Define the process of publishing versions of MatVoc

[TBoonX]

"Etablierung eines Governance-Prozess, bei dem regelmäßig abgestimmte und qualitätsgesicherte Versionen des MatVoc veröffentlicht werden"

For this we need a process which we could follow.

[TBoonX]

Excerpt from Javads presentation https://docs.google.com/presentation/d/1Rw7sMEObfrUQR0vzVfZXczIlG-w7kO39MBmwx4FlRvU/edit#slide=id.g72e9cb7c4e_1_7

Ontology publishing

• Dereferencing/lookup
  • Vapour
• Namespace maintainability
  • stable URI
  • PURL
• Usability
  • Human-readable documentation
    • human-readable definitions
    • basic metadata: creator, publisher, date of creation, last modification, version number
    • Labels and descriptions, if possible in several languages
  • Content negotiation
    • formal RDF content
    • human-readable documentation

This does match the ideas from InfAI.
The CI should mark commits with good quality as valid (via tag) which creates a list of release candidates over time.
from these candidates Releases should be done manually on GitHub. This should be done on a regular basis or event basis.
These Releases will include release notes.
Based on GitHub releases we could create the human readable webpages with tools like specgen. In order to create such a website additional triples have to be add and also styles have to be improved.
Additionally the vocabularies from MatVoc should be published to common channels/platforms, e.g.: lov, schema.org, old.datahub.io

[TBoonX]

Instead of specgen we could use https://github.com/RDFLib/pyLODE which is not outdated and creates nice pages.

[TBoonX]

I will create a document under https://docs.google.com/document/d/1cwoVhiDBEk3HedUKh5AHS6Hdh_a0h236DuD3P2Pwa9s/edit# in which I propose a workflow with a toolset. This should then be discussed.