At this moment there are 2 different client algorithms specified with the ordered mode and the unordered mode. I think it would make sense to have this in 2 separate documents: a simple and straightforward to implement document for the simple use cases, and a more elaborate specification for ordered traversal. This would probably make it much clearer and simpler on how to implement this, and would probably also open a clearer avenue to in the future evolve the client algorithms to for example RDF-Message based designs.
At this moment there are 2 different client algorithms specified with the ordered mode and the unordered mode. I think it would make sense to have this in 2 separate documents: a simple and straightforward to implement document for the simple use cases, and a more elaborate specification for ordered traversal. This would probably make it much clearer and simpler on how to implement this, and would probably also open a clearer avenue to in the future evolve the client algorithms to for example RDF-Message based designs.