I feel like the use case around RFI is not clearly fleshed out.
What parties and systems would be issuing RFI requests and under what circumstances?
What parties and systems would be responding to said requests?
Only once those two things are well documented and understood by all can we understand how to best service such a request.
As an example of why this is important... I believe only a tiny fraction of potential TAXII client systems would ever have an ability to respond to historical RFI requests... Most would not. So maybe this RFI capability should be limited to only TAXII servers who implement a "repository" specification.
Sent from IBM Verse
