Sorry I’m late replying to this conversation thread.
As @Sam noted the lineage graphs can get very complex. Some of the graphs we’ve been looking at have lines going everywhere (and that’s without grandparents or grandchildren being visible). It would be really important to be able to reduce the level of complexity to make the graphs readable and interpretable.
Table A → Table B → Table C mockup
The approach in this mockup (Table A - > B → C mockup) would be useful
It would help to be able to show/hide the grandparents/grandchildren. So if you are in the lineage graph for Table C, you could show/hide the lineage to Table A.
This could be done through another option box here:
Mockup of the Lineage Information in the Path Table in Distributions
I agree with @Michael that the Lineage Information needs to be accessible. Currently it can only be viewed by opening the distribution in the editor (then remembering to close it without saving).
However I’m not sure whether including it in the Path Table in the Distribution is the right approach (as per the second mockup). In some cases there may be a fair bit of Lineage Information that is included. This may make the Path table very long and hard to navigate.
It may be helpful to be able to either show the more detailed information for each path in a separate window or make it so the detailed information can be toggled on and off.
Lineage within a distribution
In some cases a path may have a lineage to one or more paths in the same distribution. Aristotle doesn’t currently support this type of lineage relationship. Is this something that should be supported?
Information about methods used in the lineage
On another matter related to lineages, some of our stakeholders have been asking about cases where a standard method is used to produce an item from other items (e.g. the standard methods used to code ANZSIC or ANZSCO from input items).
Would it make sense to have a metadata object to describe the methods used that could then be connected to the lineage path?
The benefit would be that the method could be described once then reused rather than needing to be described in the Lineage Information each time.