spec: 1.2+202609 version

[NicolasCARPi]

No description provided.

[FlorianRhiem]

Looks good to me!

[SteffenBrinckmann]

1. Clean up date-time format

72c72
<   "dateCreated": "2022-05-30T12:25:36+0200",
---
>   "dateCreated": "2022-05-30T12:25:36+02:00",
175,176c183,184
<   "dateCreated": "2022-05-29 16:17:38",
<   "dateModified": "2022-05-29 16:17:57",
---
>   "dateCreated": "2022-05-29T16:17:38",
>   "dateModified": "2022-05-29T16:17:57",

2. Update version 1.1 -> 1.2 in line 81

3. Clarify the hasPart rule: split across lines

112,113d111
< Subsequently, all the remaining nodes are assigned a `@type` of either `Dataset` for directories or `File` for individual files. And the `@id` corresponds to something in the `hasPart` of `./`.
< 
117c115
< If a Dataset references child Datasets (_e.g._ a parent experiment with child experiments), it SHOULD list them in its `hasPart` section. Each child Dataset that is meant to be imported MUST also be listed directly in the `hasPart` of `./`; a reference from its parent alone does not mark it for import.
---
> Every Dataset or File entity intended for import MUST be referenced directly in `./`’s hasPart. An entity that belongs to a Dataset MUST also be referenced in that Dataset’s hasPart, to record the hierarchy. An entity not intended for import, such as a previous version of a file, MAY be referenced only in its containing Dataset’s hasPart and MUST NOT be referenced directly by the Root Dataset.

4. We had "Comments exist .. at the root node". That is wrong "Comments exist ... in graph"

150c148
< All nodes with a `@type` such as `Comment` or `Person` exist at the root node (once), and can be referenced via their `@id` in other parts.
---
> Contextual entities, such as Comment and Person, MUST be represented once as separate node objects in the top-level @graph. Dataset and File entities MAY reference them using their @id.

5. Replace identifier with propertyID as it is given in schema.org. (Identifier has different meaning) Also add an example

163c161,171
<   * The `identifier` for a `@type: PropertyValue` node can be set to an IRI (e.g. the URL for an ontology entry, such as http://purl.org/dc/terms/instructionalMethod) for specifying the meaning of this node.
---
>   * A PropertyValue node SHOULD use propertyID to identify the represented property. propertyID MAY be an ontology IRI. The recorded value SHOULD be given in value.
>     ``` json
>       {
>         "@id": "#air-temperature",
>         "@type": "PropertyValue",
>         "propertyID": "https://qudt.org/vocab/quantitykind/Temperature",
>         "name": "Air temperature",
>         "value": 21.5,
>         "unit": "C"
>       }
>     ```

[SteffenBrinckmann]

More items from our discussion today:

1. Append to line 113 "All nodes with @type: Dataset SHOULD...."
   keywords are a comma separated list of words.

2. Append to line 154 "additionalType:...."
   additionalType: {"@id":url} OR [{"@id":url}]

[NicolasCARPi]

Thanks Steffen, I addressed your comments.

[SteffenBrinckmann]

Great