Add Join: What is the Data Source? Why choose a One-to-Many option rather than Relate?

[Craig_Eissler_Iceman]

Hello,
I'm a bit confused by the new 'Add Join' tool that provides the option for a 'One-to-Many' operation.
Note that I have already read all of the Help documentation on this.

First of all, the "default" operation explanation says that the join will be based on the data source.
...What, exactly, is meant by "data source"?
I'm only aware of there being an Input Table and a Join Table.

Secondly, the One-to-Many operation used to not exist as an option in previous versions ...So, if I had this table association in the past, the tool would recommend a 'Relate', and I would do that instead.
...So, what is the reasoning why this was added as a Join option? ..In other words, what advantages does this give me over a Relate?

[DanPatterson]

Data source :  are you working with locally stored file geodatabases?, or SDE data sources? cloud-based? csv ?

A lot of the differences can be ascribed to whether duplicat OBJECTIDs are permitted or not, or whether they have one, or not

The Validate Join tool was introduced to assess those scenarios and more

Validate Join (Data Management)—ArcGIS Pro | Documentation

The rest of the cavaets you have read in the help topics

[Craig_Eissler_Iceman]

My question about "data source" is specifically within the context of the "default Join cardinality". I don't understand what the data source has to do with that, unless "data source" has a different meaning when it comes to Joins.

My question about One-to-Many table association added as a chioice in new Add Join tool is, "why"?
It wasn't there before. Instead the Validation would recommend a Relate. 
In other words, the question could also be stated as, when/why would I choose this new Join option vs. Relate?...I've also posed this to Tech Support, so we'll see...

We'll see what others may say. I've also posed this to Tech Support.
Thanks,

[MervynLotter]

Hi Craig

A relate is quite different from a join in that a relate establishes a relationship between two features or a feature and a table, but it does not physically join them (if you were to open up the attribute table, you would not see the related features). You can query relates but the link/relationship is invisible. Joins on the other hand, add the fields from one feature to another feature or table. So if you open up the attribute table, you will see attributes from both. 

I have used both but generally prefer Joins for my workflows. 

As for data source, I am a little less sure about what Esri were intending, but I can guess to say that the features need to exist on disk somewhere (as opposed to Event layers or layers that are in memory only). You will notice in the next release of ArcGIS Pro, one can create temporary or in memory layers.  It is also possible that data source refers to its use in Model Builder, where the layer may not yet exist if the model has not run.

And lastly, if you open up an attribute table in Fields View (Data Design > Fields), you can view the attributes based on the layer names, or the Data Source, and the Join tool will need to reference the actual raw attributes in the data source and not the layer in the map (which can be renamed to almost anything, with special characters, which are a nono in GIS). But this is my guess only and I may be wrong. 

Best wishes,

Mervyn