FridaysForFutue, WeChange, Transition Towns or Extinction Rebellion: they all have their own maps, calendars and news – but not

 

FridaysForFutue, WeChange, Transition Towns or Extinction Rebellion: they all have their own maps, calendars and news – but not synchronised. The goal of the FairSync project is to simplify discovery and search of activity hubs, events, and activity feeds. The target group of our efforts are platform integrators, which run maps with locations, host events and offer news feeds. The demand is huge and of priority at fairmove.IT, a network of cloud and IT solution providers servicing climate change and sustainability organisations and activists.

System integrators are faced with fast evolving APIs and protocols when they try to discover and connect systems and mae search more easy. We will develop a generic master-master replication framework of metadata enriched data sets on top of the Activity Pub protocol. This way we would like to overcome the federation solutions available on the level of database replication only or specialised for file or message synchronisation. We will be looking at more flexible propagation mechanisms and add conflict resolution schemes.

For secure and un-biased search and discovery we would like to leave the harvesting-by-bots path commonly used with search engines. Instead, we are aiming at replicating public data with federating platforms and let the providers of each platform use their built-in search and navigation tool. This way, we will gain reliability and independence. The data sets need to be up-to-date (not old, not doubled). We wil try to achieve this with a lazy check mechanism in the replication procedure.

All meta data enriched map projects we have seen so far (such as transformap, yellowOSM, kartevonmorgen, TREEDAY …) may offer an API, but not a concept for master-master replication. Happily the consortium of GIP (the Austrian geographical maps consortium) has know-how in aligning maps with several data sources. They are interested to share experiences with OpenLR and semantic supported conflict resolution mechanisms in this project, so we will be interviewing them and let them test our methods.

Several platforms allow replication, but usually they are to be configured in a master-slave style, such as Mongo DB replica sets. But what our partners and the users want in the application scenarios sketched in this proposal is a non-hierarchical, friction free access to data sets, no matter where they are stored, and being able to edit them (if the permission framework allows to do so). When you look around you find solutions like StrongLoop by IBM, an Open Source JavaScript library, but it is limited to connect mobile clients reliably to a master data source. Finding a more generic approach to this cloud computing challenge requires some more prototyping and testing.

Various concepts have been discussed  since 2017. The swiss initiative Transition Connect followed a hierarchical approach but got stuck in a Proof-of-Concept phase. Transition initiatives have been planning a solution in fairkom’s gitlab which is close to the FairSync approach. We have agreed in summer 2019 to collaborate and if the FairSync funding is successful we will rename the project.