Replace vague owl:Thing ranges with specific types

[melvincarvalho]

Severity: WARNING

Problem

Three properties use owl:Thing as their range, which is too generic for proper validation and type checking:

Affected properties:

{
  "@id": "nostr:tags",
  "rdfs:range": { "@id": "owl:Thing" }
},
{
  "@id": "nostr:profile",
  "rdfs:range": { "@id": "owl:Thing" }
},
{
  "@id": "nostr:follows",
  "rdfs:range": { "@id": "owl:Thing" }
}

Why This is Problematic

owl:Thing is the most generic type in OWL - it means "any RDF resource." This provides:

• No structural validation
• No type checking
• No schema documentation value
• No tooling support for data validation

What Should Be Used Instead

Each property has a specific structure in Nostr:

1. tags (Array of Arrays)

"tags": [
  ["e", "event-id", "relay-url"],
  ["p", "pubkey-hex"],
  ["t", "hashtag"]
]

Proposed fix:

{
  "@id": "nostr:tags",
  "@type": "rdfs:Property",
  "rdfs:comment": "Event tags as an array of arrays. Each tag is an array where the first element is the tag type.",
  "rdfs:range": { "@id": "nostr:TagArray" }
}

Define nostr:TagArray:

{
  "@id": "nostr:TagArray",
  "@type": "rdfs:Class",
  "rdfs:comment": "An array of tag arrays, where each tag is itself an array of strings",
  "rdfs:subClassOf": { "@id": "rdfs:Resource" }
}

2. profile (Kind 0 Event Object)

{
  "name": "username",
  "about": "bio",
  "picture": "https://...",
  ...
}

Proposed fix:

{
  "@id": "nostr:profile",
  "@type": "rdfs:Property",
  "rdfs:comment": "Profile metadata from a kind 0 event",
  "rdfs:range": { "@id": "nostr:ProfileMetadata" }
}

Define nostr:ProfileMetadata:

{
  "@id": "nostr:ProfileMetadata",
  "@type": "rdfs:Class",
  "rdfs:comment": "Structured profile metadata containing name, about, picture, etc.",
  "rdfs:subClassOf": { "@id": "rdfs:Resource" }
}

3. follows (Array of Pubkeys)

["pubkey1", "pubkey2", "pubkey3"]

Proposed fix:

{
  "@id": "nostr:follows",
  "@type": "rdfs:Property",
  "rdfs:comment": "Array of pubkey hex strings representing followed accounts (from kind 3 events)",
  "rdfs:range": { "@id": "nostr:PubkeyList" }
}

Define nostr:PubkeyList:

{
  "@id": "nostr:PubkeyList",
  "@type": "rdfs:Class",
  "rdfs:comment": "An array of 64-character lowercase hex strings representing Nostr public keys",
  "rdfs:subClassOf": { "@id": "rdfs:Resource" }
}

Benefits of Fixing

1. Type safety: Validators can check structure
2. Documentation: Developers understand expected format
3. Tooling: IDEs and validators can provide better support
4. Interoperability: Standard RDF tools can reason about types
5. Future-proofing: Easier to extend with subclasses

Implementation Notes

If you want to keep it simpler without defining new classes, you could use:

• JSON Schema types (via @type: "@json" in JSON-LD)
• More specific XSD types (xsd:array if available)
• Or simply better documentation in rdfs:comment

But defining specific classes is the RDF/OWL best practice.

Priority

MEDIUM-HIGH - Doesn't break functionality but limits schema usefulness for validation and tooling.