Open user34ngrau opened 2 months ago
In fact, at its simplest it would be:
This will allow us to load submissions in any order (relative to versions).
The main use case is when we load an ontology for the first time but we do know that there are multiple versions of this ontology that we would like to also archive in AgroPortal. Starting at 100 will allow us to "come back later" to use the number 1 to 99 to store "previous versions".
we change the number of a first submission to 100 instead of 1 => certainly easy
Changing the submission ID number, means, reloading all the ontology data into a new graph, as submission data are saved in the submission ID graph, e.g the data for E-PHY is saved in the graph "https://data.agroportal.lirmm.fr/ontologies/E-PHY/submissions/2
)
In addition, you will need to migrate all the metadata that have as subject the old submission ID to the new one, e.g for E-PHY it means fetching all the tripes that matches <https://data.agroportal.lirmm.fr/ontologies/E-PHY/submissions/2> ?p ?v
and migrate them to the new ID
In summary, changing a submission ID, means preparing the ontology with new ID and migrating all the metadata.
Ez-publish propose un système intéressant pour ordonner des contenus On peut mettre des nombres positifs ou négatifs et on peut changer le numéro (ce n'est pas un identifiant)
Est-ce que ce on pourrait réfléchir a investir plutôt sur un script qui serait capable de reconstruire complètement l'historique d'une ontologie... ?
From discussion with Clément about the possibility to influence the "ID submission" of an ontology. Possibility to start with a latest version first and then can have the option to upload oldest version after, if they existe.