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
The goal of this feature is to be able to quickly find all instances of an arbitrary attribute in the entire graph.
Originally I was going to implement this by creating an "attributes" node, and then for each unique attribute, create a node for it and connect it to attributes. Then each instance would be connected to the archetype node of that attribute name. Agdb also has an indexing system for quicker search of specific attribute names. It could be a better fit for this feature.
The benefits of having it all be nodes is that they could utilise relative positions when viewing the context of a certain attribute archetype. The downsides are performance and bloat in the database. The root attributes node would need to exist in either case, because it would need to store a list of all existing attribute names. Conceivably it could also store some settings for how the attributes are laid out or sorted when in that view. Maybe it's unlikely that there would need to be anything more granular than that.
The text was updated successfully, but these errors were encountered:
The goal of this feature is to be able to quickly find all instances of an arbitrary attribute in the entire graph.
Originally I was going to implement this by creating an "attributes" node, and then for each unique attribute, create a node for it and connect it to attributes. Then each instance would be connected to the archetype node of that attribute name. Agdb also has an indexing system for quicker search of specific attribute names. It could be a better fit for this feature.
The benefits of having it all be nodes is that they could utilise relative positions when viewing the context of a certain attribute archetype. The downsides are performance and bloat in the database. The root attributes node would need to exist in either case, because it would need to store a list of all existing attribute names. Conceivably it could also store some settings for how the attributes are laid out or sorted when in that view. Maybe it's unlikely that there would need to be anything more granular than that.
The text was updated successfully, but these errors were encountered: