Potential Media Network / Baobáxia
A joint proposal by the Synap.tk Free Technology Kollective, Rede Mocambos and Metareciclagem* addressing the digital divide, and connecting communities lacking communication infrastructure.
The Myth of "Always On"
Although the new dominant paradigm of online social networking is total connectivity, where users are constantly online, reality begs to differ.
Eventually-Connected Federated Network
Potential Media Network/Baobáxia involves local communities that are not necessarily online at all times: Internet connectivity can happen once-a-day or less, over a low-bandwidth link such as a satellite connection, or completely offline with a hard-disk being shared among villages. In order to address this situation, we are thinking of using git-annex to share meta-data about git repositories of media that can be synchronized over time. E.g., a user would have access to the list of files, and could view or use the files that are already there, or queue the ones that are not yet available for later download. That is, it would be like using the Internet with a lot of lag between nodes.
The Quilombos form our primary target community, as they have an actual and immediate need for the technologies we want to implement, and offer as well a unique use-case in social networking: "eventually-connected federated networking".
Potential = not (yet) actual.
Potential Media Sharing
When dealing with sharing information with low or no bandwidth, and arbitrary long delays...
Potential Media Browser
Using git-annex, we want to make available the meta-data of potentially-shared media: using a well-known interface, users can browse a catalog of shared audio files, videos, archives, etc. When the chosen media is available locally, it is displayed. Otherwise, it is queued for download according to local connectivity and rules. In some case it could be immediate, in others, added to the daily batch. In the worst case, it could be "ordered for synchronization" by a traveler who will eventually get access to the original media in a better connected place. In any case, the potential media browser breaks the expectation of "unity of time" by allowing options and futures on exchange of valuable data. Methodology The main goal of the project is create solid self-managed infrastructure through real technological appropriation. In order to get there we need to create a step by step development, where people actually undestand, and participate to a step before taking the next one.
In the first step, we setup a local media server to host a central media repository for the community, to collect data in the field, that is actually used and shared by members of the local community.
- Create doc and tools to install a solid local server
- 3 hardware prototypes: common machine, more robust server, embedded/micro server (including pratical info like costs, where to buy, etc)
- Debian/GNU/Linux support with a step by step installation tutorial
- create packages, if necessary, for the main software
- eventually create a Debian Pure Blend for easy distribution of the bundled software
- define specification for basic organization of media files
- define some policies/pattern for filenames/foldername (for example "20120800_Campinas_VisitaHellekin" could be the name of a folder that provide cronological and location index and some kind of identification directly on folder organization)
- Provide some basic access to data through ZASF.
In a second step, we analyze collected data to research an optimal organization for indices.
- Refine indices (starting from folder organization) and propose some more advanced organization that should be supported by the system.
Make Data Scale
In a third step, we organize available data into git and git-annex repositories, so they can be shared across space and time, recording history of their production and evolution.
- Create one or more git/git-annex repositories to insert the collected media files (Potential Media Repository)
- Define what contents can be/should synchronized (Potential Media Preview)
- Automate common procedures (ex. create thumbnail and preview that can be easily synchronized, define metadata to allow local browsing in a catalog that does not necessarily provide access to the actual data.)
The fourth step concerns basic synchronization between distinct nodes across the network, using direct access or offline, asynchronous sharing.
- Define basic map of local and online server to synchronize contents
- Automate synchronization across the network (at specified time, when network is idle, etc.), through media (HD) or through movable micro-server (mule, bus, car, spaceship...)
Applications and Interfaces
Finally, the fifth step creates modular applications with really usable interfaces to the system. Usability here is critical: it will condition the adoption of the product.
- Create a web app/interface to publish and manage contents on local server (Potential Media Web Manager)
- Create a standalone app/interface to publish and manage contents on local machine (Potential Media Manager)
- Create plugins/API/app to browse and use media in the browser, on the desktop, on mobile devices (Potential Media Browser)
Annex -- The Plan
In the long term, we're aiming at creating autonomous local and public communication infrastructure including Internet connectivity, telephony, collaboration tools (lorea plataform), 3D printing, fabrication, etc.
Annex -- Resources:
: Synap.tk / Free Technology Kollective is an informal collective of people working in organizations promoting free software, hardware and culture in Latin America | Rede Mocambos is a network of mainly brazilian Quilombos communities | Metareciclagem is a network of groups and people that works on appropriation and hacking of technologies in a decentralized and open way, promoting social transformation.