Hi Gary,
Gary Lyons wrote:
I am considering writing an overview of relational database theory.
I'm glad to hear that you want to jump in and contribute.
The concern I have is that there is an abundance of published material
on the subject--a Google search turned up ~16,000 hits. I have read
quite about about the subject and I can't be sure that I wouldn't
unintentionally repeat some of it.
Is it enough to provide a brief history and expand on the bits that
would be of interest to the audience? After all, it *is* a well known
theory, and pretty rock solid to boot. Most of the differences in the
documentation I've seen revolve around implementation of the various
database engines, and I hope to steer clear of that. But I will still
need to use some sort of reference material.
What kind of document are you hoping to write? Would this be a chapter
in the OOo Database Guide? It sounds like you want it to be as generic
as possible. Would there be parts that are specific to OOo? What aspect
would make your document unique?
It would make me more comfortable to find a couple of good sources and
paraphrase or quote from them. Would it be enough to provide
bibliographic references, or should I get the Author/Copyright holders
permission and list them in the acknowledgments section?
how would this reconcile with the PDL?
This very much depends on the source of the material and how it is
licensed.
If the source material has a "straight" copyright, quoting or
paraphrasing more than a brief section would be a copyright violation.
You would definitely need to get the author's permission to use any
substantial portion. And they might or might not be OK with you
publishing it under our PDL/GPL/CC license, which potentially could
allow unknown third parties to also re-publish it. An author who is not
in the open source world might not feel comfortable with that. And if
they say no, you can't use their work. End of story.
If the material is published under some sort of open source license,
then you would need to abide by the terms of that specific license. The
license in question might preclude the triple-license option we are
currently using for the OOo Guides. So that would still be an obstacle.
Assuming that the author grants permission to use their work in an OOo
Guide, either implicitly through a compatible license or explicitly to
you personally, I think you should list them in the Authors section of
the chapter, to abide by our license(s). It would be good form to also
list them in Acknowledgements and provide references. At least that's my
take. I am not a lawyer. We recently published a chapter in which
sections were based on a Linux Journal article by Bruce Byfield, but he
was not listed as an author. JEAN: what was the reasoning there?
You can avoid all this hassle by avoiding quoting or closely
paraphrasing your source material. You can, of course, refer to existing
sources. But you should try to express the concepts in your own way.
If you find that there some really good existing sources, you might want
to simply write a sort of annotated bibliography that directs the reader
to those sources, with your comments about why they're good. I don't
mean literally a bibliography organized by author name, but rather an
overview of the concepts, with references to the best sources of
information about each concept.
I don't mean to come across as discouraging. On the contrary, I want to
encourage you to create your own content rather than mimic someone else's.
Have you had a chance to review some of the chapters we're working on?
That would give you an idea of the audience, tone, and level of detail
we're trying to address.
--
Janet Swisher --- Senior Technical Writer
Enthought, Inc. http://www.enthought.com
