CMPs Handling cookieRefresh from Device Disclosures

[kenario]

Referencing this documentation: Device Storage Duration & Access Disclosure.

How should CMPs handle a disclosure that is missing the cookieRefresh key but is of type cookie. The case being that cookieRefresh is required when the disclosure type = cookie. Should we add the cookieRefresh key ourselves with a default value of false?

[dmdabbs]

Thanks for opening this @shortaflip. I'll recap our discussion.

I believe code that parses/loads deviceDisclosure.json files should not add any value where one was not provided (unless the value has a specified default value). The attribute would have an undefined value to reflect the file's content. The CMP would disclose all the other valid information for the file entry.

As for what CMPs should display in this or similar cases, @alex we should get TCF MO guidance for the specification. I will raise for feedback if this week's Framework Signals Quality WG agenda permits.
cc: @Facens , @lon-pilot-mpf

[kenario]

@dmdabbs, let me make sure that I am understanding it correct. The parsed object should contain the key but the value would be undefined?

[dmdabbs]

Hi @shortaflip . Discussion in today's WG meeting the consensus was that CMPs or library code should return the value undefined when a caller accesses this attribute but it was absent in the underlying deviceDisclosure.json file. CMPs are free to display that as they deem appropriate.