O'Reilly's German Svnbook and our efforts...

C. Michael Pilato cmpilato at red-bean.com
Sun May 31 07:17:00 CDT 2009

Ben Collins-Sussman wrote:
> On Fri, May 22, 2009 at 6:53 AM, Karl Heinz Marbaise <khmarbaise at gmx.de> wrote:
>> I think the German part of O'Reilly has not understand the basics of Open
>> source and the basic idea of this kind of projects.
> It's not just the German part of O'Reilly, it's all of O'Reilly.
> O'Reilly produces excellent computer books, but ever since we
> published the 1st edition of the book with them in 2005, I was shocked
> to discover that they are a "conventional" book publishing company in
> every way.  The editors who work there aren't any more technically
> advanced than editors working on fiction books.
> *  They use Microsoft Word, proprietary layout systems like
> FrameMaker, and have *no* idea how to use version control themselves.
> *  They tried to get us to assign 100% copyright to them, because
> that's the norm in the book industry.  We instead had to persuade them
> to modify the generic contract such that the authors held onto the
> copyright, and to use the Creative Commons license.
> *  They completely ignore the online edition of the book:  they
> operate as if whatever text we send to them is the only copy of the
> book in the world.  Even though we've explained that the website runs
> in parallel (and is always more up-to-date), they still believe they
> have the "master" fork of the book.  That's why they publish "errata"
> on their website, and completely ignore our open source repository.
> *  As with all of their books, they hire their own translators and
> publish their own translations.  They are completely uninterested in
> our opensource efforts.  We simply do not exist.
> Really, the *only* thing that they do differently than a generic book
> publisher is that their lawyers "tolerate" open source licenses on
> books.   All their other processes are traditional and are exactly the
> way publishers have been behaving for decades.

Actually, I think some defense of O'Reilly is due here.  Things have changed
some since 2004 when we put out the first printed edition of the book.  I
handled most of the authors' interactions with them during our second
edition work, so it's reasonable that Ben might have forgotten how far along
they've actually come.

O'Reilly's copyeditors, indexers, tools team, etc. still use some
proprietary software, I believe, but DocBook XML seems to be the One True
File Format -- or at least, anything given them in DocBook XML can remain
there for the duration of the production process without any glitches.  But
what's cooler is that they use Subversion to house their work now!  In fact,
they gave me commit access to their repository so I could merge edit changes
back and forth between our repository and theirs during the latter phases of
the second edition wrap-up.  (Which would have been super-easy thanks to
Subversion 1.5's foreign repository merge support if their tools didn't
constantly rewrap and reformat the DocBook structures.)

These changes still fall short of some ideal, where perhaps O'Reilly's
employees and contractors interact with an open source community to produce
a high-quality manuscript in the public's eye.  Wow ... that would be truly
something to experience!  But, the changes in their process made between our
two editions were still pretty remarkable, in my opinion.  And this makes me
optimistic that there's still room for -- and interest in -- improvement.
Clearly O'Reilly has already come to terms with the fact that in many cases
(such as ours), they aren't making money on unique content.  The power is in
the packaging, or something like that.  If we could convince them to extend
that mindset to the translation phases, I think we could see outright
donation of translated content back to the open communities.

C. Michael Pilato <cmpilato at red-bean.com> | http://cmpilato.blogspot.com/

More information about the svnbook-dev mailing list