Defining an Ideal Index

What’s an ‘ideal index’? What’s an ‘index’? So say we were to design a public index? What would it be like?

What standards would guide us? What would the ideal index look like? How would it work?

Complete It would index all available resources.
Fast It would return good results quickly.
Simple It would be easy to use.
Timely An ideal index would be up-to-date. It would be able to quickly discover and index resources and revisit those web sites at appropriate times to record changes.
Relevant Results from searches would be useful – as much as possible.
Accurate, Honest and Fair It would present search results fairly and without bias.
Private Users of the index – and the operators – would be protected from surveillance and identification. By design, it couldn’t be used to spy on you, target you, or track you.
Secure It would be impossible to interfere with the collection, presentation, or use of data, or with the network or servers.
Robust The index would be robust – that is, unbreakable. The system would be designed to withstand any kind of disruption. It would always be available.
Representative It would provide equal service to all users, with no one receiving special treatment.
Unencumbered There should be no obstacles to the use and distribution of the index or its parts. It should be free.
Automatic It should run on its own. It should be able to do load balancing and task assignment based on its own monitoring of the network. It shouldn’t require expertise to set it up or to operate it.
Efficient It should make conservative use of resources available to it.
Affordable It should be possible for anyone with basic computers and a network to use and operate it.