The user response level

From the server level (or to begin with, the enquiry level) we have two pieces of information from each participating server:

1. the number of matches
2. the list of gaps, each unanswered question with a number of times it has been required.

It is a straightforward matter to put these together into

1. an overall number of matches across the whole service;
2. a complete list of gaps, with their frequency across the whole service.

The number of matches is absolutely fundamental to the way that processing continues.

• If there are already any non-zero number of matches, there is no need for the actively enquiring person to answer more questions at this stage.
• If there are many matches, the feedback given will be to add requirements, or make the requirements more stringent or demanding.
• If there are just a few matches, this is the ideal answer, and the enquirer can proceed to reviewing those “correspondents”. If there are also gaps, the enquirer may be asked if they want more correspondents, and if so, try answering these questions and re-submitting the enquiry.
• If there are no matches, but there are some gaps, then perhaps at most three questions would be enough to present to the enquirer. They can answer those and re-submit the enquiry.
• If there are no matches and no gaps, then it will be up to the enquirer to relax requirements, or simply to leave the enquiry on the system as passive.