You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently when requesting _metadata for e.g. thing modifications with the enrichment feature (?extraFields=_metadata) the whole metadata of a thing is returned, although the messages path and values only comprise a sub-set of the thing.
It should be possible to only request metadata for the effected features and properties.
I suppose it would be reasonable to send as little metadata as possible keeping requests and streams small by default. It then should be possible to request an arbitrary amount of metadata. At least it should be possible to request all metadata, or for a whole thing which will make a good start for this.
Maybe always attaching metadata is desired, that's up for debate I'd say. If not always attaching this we have to think about how to configure this in a request. Wildcard support for the enrichment feature (#710) certainly plays a role here too, since the enrichment wild-cards could also be used for metadata, although I admit I don't yet see how I could define elegantly to only provide metadata for the fields being part of this modification.
The text was updated successfully, but these errors were encountered:
Currently when requesting
_metadata
for e.g. thing modifications with the enrichment feature (?extraFields=_metadata
) the whole metadata of a thing is returned, although the messages path and values only comprise a sub-set of the thing.It should be possible to only request metadata for the effected features and properties.
I suppose it would be reasonable to send as little metadata as possible keeping requests and streams small by default. It then should be possible to request an arbitrary amount of metadata. At least it should be possible to request all metadata, or for a whole thing which will make a good start for this.
Maybe always attaching metadata is desired, that's up for debate I'd say. If not always attaching this we have to think about how to configure this in a request. Wildcard support for the enrichment feature (#710) certainly plays a role here too, since the enrichment wild-cards could also be used for metadata, although I admit I don't yet see how I could define elegantly to only provide metadata for the fields being part of this modification.
The text was updated successfully, but these errors were encountered: