XMP namespace

An XMP namespace is a stable URI that acts as a unique identifier for a vocabulary of metadata properties. In RDF/XML, properties are written with a prefix — like dc:creator or llmind:entities — and that prefix is bound to a namespace URI in the document. The URI is what makes the prefix unambiguous globally.

What a namespace is

Imagine two different organizations both define a property called author. Without namespaces, they would collide. With namespaces, one organization uses org-a:author (bound to http://org-a.com/ns/) and the other uses org-b:author (bound to http://org-b.com/ns/). The same file can contain both without ambiguity. Namespaces are how RDF solves the naming problem.

URI stability requirement

The namespace URI must never change or every consumer breaks. That's why LLMind's namespace is versioned: https://llmind.org/ns/1.0/. If a future major-version-incompatible change is needed, a new URI — https://llmind.org/ns/2.0/ — is created. Old files can still be read with the old namespace; new consumers opt in to the new one.

Prefix conventions

Prefixes are local to each file — you can bind whatever prefix you want to a namespace URI. However, some prefixes have de-facto conventions: dc: for Dublin Core, xmp: for Adobe XMP, xmpRights: for rights, llmind: for LLMind. Tools and humans recognize these conventions, making metadata readable and navigable.