Social-Context

This is a sophisticated AD4M Language for link sharing in Neighbourhoods which has the following AD4M type signature.

AD4M Links are a shape much like an RDF triple. This shape allows us to associate a given subject to a given target with a predicate as another data point in this link. From such a structure you can start to associate pieces of data (in this case AD4M expression references) to each other and begin encoding application logic.
The ad4m linkLanguage provides the following interface for querying/adding links: LinkLanguage.

The ad4m language defined here will call the included holochain DNA (found in hc-dna). This holochain DNA is configurable based on the dna properties that are passed when being installed (see hc-dna/workdir/*.yaml) for some examples. Overview of the DNA properties and what they do are as follows:

• enforce_spam_limit: number of links a given agent should be able to make per max_chunk_interval before failing validation due to spam protection.
  
• max_chunk_interval: This value if read by holochain-time-index and is used to determine the depth of the time tree that is generated in order to index links under a given time index.
  
• active_agent_duration_s: Length of time that an agent is considered online for after adding an active_agent link pointing to their agent address.
  
• enable_signals: Determines if holochain signals should be sent to active_agent(s) when adding a link
  
• enable_time_index: Determines if links should be added to a time index that makes links queryable between time bounds, see LinkQuery
  
• index_strategy: Determines what values from the triple are indexed and thus queryable in the future. Options are FullWithWildCard, Full & Simple. Full with wildcard will make links discoverable by subject, predicate, target & *. Full will make discoverable by subject, predicate, target. Simple by only discoverable subject. It's an input parameter of add_link zome external function and may be different for each call.
  

How is this used in Junto?

Neighbourhoods (and this ad4m language as the backbone) are used in the Junto communities app to represent a community and more specifically share links agents create on a community with each other.
Via the ad4m-executor we can create neighbourhoods + add links into the neighbourhood to be retreived by agents and used to construct application logic/experience.

The source, target, predicate values found in these links are usually ad4m expression references in the form: languageHash://expressionId and resolvable by the given ad4m language to expression objects. An example of a link you may find in a junto community neighbourhood may look like:

{
    source: "neighbourhood://neighbourhoodHash",
    predicate: "hasPost",
    target: "shortFormLanguageHash://expressionHash"
}

Here the target would point to an expression found in the ShortForm language and be resolvable to an object found there. The hasPost predicate tag tells us this is a post on the neighbourhood found at source: neighbourhood://neighbourhoodHash.

Development

Build

npm i
npm run build

This builds the Holochain DNA first, using nix-shell (which needs to be installed) and then packages the whole language including the DNA in build/bundle.js.

Test

Run command in nix-shell,

make test

Add RUST_LOG=debug TRYORAMA_LOG_LEVEL=debug to test commands in package.json to view debug logs.