If only there were some sort of system for mapping human-readable names to numeric addresses! Alas, such a service is only a fantasy here in the future.
The problems I see with such a scheme are (by necessity):
Centralize the discovery of such mappings. One could imagine that this would allow for arbitrary disruption of access by removing mappings from that index.
The above centralization could allow for a small cartel of companies or governments to do major rent-seeking. Imagine if you had to continually pay for the privilege of having one of these public mappings. It would be trivial for the cartel to agree on some “reasonable” fee and then to force everyone to engage in such payment, escalating over time. Nobody would put up with that for five years, much less the decade or two one would expect such a scheme to remain in service.
How would such a system deal with name collisions? What if you wanted the name “hank” and I had a server who I wanted to refer to as “hank”? Even worse, what if somebody quietly compromised the system to broadcast as “hank” while silently forwarding the traffic to the real “hank”, snooping on everything that happened? The user of the system would have no way of knowing!
Overall, I think that such a system is bound to fail, due to its glaring shortcomings in both security and economics. It’ll never take off. :)
I find this to be eerily applicable to human beings…
Centralized the discovery of the real world. One could imagine that this would allow for arbitrary disruption of access by removal of various organs.
The above centralization could allow for a small cartel of viruses or bacteria to do major infectious take-over of body prats. Imagine if you had to continually pay for the privilege of vaccinations and medications. It would be trivial for the cartel to evolve, raising the “reasonable” fee of medication, forcing everyone to engage in such payment, escalating over time. Nobody would put up with that for five years, much less the decade or two one would expect such a scheme to remain in service.
How would such a system deal with conflicting input? What if you want to see a cloud in the form of a rabbit, and I saw the same cloud as a duck, which is it? Even worse, what if somebody quietly compromised the system to hallucinate or comatose? The user of the system would have no way of knowing!
Overall, I think that such a system is bound to fail, due to its glaring shortcomings in both proving reality is real and countless methods of failure. It’ll never take off. :)
If only there were some sort of system for mapping human-readable names to numeric addresses! Alas, such a service is only a fantasy here in the future.
The problems I see with such a scheme are (by necessity):
Centralize the discovery of such mappings. One could imagine that this would allow for arbitrary disruption of access by removing mappings from that index.
The above centralization could allow for a small cartel of companies or governments to do major rent-seeking. Imagine if you had to continually pay for the privilege of having one of these public mappings. It would be trivial for the cartel to agree on some “reasonable” fee and then to force everyone to engage in such payment, escalating over time. Nobody would put up with that for five years, much less the decade or two one would expect such a scheme to remain in service.
How would such a system deal with name collisions? What if you wanted the name “hank” and I had a server who I wanted to refer to as “hank”? Even worse, what if somebody quietly compromised the system to broadcast as “hank” while silently forwarding the traffic to the real “hank”, snooping on everything that happened? The user of the system would have no way of knowing!
Overall, I think that such a system is bound to fail, due to its glaring shortcomings in both security and economics. It’ll never take off. :)
The only thing that could make this nightmare even less practical was if it worked over UDP! Think how hilarious it’d be!
I find this to be eerily applicable to human beings…
Overall, I think that such a system is bound to fail, due to its glaring shortcomings in both proving reality is real and countless methods of failure. It’ll never take off. :)
Its a non-issue. The IETF already changed changed the address format for IPv9 in RFC1606 for support of more granular hierarchies.