Dataverse Primary Name Column

[todaytoo]

Hello! I keep seeing conflicting details around the 'Primary Name Column' in dataverse when creating a table. What should be populated in this column or is it even needed? Some documentation I see says that it should have your Primary Key in this column, others say to leave this column alone.

Can someone help to explain what this column is and how it should be used, or point me to clear documentation around this topic?

Thanks!!

[AhmedSalih · Accepted solution]

Hello, @todaytoo, A Primary Name column is always created by Microsoft Dataverse when you create a new custom table. This is the first column that is listed and available so you have a way to identify a record in the new custom table by using a business value or an autogenerated whole number. In the model-driven apps, the value in the Primary Name Column will be clickable so you can navigate to the Form of that record. So, it needs to be user friendly value, like company name, or Order number. 

Also, when you have relationship between tables through lookup column, the Primary Name column value will show as the value of the Lookup column in the Child tables. 

It is needed and as a best practice you should use it and you can't avoid using it.

 

If my reply helped you, please give a 👍 , & if it solved your issue, please 👍 & Accept it as the Solution to help other community members find it more.

I am primarily available on weekdays from 6-10 PM CT and 5-10 PM CT on weekends.

Visit my Blog: www.powerplatformplace.com

[carlilelance]

It's an excellent question, and the truth is, the way Microsoft has implemented it, in both SharePoint and Dataverse is kinda clunky. In most databases you would create a unique (only one value, never repeated) primary key. Many times this would just be an auto-number (like 1,2,3,4, etc.,). Then you would have a more human-readable value (George, Tom, Susan, etc.) which would also frequently be unique. Microsoft, in order to appear user friendly, auto-creates a unique identifier (primary key) with the same name as the table (by default) and then forces you to use the primary name field (and assumes you will probably make it human-readable). This, unfortunately, tends to confuse new users AND old database gurus. So if you are confused, just know you are not alone.  Sometimes, you don't want or need to have a human readable primary name field, but you are forced to anyway. I've seen data normalization broken because of this issue. Another issue is Microsoft forces it to be a text field. This can break common database features (try sorting if you make it numbers, for example). One alternative is to just use the auto-number feature, or ignore the primary name field. 

[AnttiKurenniemi]

I wonder what we should put in the "Primary Name" column when we have an "intermediary" table in a many-to-many relation, where we need other values besides just the id's of the two connected tables.

For example we have table of organizations, and another of persons, and an intermediary table of "person to organization" where we have the starting and ending dates, when a person has been in the organization and maybe even a reference to a role. Thus, we can keep history for persons; in which organization have they been, at which times and in which role, and we can easily check the current position of a person by looking at the latest entry in this intermediary table.

So we have

Table "Organization" -> field "id"

Table "Person" -> field "id"

Table "Person-to-Organization" -> fields "organization-id", "person-id", "starting-from" and "ending-to"

All of this is doable and ok, but what on earth should we come up with as the Primary Name column? How do we populate something there? We can't use Formula columns, so we'd need to just put... something, from the application code?

...yes we could do this by just having the auto-generated many-to-many relation and then perhaps a log table or some such, with the dates, but I'm just wondering about this because this is how we have used SQL Server in many cases, and the approach in Dataverse is a bit difficult for me to grasp.

[ivantohelpyou]

Here's something I did with a similar situation to track social media posts.

Post Queue table contains ideas for posts and articles.

Social Network table is a lookup with LinkedIn, Facebook, etc.

Post table has customized versions and images for each post. 

Solution: 

1. For Post, set the "Required" property in the Primary name column to "Optional."

2. Create a business rule for Post that automatically populates Name to "Posted" or "Not Posted" based on whether there's anything in the timestamp. 

3.  Delete "Name" from the main form for Post, show it (next to Social Network) on the Associated View.

The Post Associated View for an item in the Queue may show: 

LinkedIn   Posted

Facebook  Not Posted  

 Business rule based on timestamp