece of software and its documentation by studying the source code.

So effectively, just like non-free software, non-free documentation is quickly degrading in its usefulness due to lacking maintenance capabilities.

Additionally, proprietary documentation of Free Software is putting limits on equal chances and empowerment, because northern licensing fees are often unaffordable to students in poorer regions.

So Free Software plays a key role in upholding freedom not only on paper but also in reality for all people in the information societies to come.

GNU Press

GNU Press [6] is the publishing project of the Free Software Foundation (FSF) complementary to the GNU Project, into whose development all proceeds from GNU Press go. It is the goal of GNU Press, to publish affordable computer science books under Free licenses for Documentation [7].

GNU Press is being managed by Lisa "Opus" Goldstein, a long-term activist for Free Software, who knows the Free Software Foundation From its early days and is also managing the office of the FSF in Boston.

Green Tea Press

Another -- at the moment rather small -- publisher for Free Books is Green Tea Press [8] by Allen B. Downey and Lisa Cutler. It was originally started to publish the book "How to Think Like a Computer Scientist: Learning with Python" of which now also a C++ and Java variant exists.

Another new item in the inventory is "Learning Perl the Hard Way" and the future will hopefully have us see more Free Books from Green Tea Press. To a large extent, this will be up to the readers.

On the page, Allan Downey describes five easy ways [9] with which a reader can further Free Books -- beginning with linking to them. Another way is to support the buildup of distribution channels through requests at local book shops and the large vendors. And finally recommending them to libraries and teachers is another way of supporting Free Documentation.

Network Theory

And last -- but definitely not least -- Network Theory [10] by Brian Gough should be mentioned. He was also seminal to initiating this focus issue of the Brave GNU World because it was his idea to write about publishers of Free Documentation and the corresponding issues.

Network Theory was founded two years ago after the realization that maintaining the freedom of Free Software also requires Free Documentation and Brian Gough had just finished writing the "GNU Scientific Library Reference Manual."

As it was too large to print and his office was already swamped by printouts of other manuals, he decided to use the knowledge about publishing acquired in a previous job and founded Network Theory to publish these first books.

By now seven titles have been published, among them the official CVS handbook. Another five are in preparation. Also the profits are not only used to fund Free Software and organizations like the Free Software Foundation, they are also used to hire authors to write more Free Documentation.

So two new titles "An Introduction to GCC and G++" and "An Introduction to GNU Bash" were both paid for by Network Theory and should come out November 2003 in print and under the GNU Free Documentation License (GFDL).

Brian also sees the best way of supporting Free Documentation in becoming a loyal customer and encouraging companies and libraries to buy these books.

Also Brian would very much like to hear about demand for new books as well as being contacted by authors interested in publishing Free Books.

Other Publishers

In other, more "traditional" publishing companies, Free Documentation has also become more of a topic recently. For instance O'Reilly Associates now has several Free Books on their list. A list can be found online. [12]

Even publishers like the German Springer Verlag show first steps in this direction -- Peter Ganten for instance received permission to at least distribute his Debian GNU/Linux book without charge in digital form. This still does not permit modification and reprint, so it is clearly non-free, but it is a first step into the right direction.

But especially the permission for reprinting is likely to have any traditional publisher turn it down as a knee-jerk reaction -- for understandable reasons.

Copy vs. Create?

The consideration goes like this: if a publisher can reprint a book without having to pay authors, editors and printers, that book will be considerably cheaper than the original and therefore put the original book out of business.

The result would be a negative selection in which -- if taken it its extreme -- only reprinting companies survive and therefore at some point in the future no new books would be written.

But there are several reasons why this is a gross exaggeration.

Firstly, humans have been creative at all times and there is no reason to assume that this would change fundamentally. Of course financial reward is one of the motivating factors, but what is done for financial reasons only is usually what people refer to when they talk about the downfall of cultural life.

In fact those works that are not mass-compatible but culturally and socially valuable do not allow their authors to make a living today. They effectively disprove the "no creation without property rights"-myth so often used for expanding monopolization of knowledge.

Additionally, thanks to global networking and the possibilities of a global, decentralized distribution network though multiple channels, a reprint could never fully replace the original even if it comes at a lower charge. In many cases, booksellers and the distribution channels would -- because of their knowledge of the background -- still prefer the original.

Also, reprinting can be a very socially useful activity -- when for instance the book has gone out of print and the original publisher has decided to not reprint it for some reason. In this case the original publisher -- having decided freely to ignore the demand and not reprint -- does not suffer any serious disadvantage.

Still the considerations about the socially harmful effect of pure reprinting is not entirely without truth.

Pragmatically speaking, the optimal solution would be one in which the freedom of readers is not restricted, socially useful reprinting is possible and socially harmful reprinting put at a disadvantage.

Equalization as Problem

Interestingly enough, it were considerations like these -- although on a much simpler niveau and with other framework parameters -- that led to the creation of Copyright (and later authorship right) around 1476.

Copyright was introduced as a purely industrial regulation between publishers to regulate reprinting. Authors were only allowed to purchase rights in their works around 1710, something that the system was neither built for nor was it planned.

Unfortunately, our days are dominated by a mental doctrine that replaces the original sense of the system -- furthering creativity in a socially useful way -- by a pure perception of entitlement and possession. Limitation of access and monopolization of knowledge have ceased being means, they have become ends in themselves.

This mindset is contained in and furthered by the term "intellectual property", which despite its obvious non-sense -- what does it mean to own a thought -- is widely accepted without questioning and lumps together lots of very different issues. Among them are Copyrights and Patents as well as Trademarks and business methods. Each a very large and distinct area in its own right and worthy of differentiated consideration.

In fact it would be socially useful to take a close look at these areas individually, because -- to come back to Copyright/Authorship right -- why exactly do the instructions for my washing machine a protection of 70 years after the death of its author? With all due respect to German engineering, I do suspect that no washing machine in use today will still be functional in 100 years.

Also, technical instructions usually contain much less artistic expression than poetry. It is an obvious weakness of the system that both are considered and treated similarly.

But as long as we do not revise the system, Free Documentation has to be aware of these weaknesses. The licenses have to aim at reestablishing the balance that the system itself has lost.

GNU Free Documentation License (GFDL)

One attempt at finding a better balance is the GNU Free Documentation License (GFDL) [11] by the Free Software Foundation.

With it, identical reprints can be prevented by means of specifying an optional cover and back-cover text that cannot be reproduced identically without consent of the original publisher and has to appear in a "History" section of any reprint.

This allows readers to identify reprints because of the different title and also gives information about the original version. At the same time it protects the possibility to build upon and distribute the knowledge inside a book including its history.

Also it is possible to specify secondary sections which inform about the relationship between author, publisher and/or the general context of the book. These can also -- in order to protect possible artistic, personal or political statements of an author from being represented falsely or out of context -- be marked "invariant" and therefore exempted from the permission to modify.

Of course that invariance can cause problems in particular for technical documentation when used in a thoughtless or excessive way, since it obliges all future versions to not modify these sections -- including not to remove them.

As discussed extensively in the Debian project, this can lead to serious practical difficulties, since invariant sections also need to accompany short references of handbooks and when merging two handbooks into one, the invariant sections of both would have to be present in the result.

Sometimes a restriction on absolute freedom brings an increase in overall freedom for the majority -- this is also a guiding principle for the GNU General Public License (GPL), in which the freedoms are granted as long as the recipient respects the same freedoms of others.

If an author feels strongly to only make his or her works available provided that some accompanying notice is preserved, this is probably part of his or her personal freedom.

But authors should consider very carefully whether these statements have to go into that document as invariant section. The attempt of forcing opinions upon others hardly shows the intended effect and as the discussions within Debian show, acceptance for such measures is pretty low. So the direct result would be for such documentation to achieve much lower distribution.

Also for that reason the GNU Free Documentation License by default does not define any title or invariant sections. And even though it provides means for them, Network Theory for instance foregoes using them for some books at all.

Similar to the GNU General Public License and GNU Lesser General Public License, the GNU Free Documentation License also means for recipients of a Free Document to be informed about the freedoms provided as part of the document and to give them the possibility to excert them.

Even though this may seem impractical at times because of the rather long license text, it is essential to not only protect freedom on paper but also in reality.

Enough about Free Documentation for now and towards an initative From the more traditional corner which allows furthering Free Software by reading books -- be they free or non-free.

Bookzilla.de

The Hamburg-based company freiheit.com technologies gmbH has implemented the shopping portal for Libri.de, the online-daugther of the LIBRI consortium. The book wholesaler LIBRI delivers books to over 4000 bookstores in Germany as well as virtual stores and internet book-sellers.

Any bookstore can register as a partner and profit from internet orders via Libri.de. Also the new concept allows opening so-called affiliate shops. These allow modifying the themes and parts of the standard web layout to create a personalized online bookstore accessing the whole sortiment of Libri.de.

One such affiliate shop was now opened by freiheit.com itself under the name Bookzilla.de [13] with the goal of supporting the work of the FSF Europe for Free Software. The entire sales provision for every book sold through Bookzilla.de is donated to the FSF Europe. [14]

Complementaty to the idea of UKFSN to further Free Software by choosing the right ISP, this is another interesting possibility to indirectly support Free Software.

Enough

That should be enough for this issue. There would be news to report From the World Summit on the Information Society (WSIS), but that would exceed the time and space available. If interested, information can be found via the FSF Europe project page. [15]

So I can only wish you a good month and ask for numerous questions, comments, ideas, suggestions and project recommendations that can be sent to the usual address. [1]

Also I would like to point out the mailing lists [16] that have recently been created to make it easier for volunteers to get involved in the Brave GNU World -- especially for translations and the web pages. Without the help of so many people the Brave GNU World would not be possible.

Until next month.