handling attributions for Docs

Eric Christensen eric at christensenplace.us
Mon Nov 16 02:47:07 UTC 2009


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, Nov 16, 2009 at 11:26:09AM +1000, Christopher Curran wrote:
 > Karsten Wade wrote:
 > >Now we have a clear need for attributing source information: thanks to
 > >being relicensed under CC BY SA 3.0 means we can import lots of good
 > >content and so we're likely to do so.  Efficient attribution is now
 > >important.
 > >
 > >There are two parts to this discussion:
 > >
 > >* Technical - how it works in DocBook and Publican
 > >* Brand - how does the Fedora Documentation Project want to attribute
 > >
 > >I was going to look at it from a social angle, but I think the brand
 > >angle overrides the social angle.  Read on for more.
 > >
 > >== Technical proposal ==
 > >
 > >To handle copyright attribution, I recommend we adopt this approach:
 > >
 > >* For single or a few imports of a chunk of attributable content, use
 > >  a <footnote>.  Attribution happens on the page it occurs.
 > >
 > >  Example: Pulling a description of AES encryption from Wikipedia for
 > >  the Fedora Security Guide.
 > >
 > Footnotes come from one school of attribution. There is another
 > school of thought that preaches endnotes. I would recommend neither
 > and suggest we go with a page of attribution and thanks were we
 > mention various voluntary and 'acquired' contributions. This gives
 > us a wider scope for referencing conversations, technical input,
 > code, and other miscellaneous contribution. Useful as we will be
 > without a formal and ridged format which would normally exclude
 > those types of contributions.

While it is great to say "we used all these references and people to
make this guide" I don't think that meets the 1) letter of the license
or 2) the spirit of sharing.  If the text we use comes from someone else
who says to use it but attribute it back we should do that.  We would
need some way to point from the text to the attribution so it is proper.

 > >* For longer imports, blends, or remixing of content, use
 > >  <legalnotice>.
 > >
 > >  Example: All of the people who work on a guide over the years would
 > >  be lists in a standard format under the primary Fedora/Red Hat legal
 > >  language.
 > >
 > That isn't relevant, really. If something is exceedingly, heavily
 > rewritten it counts as an original work. Credit where credit is
 > due, of course. However, if the original is unrecognisable all it
 > does is make notices longer and unwieldy.

Again, we have to follow the letter of the license.  We would want
others to do the same.

 > >For the <legalnotice> usage, we would need:
 > >
 > >* A standard format for all attributions, to make it fair, clear,
 > >  equitable.  Alphabetical, for example.
 > Always by importance. Alphabetical is rubbish for translations and
 > other languages.

How do you determine importance?

 > >* A change to Publican(?) to look for and use a file,
 > >  en-US/Attribution.xml, if it is present.  This allows attribution to
 > >  be kept within the main document source tree.
 > +1
 > >== Brand proposal ==
 > >
 > >This is a proposal only affecting Fedora-branded works.  An upstream,
 > >such as the "Linux Security Guide"
 > >(https://fedorahosted.org/securityguide/), can attribute as it sees
 > >fit, just as a downstream "Red Hat Enteprise Linux Security Guide" can
 > >attribute as it sees fit.
 > >
 > >Currently, for some works, we have primary authors in a long list on
 > >the front cover of a work.  We've long discussed swapping that for
 > >"Fedora Documentation Project".
 > >
 > In all writing projects there is always one to three actual
 > writers. Primary writers should be up the front and secondary,
 > tertiary and other contributors up the back/front in a
 > references/attribution section.

Maybe but there are times that distinction is always clear.

 > >Especially as we work with a larger group, the list of authors on the
 > >front cover grows.  It visually competes with the Fedora branding.
 > >
 > And we should note that an important part of Open Source is
 > personal gain. People's names are their brand. Most people involved
 > in Open Source want their name present in patches and so forth as
 > they are trying to grow their own brands. This should be
 > encouraged, not hidden away. It brings personal pride to the work
 > involved and enforces accountability.

I don't think anyone is suggesting that we hide contributor's names in
the metadata somewhere.  But we aren't building these guides as a
billboard, either.

 > >In other parts of the Fedora Project, we don't see the authors
 > >presented in that fashion.  Anaconda is "Anaconda Team
 > ><anaconda-devel at redhat.com>".  This is more the norm for FLOSS
 > >projects.
 > >
 > There is often an AUTHORS page or similar included in the code or
 > the program. Man pages are full of personal attributions. We should
 > no abstract or removed this for previously mentioned reasons.
 > >This is how a "Fedora Docs Team" focus looks in practice:
 > >
 > >http://docs.fedoraproject.org/release-notes/f12/en-US/html/
 > >
 > >A good bonus to that would be having either a link to the project page
 > >or text inline that links to the mailing list or project page.
 > >
 > >Another idea is to have a notice in the authorship/editorship section,
 > >"For full attribution of contributions to this work, refer to the
 > >[legalnotice]."
 > I agree, but we should keep a focus on primary contributors.
 > >One reason for picking a standard is to set the expectation for how we
 > >attribute under the Fedora brand.  When we put up individual names, it
 > >creates a competitive space.  External content originators that are
 > >remixed may demand front-page attribution.  This causes the visual
 > >appeal to diminish while increasing the attribution maintenance.
 > >
 > If the remix is sufficient, we can give them a middle finger and
 > suggest they take a hike. The world of publishing and technical
 > documentation is nothing like the world of academic writing.
 > Attribution is handled differently. If a work is sufficiently
 > distinct it counts as an original work. Only original or patented
 > ideas /must/ be attributed. I'm not saying we act like jerks to the
 > community, I am just saying we don't have to comply when they act
 > like jerks.

Again we have to follow the license.  Personally, I've never run into
someone that was demanding, yet, but if I did I'd simply remove their
text from the document and move on.  But I think we still have to
attribute back to the original even if we modified the work afterwards
due to the original license.  Now a complete rewrite may be different.

 > 
 > I think we want to create a competitive space. We want to work as a
 > team but really bring credit to our standout players. We want to
 > put them out there and say, "This person is really great". That may
 > change from version to version and that is a Good Thing.

I think the end result is having a product that is up to date and
helpful.  People come to our guides because they have questions.  We
should want to help answer those questions.  If you are in it just to
have your name in lights go get yourself a blog.
  
 > 
 > >Having _all_ Fedora-branded guides follow the same standard that puts
 > >the Fedora brand first does the best service to the Fedora Project.
 > >It gives us the least headaches.  I think it is the right thing to do.
 > >
 > >What do you think?
 > >
 > >- Karsten
 > IANAL
 > 
 > Chris
 > 

- --Eric
- ---end quoted text---
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iEYEARECAAYFAksAvSoACgkQfQTSQL0MFMEeygCgjZANfLx1lyxftEB9/V5PHwga
EjAAoMqICEI8Boc+phkXE4JDSHg4Os3D
=6Tj2
-----END PGP SIGNATURE-----




More information about the fedora-docs-list mailing list