#+TITLE: MediaGoblin's application to Knight News Challenge
#+AUTHOR: Christopher Allan Webber <cwebber@dustycloud.org>
#+EMAIL: cwebber@dustycloud.org
#+DATE: April 28, 2014

#+LATEX_CLASS: article
#+LATEX_CLASS_OPTIONS: [12pt]
#+LATEX_HEADER: \usepackage[margin=1in]{geometry}
#+LATEX_HEADER: \usepackage{hyperref}
#+OPTIONS: toc:nil


#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./explaining_federation.jpg]]
#+BEGIN_LATEX

\emph{
  MediaGoblin is a decentralized media publishing system.
  Users can share across nodes thanks to built-in federation
  tooling.}

\end{center}
#+END_LATEX

* Project summary

MediaGoblin is a decentralized and social media publishing system
focused on user freedom. It supports many different media types;
images, video, audio, 3D models and more. Instead of needing a
separate service for each one (YouTube + Flickr + SoundCloud + Vimeo +
Thingiverse), each controlled by a specific company with a walled-off
network, when you use MediaGoblin you need just one account -- and
it's yours to control. For maximum independence you can host your own
network node, or just create an account on a node you trust. It
archives all your media and lets you share with as large or small a
circle as you wish. MediaGoblin is free and open source software,
making it the censorship- and surveillance-resistant solution for
global media sharing.

** Detailed description


MediaGoblin is a highly extensible free and open source media
publishing system with a great plug-in architecture. This means that
its evolution can be user-driven, allowing it to be adapted to a great
variety of user purposes. Nodes focused on citizen journalism may have
different requirements than music publishing nodes. If a person,
project or business using MediaGoblin needs an extra feature, they
will always have the right to extend the software or hire someone to
extend it for them -- as long as they contribute the code back to the
community they build on.

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./video_media_type.jpg]]
#+BEGIN_LATEX

\emph{
  Thanks to MediaGoblin's built-in extensibility, many media types are
  supported.  Here we see the film "Caminandes: Gran Dillama" playing
  on a MediaGoblin node.}
\end{center}
#+END_LATEX

This extensibility is already evident in the areas of theming, media
types, and plugins providing additional features like embedded
OpenStreetMap support, and flexible authentication options including
OpenID and Persona. MediaGoblin's codebase and legal tooling make it
ideal for extension, and because it is written in Python, it is
accessible and easy to modify.

The internet is becoming a fragmentary landscape of competing but
incompatible services, with business models commonly designed to
exploit and endanger users rather than aid them. But it doesn't have
to be that way. MediaGoblin is designed for federation, the way email
is. Federation helps people talk across different systems. Imagine if
every email user of Blahoogle.com could only email other Blahoogle
users, and needed another subscription at CatMail to talk to users
there. That is what is happening across the media-rich social
web. MediaGoblin doesn't want to see the internet become a hopelessly
fractured and chaotic space where everyone has to keep up with a
zillion services and networks, always chasing their feature and social
needs. These services are not prioritizing the open standards and user
rights that protect the internet itself while keeping it an
egalitarian and safe zone to live in. Rather than cultivate its own
private commercial fiefdom that relies on preventing interoperability
with everyone else, MediaGoblin builds tools and communities that
benefit everyone. MediaGoblin is a powerful framework that people can
build on, benefit from, and share the benefits back to the common
good. We don't think people should have to trade their freedom for
usability. MediaGoblin is designed to offer people the same kind of
social media sharing features they find in their existing services,
but better: simpler, more powerful, connecting a global user base,
growing with every bit of user investment, and all while protecting
user freedom.

The call for an independent media publishing system that resists
censorship grows urgent as we see increasing cases of the centralized
commercial services exerting censorship and restricting content -- on
behalf of their own interests, and in collusion with regional
governments. The silos that are the main current points of media
sharing are not only vulnerable to attacks on free speech, they also
hamper important grassroots economic activity by privileging the
interests of a tiny minority over those of most of the world. As more
and more people conduct their lives online, the accumulation and
commodification of granular personal data presents a tremendous
encroachment on privacy, security and anonymity -- not just the
hallmarks of free society, but even the preconditions of our freedom
of thought.

** One-sentence summary

#+BEGIN_QUOTE
In ONE sentence, tell us about your project to strengthen the Internet
for free expression and innovation.
#+END_QUOTE

MediaGoblin builds a decentralized, social media publishing system,
and a set of tools any other project can use to be social /and/ protect
user freedom on the internet.

** Who will benefit?

#+BEGIN_QUOTE
Who will benefit from what you propose? What have you observed that
makes you think that?
#+END_QUOTE

Media authors -- whether artists, journalists, individual users or
organizations -- gain an innovative publishing system without
censorship, centralized surveillance, or corporate monopoly, on a
social platform the user can customize and control. In place of being
locked into centralized and vulnerable corporate media silos, users
gain the enormous benefit of a world-wide independent media sharing
network that can be extended to defy even national and language
boundaries.

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./jpope_instance.jpg]]
#+BEGIN_LATEX

\emph{
  MediaGoblin is flexible enough to fit the needs of whatever
  individual or organization is running it.  If the project does not
  yet contain the features you need, it can be easily extended.  Here
  a user has installed their own custom theme.
}

\end{center}
#+END_LATEX


** What progress have you made so far?

After one year of crowdfunded development, MediaGoblin has issued 5
major releases and built the groundwork of an ambitious system that
could reverse the trend of fragmentation into proprietary media silos,
and empower the future of user-driven media communications on the
web. Its active community of over 75 volunteers need to be augmented
by full-time developers in order to implement the next key targets of
MediaGoblin's roadmap such as federation and private media
sharing. MediaGoblin would greatly benefit from more resources and is
working hard to become a self-sustaining project, having launched
successful community-funded crowdfunding campaigns while providing
complete financial transparency related to the expenditure of those
funds.

** Successful outcomes

#+BEGIN_QUOTE
What would be a successful outcome for your idea or project? 
#+END_QUOTE

MediaGoblin's success can be measured by the following criteria being
met: complete and usable decentralized/federated media publishing
software, the presence of well-functioning privacy features and
private sharing, and growing public adoption with a multitude of
MediaGoblin nodes.

** Team and experience

#+BEGIN_QUOTE
Who is on your team, and what are their relevant experiences or
skills?
#+END_QUOTE

Christopher Allan Webber is lead developer of the MediaGoblin project
and provides the technical vision of the project, as well as the
goblin drawings. He previously has worked as tech lead at Creative
Commons and began development of Miro Community at the Participatory
Culture Foundation. Deb Nicholson is community manager and
additionally handles much of our communication. She has a background
in social justice, currently works at the Open Invention Network, and
previously worked for the Free Software Foundation. We also have a
team of over 75 contributors from around the world that do everything
from coding features to graphic design to documentation to
translations.

** Location

Madison, WI, USA

* Questions and answers

** Users and target customers

#+BEGIN_QUOTE
Who are the users or target customers of your project, and
what have you learned from them so far? Please give specific
examples.
#+END_QUOTE

MediaGoblin aims to be usable for anyone who is interested in
publishing media on the web.  However, we are particularly interested
in targeting 1) artists who have a wide portfolio of media types to
publish as social activists, 2) communities who are concerned about
censorship and surveillance, and 3) those who want a media hosting
platform that can be tailored to their individual needs.  We are also
interested in targeting galleries, libraries, archives, and museums.

MediaGoblin's early adopters are presently by fairly technically savvy
communities or those who are more keenly aware of issues of technical
freedom (including the Free Software Foundation hosting their
conference footage from LibrePlanet on their own
node[fn:libreplanet]).  However, we are working actively to expand
beyond those areas.  We have growing adoption in other spaces;
MediaGoblin is presently working on a small grant to make the project
more useful for archival and academic institutions, and we have seen
at least one university adopting MediaGoblin, as the Rochester
Institute of Technology is running their own node.[fn:rit]

# Our users provide regular feedback; many of them become parts of our
# community directly.  Additionally, we have participated in user
# experience testing (expanded upon below).  Our general sense based
# upon this input is that users generally appreciate and enjoy the
# features we provide, but wish us to fulfill the main things

[fn:libreplanet] [[http://media.libreplanet.org/]]

[fn:rit] [[http://yacht.rit.edu/mediagoblin/]]

** Assumptions and testing

#+BEGIN_QUOTE
What assumptions are you making in what you propose, and how
will you test them?
#+END_QUOTE

We assume that artists, musicians, and other authors or publishing
institutions of media are interested in a decentralized media
publishing system with the features MediaGoblin provides.  The large
amount of discourse on the internet and in the press over the last
year surrounding controversies of governmental and corporate
censorship and surveillance suggest that many individuals would like
to move to platforms that better support their freedoms, but feel as
if they have no place to go.  The best way to test these assumptions
are to build MediaGoblin, make it easily accessible to use and run,
while continuously seeking user feedback and talking to those in our
target groups.

To get a better sense of what areas we are lacking in, we have also
begun user experience testing sessions.  Recently we ran an event
called SpinachCon[fn:spinachcon] to collect user experience stories
from actual users.  We have collected this data and are working
towards acting on it.  The event was successful
enough[fn:spinachcon-success] that the free software community
diversity and outreach group OpenHatch[fn:openhatch] are looking to
take it on as a recurring event.

Software design involves making many decisions based on assumptions,
and MediaGoblin is no exception to this.  Fortunately, MediaGoblin has
the advantage of being transparent while encouraging user feedback,
having an open bug tracker, and fostering community participation.

[fn:spinachcon] [[http://mediagoblin.org/news/spinachcon-2014.html]]

[fn:spinachcon-success] [[http://openhatch.org/blog/2014/spinachcon-zero-a-huge-success/]]

[fn:openhatch] [[http://openhatch.org/]]

** Getting the project out there
#+BEGIN_QUOTE
How will you get your project in front of the necessary
people or organizations?
#+END_QUOTE

# Deb:
#   Asking everyone individually a million times and working with them to meet
#   their needs?

# <larjona> paroneayea: "How will you get your project in front of the necessary 
#           people or organizations?" > More visibility to the existing          
#           deployments + agreements with certain projects to help them the      
#           deployment and they become "living examples" of what can be done     
#           with GMG                                                     

#  - marketing
#    ^--- Morgan, says this is what they're looking for
#         explain what we mean by then

#  - MediaGoblin is free software so anyone can use and run it, and orgs
#    and individuals running it themselves is ideal
#  - Interest in connecting to nonprofits, activist groups, becoming a
#    resource to them.
#  - However this is not always so easy
#  - Exploring this as a path to sustainability

We intend to promote visibility of existing MediaGoblin deployments
and also work on agreements with high profile and related communities
and projects to assist them with MediaGoblin deployment.  Thus, we can
help build living examples of what can be done with MediaGoblin.

For those projects and individuals that can afford it and are
interested in using the software but perhaps not maintaining a running
node themselves, we are planning to offer paid hosting.  We are also
interested in this as a path toward sustainability of the project.

Additionally, MediaGoblin has put effort into building clear materials
to make what MediaGoblin is easy to understand for newcomers, as well
as the surrounding issues that MediaGoblin is attempting to address.
We intend to continue to put resources into these efforts.

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./censor_scan.jpg]]
#+BEGIN_LATEX

\emph{
  MediaGoblin's messaging team works to build materials both to market
  the project and also to explain the underlying issues.  This frame
  is from an animated video explaining the issues of automated censorship
  in a highly centralized environment.}

\end{center}
#+END_LATEX


** Addressing obstacles

#+BEGIN_QUOTE
What are the obstacles to implementing your idea, and how
will you address them?
#+END_QUOTE

*** Technical obstacles 

From a technical perspective, much of MediaGoblin already works, and
while there are certainly many technical areas that need polish,
refinement, and advancement, the core of MediaGoblin as a publishing
tool functions well.

There is, however, one major piece missing: it is not enough just to
"push media out there" in modern times with media publishing; users
are accustomed to many social features such as favoriting media,
adding media to collections, subscribing to channels and content, and
so on.  Adding these features to a decentralized environment is called
"federation" and is a significant technical undertaking, one that we
have begun to embark on but which will take significant effort to
complete in a user-friendly way.

We have taken the "standing on the shoulders of giants" approach and
are presently working towards implementing inter-site communication
via a protocol called the Pump API.[fn:pump-api] We are adopting this
to try to forge a future where many websites can be interoperable.  It
is possible that this assumption in protocol choice may not turn out
to be correct, though we do believe that trying to collaborate with
likeminded projects is the best way to move forward in advancing
federation on the web.  We also believe that if there are assumptions
in the protocol that are proven to not be ideal for a project such as
ours which is media-heavy, we can collaborate with those authoring the
specification.

[fn:pump-api] [[https://github.com/e14n/pump.io/blob/master/API.md]]

*** Social and adoption obstacles

Much of the issue of social challenges has already been addressed in
this proposal, but it is worth noting that for all truly decentralized
web services, gaining adoption is difficult because of the challenge
of network effects.  While media publishing systems have less of a
challenge than purely social networking sites,[fn:social] users have
come to expect the ability to interact with other users in relation to
said media.  If their family, friends and colleagues are not on the
same network as them, this transition can be difficult.  As such, we
would like to make MediaGoblin as easy to migrate to as possible by
providing easy transitional tools to migrate one's data from other
larger sites.  Making MediaGoblin easy to use, well featured, and well
marketed will all of course be necessary as well.

[fn:social] We say that MediaGoblin has less of a challenge than a
  more "pure" decentralized social networking system (such as one
  challenging Facebook) because the social network is not the only or
  even primary element of value being delivered, the media is.  It is
  easy to share a video or an image with friends even if they do not
  have an account on that site or network.

*** Community obstacles

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./mediagoblin_hackathon-scaled.jpg]]
#+BEGIN_LATEX

\emph{
  MediaGoblin has both strong leadership and a sizable community which
  is tightly integrated into the project.  Here a group of
  contributors are collaborating on improving the codebase at a
  hackathon at the GNU 30th event.
}

\end{center}
#+END_LATEX

One challenge that many free and open source software projects suffer
from is that once the project grows large, the community suffers, or
perhaps the primary organization funding the software chooses to make
it proprietary.  We have sought to avoid this issue from the very
beginning of the project.  For this reason, the project is under the
stewardship of a nonprofit (the Free Software Foundation) which has a
mission to preserve software freedom and maintain its integrity, and
which has a strong track record of doing so.  Additionally, we have
chosen the GNU Affero General Public License because of its
requirement to give back modifications to the community.  Since there
are many contributors to the software and no requirement for copyright
assignment, it would be very difficult for someone to lock down the
codebase by making it proprietary.

Despite these safeguards, community integrity remains on the forefront
of our thoughts and actions.  We have monthly contributor meetings,
project decisions are made out in the open in a public chatroom, on
mailing lists and on the bug tracker.  While MediaGoblin is a project
with strong leadership, it is also a project that emphasizes
both transparency in decision making and community.

We also actively encourage and recognize contributions of all
kinds... not just code, but also graphic design, documentation
writing, translations, marketing, and so on.  We work hard to make all
contributors feel welcome, no matter what their skills are.  We are
also committed to diversity and are proud to have participated in
projects such as the Outreach Program for Women, and will continue
being proactive in building a diverse and welcoming community.

** Project cost and expenses

#+BEGIN_QUOTE
How much do you think your project will cost, and what are
the major expenses?
#+END_QUOTE

The large majority of MediaGoblin's costs will go into development of
the software itself, which means paying developers and designers.  (We
have some advantage in that some of the administrative overhead can be
reduced through our fiscal sponsorship through the Free Software
Foundation.)  We can make use of any amount of funding we can get,
however the amount of funding brought in affects what kind of work we
will be able to do.  More details are given below in a range of three
potential funding levels.

*** Basic plan budget: $125,000

 - MediaGoblin as a well functioning project.  Full-time developer
   resource to focus on the programming of the project.
 - Community management of the project, including reviewing and
   stewarding community contributions.  This will be a full time
   management plus developer resource.

*** Polish MediaGoblin for everyone budget: $250,000

 - MediaGoblin as a polished, well functioning, well featured project.
   Multiple full-time developers to focus on the programming of the
   project.
 - Graphic design and user experience resource to make the project
   smooth, pleasant, and easy to use for everyone.
 - Community management of the project, including reviewing and
   stewarding community contributions.  This will be a full time
   management plus developer resource.
 - Actively reach out to interested institutions and assist them with
   adopting MediaGoblin.
 - Good marketing materials explaining MediaGoblin and why it matters
 - Exploring MediaGoblin funding sustainability by partnering through
   interested resource to provide MediaGoblin as a service.


*** Strong sustainability budget: $500,000

 - MediaGoblin as an amazingly polished, well functioning, well
   featured project.  Multiple full-time developers to focus on the
   programming of the project.
 - Graphic design and user experience resource to make the project
   smooth, pleasant, and easy to use for everyone.
 - Community management of the project, including reviewing and
   stewarding community contributions.  This will be a dedicated and
   full time resource.
 - Actively reach out to interested institutions and assist them with
   adopting MediaGoblin.
 - Actively working to bring the general public into MediaGoblin
   adoption.
 - Strong marketing materials explaining MediaGoblin and why it matters.
 - Multiple paths to exploring MediaGoblin sustainability and a
   resource focused on developing future paths to sustainability and
   self-sufficiency.
 - Enough of a funding buffer to ensure that MediaGoblin has time to
   properly develop to the point of achieving sustainability.

*** A note on financial transparency

#+BEGIN_LATEX
\begin{center}
\includegraphics[scale=.7]{./finances_pie_chart.jpg}

\emph{Summary of MediaGoblin expenses in 2013.}

\end{center}
#+END_LATEX

Note that it is already possible to get a sense of how MediaGoblin
spends its funds.  We have done a
full financial transparency report[fn:financial-transparency]
on past expenses, and we intend to continue in this direction.

[fn:financial-transparency] [[http://mediagoblin.org/news/financial-transparency.html]]

** Data collection and implementation

#+BEGIN_QUOTE
Who will do the data collection described in your proposal?
#+END_QUOTE

# Deb:
#   Hmm, sounds like we should do some data collection in the assumption
#   testing area then. How about we pay someone?

#  - Talk about programming, design
#  - Majority of data collection is actually providing a platform so
#    that users themselves have a way to gather and collect data.
#  - As for data, user experience testing to see if we're meeting usability
#  - We are willing to engage on other data collection tasks if useful

# Morgan: they want people who have a sustainable project... in order
#   to make this a viable business we'll have to do more market research

The majority of the project's data is actually user-submitted data,
i.e. the media that users are publishing to their MediaGoblin nodes.
Thus the vast majority of development is aimed at making the software
work well so users can do this.

Additionally, we do data collection in the sense of bug tracking, our
community wiki, and a development process which allows for public
contribution.  This helps us determine what we are currently being
successful at and what we are lacking in.

We are also interested in both user experience testing (which we have
begun doing through SpinachCon) and market research to better
understand how users react to MediaGoblin and how to improve the
software to those ends.  If we receive funding, we will allocate some
resources towards such projects.

** Learning from projects

#+BEGIN_QUOTE
What have you learned from other projects in this space that
informs your approach?
#+END_QUOTE

# Deb:
#   Feel free to speak broadly and include my experience and connections here.

# Miro and the Participatory Culture Foundation, knight funded project
#   miro community
# Deb's connection w/ OpenHatch
# Bring up CC and FSF connections, including Liberated Pixel Cup

# First of all, MediaGoblin has the advantage of being founded by
# individuals with extensive experience in the free and open source
# software space.  Chris and Deb's past experiences in organizations
# working in a similar space have greatly influenced the project.  (Work
# on Miro Community helped partly inspire MediaGoblin from the feeling
# that aggregation 

# Does this sound condescending?  Definitely not the goal.

# The early ideas of MediaGoblin itself partly came out of Chris
# Webber's own thinking around the issues that Miro Community succeeded
# at (software that would allow communities to pool together video
# resources), but also the large issues that as a media aggregator it
# could not solven on its own (communities are still bound to the whims
# of large, proprietery media silos, which can be especially hazardous
# to activist groups suffering from censorship).

# Also talk about the existing foss social media projects that have failed
#   Why did they fail or are struggling, how will mediagoblin avoid that?

One lesson we have taken in observing other projects is in the
construction of our community.  When Diaspora first made its
announcement during its initial crowdfunding campaign, it did a
successful job of capturing the hearts and minds of many people.
Unfortunately, the developers decided to do development privately and
not until later did they open up to show their current progress.  We
feel this was a mistake: there was an opportunity lost to build a
community from the excitement of this group of people.  Instead when
Diaspora's code was first opened, there was much frustration and
confusion in that the codebase was not as well developed as many had
anticipated.[fn:diaspora-launch] Diaspora now has a much stronger
community, but there could have been peer review by domain experts and
an opportunity to build a community if all code had been built
publicly from day one.  For this reason MediaGoblin has always had a
public and transparent development process.


#  Not enough to just rely on external services
#  - Miro Community

#  Interoperability and marketing
#  - StatusNet (now GNU Social) to Pump.io
#    - StatusNet originally developing protocol OStatus 
#    - Doing the right thing as in terms of building a
#    - Protocol now much easier for developers to understand, but hard
#      for users.
#    - Hence, our approach is to build off of the hard work of the
#      easy to understand for developers pump API

We have also learned by paying close attention to what has happened
between the decentralized social media systems StatusNet and Pump.io
(which implement and are the main examples of the federation protocols
OStatus and the Pump API, which we are using for federation,
respectively).  StatusNet achieved a significant amount of success as
in terms of user popularity.  The project was mostly fairly easy to
use from a end user perspective.  However, most users stayed on one
very large server hosted at identi.ca, and there was not much success
with true federation in this sense.  Additionally, the federation
protocol OStatus, while designed to be generic, was not well adopted
by other software projects, partly because it was fairly difficult to
understand, being cobbled together from multiple existing standards.
When efforts transitioned to a new project named Pump.io, the new
software instead switched from a user focus to a developer focus.  The
Pump API is much easier to understand for developers, and the goal of
Pump.io is to become infrastructure to make developing federated
software easier, but the software is harder for end users to make use
of.  Thus, we are taking the approach of making use of the
developer-friendly Pump API and intelligently incorporating these
technical efforts into our software to simplify efforts around
decentralization technology decisions, but much of our focus will be
around making these features easy to use for end users.

Lastly, MediaGoblin is itself founded by individuals who have worked
extensively in many of these spaces.  Chris Webber's experience as
tech lead at Creative Commons and work on Miro Community and Deb
Nicholson's employment history at places such as the Free Software
Foundation and the Open Invention Network, coupled with years of
community activism and nonprofit work, have greatly informed the
direction and mission of the project.  We work hard to incorporate the
skills and experiences of our broad community, and stay in close
contact with other projects at the forefront of the fight to keep the
web decentralized.

Our community is strong, our project is growing, but it lacks
significant resources.  We believe with the support of the Knight
Foundation we can bring MediaGoblin to the next level.  Together, we
believe we can bring freedom back to media publishing on the web.

[fn:diaspora-launch] See
  "'Open Source' is not a Verb; thoughts on Diaspora" at
  [[http://yergler.net/blog/2010/09/23/diaspora-community/]] for a good
  summary of this situation.


* COMMENT Additional media

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./audio_media_type.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./mediagoblin_profile.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./jpope_train.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./presentation2.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./infected_node.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX

#+BEGIN_LATEX
\begin{center}
#+END_LATEX
[[file:./surveillance1.jpg]]
#+BEGIN_LATEX
\end{center}
#+END_LATEX