Hello, is there a change in behavior that if a label of an attribute it

null

in a table, then the attribute value itself is used instead? If so, can it be turned off?

New message text here

Hi Vaclav, thanks for reaching out! I’d like to better understand the scenario before checking whether this behavior has changed. Could you please share a bit more detail about where you’re seeing this? • Could you share an example of the attribute/label setup, maybe with a screenshot or sample result showing how it currently behaves? • Was it previously showing a blank/null value,? • Are you referring to this behavior in the GoodData UI or via the API/SDK? This information will help us better understand the context and provide an appropriate response

If the label is

null

, we would prefer to show a grey dash. The reason is that we use a star schema, and when a data set points to an entity, it has to exist Conversation points to an Agent, but some conversations do not have an associated agent, so we point to an Agent data set record that represents "no agent" instead of using null as a reference (because null causes unexpected behavior by users by some records being excluded due to left joins). This dummy record has a primary ID of

00000000-0000-0000-0000 000000000000,

but the name label for this record is intentionally

null

. And we do NOT want to show

00000000-0000-0000-0000-000000000000

.

Thank you for the clarification and for explaining your setup. For clarification, what you’re seeing is related to a recent change in how tables handle

NULL

label values. When a label is empty, the platform currently replaces it with the corresponding attribute value (for example, the primary key). This behavior was introduced with the new pivot table version in dashboards and Analytical Designer, which it is a bug and currently being fixed by our engineers. As a temporary workaround, we can disable the new pivot table feature in your organization. This will revert tables to the previous version, where

NULL

labels are shown as empty instead of displaying the attribute ID. Please note, though, that disabling this will also remove some of the newer enhancements such as multi-column sorting, text wrapping, and conditional formatting. Alternatively, if your goal is to display a grey dash or placeholder instead of a blank value, one practical approach would be to replace those

NULL

label values directly in your data with a string such as

'-'

, or any desired display text (I would not recommend a special character). Since you already use a dummy record to represent “no agent,” setting its label value explicitly to a dash would produce the exact display result you’re looking for, without relying on

NULL

handling in the platform. Let me know how you would like to proceed.

If some user used those features already will the table old table still work (just not using the new features)?

Hi Vaclav, yes, the tables will continue to function, but they will not include the new features I mentioned in my last response. If you're wondering why the tables are not available in Early Access, it's because they are no longer experimental; they are now in production for the entire organization. These features can be activated or deactivated by us at the organization level. I also want to inform you that our engineers are currently aware of this bug and are working on a resolution. However, at this time, we do not have an estimated time of arrival (ETA) for the fix. Please let us know how you would like to proceed: would you prefer to deactivate the feature for your entire organization, or would you like to wait for the bug fix?