When the network is down, the PNI altruistic servers handle the entire workload by way of gateway relays. This is nice, but undesirable in a few aspects.
it’s highly centralized.
it’s expensive to PNI
it’s really bad PR which demonstrates rather dramatically how massively under utilized the network is.
What if, we developed a list of persons/ organizations that were willing to allow the gateway to use their backend when the P2P net was down? After all, the backend is not doing anything during those periods anyway.
The old PNI altruist for 0021 is 64 cores; a massive server and a massive expense. This will not be the case in the new architecture under PNF and I would welcome the chance to add more distributed altruists.
This open issue is highly related and about to go into development now:
I’ll work with you over Discord if you’re willing to be part of the first experimental set in the new altruists.