I was watching how a package moves through a delivery network. The strange thing is that speed is not always about moving. Sometimes it comes from knowing where not to send the package.
That thought came back while reading about Kadcast in Dusk. Most people see message propagation as a networking detail. I do not think that is correct. Dusk relies on Kademlias XOR distance, where routing decisions are shaped by how nodes are from each other in the Dusk network. The away a destination sits in the Dusk network the fewer unnecessary relays are needed in Dusk.
What interests me is the pressure this creates beneath the surface of Dusk. Traditional broadcast methods often solve reliability by sending the message everywhere in the network. Dusk takes a path. Kadcast in Dusk selectively forwards messages across increasing XOR distances in Dusk reducing traffic and bandwidth consumption in the Dusk network.@Dusk
The catch is that efficiency and resilience do not always pull in the direction in Dusk. If Dusk keeps reducing noise in the Dusk network does the Dusk network become more dependent on routing quality and healthy node distribution in Dusk? That is the part I rarely see discussed about Dusk.
Maybe the real test for Dusk is not whether Kadcast works when conditions are ideal, in the Dusk network. It is whether the shortcuts still hold when the Dusk network becomes crowded, uneven or unpredictable. #Dusk @Dusk $DUSK
$ACE
$XOS.US
That thought came back while reading about Kadcast in Dusk. Most people see message propagation as a networking detail. I do not think that is correct. Dusk relies on Kademlias XOR distance, where routing decisions are shaped by how nodes are from each other in the Dusk network. The away a destination sits in the Dusk network the fewer unnecessary relays are needed in Dusk.
What interests me is the pressure this creates beneath the surface of Dusk. Traditional broadcast methods often solve reliability by sending the message everywhere in the network. Dusk takes a path. Kadcast in Dusk selectively forwards messages across increasing XOR distances in Dusk reducing traffic and bandwidth consumption in the Dusk network.@Dusk
The catch is that efficiency and resilience do not always pull in the direction in Dusk. If Dusk keeps reducing noise in the Dusk network does the Dusk network become more dependent on routing quality and healthy node distribution in Dusk? That is the part I rarely see discussed about Dusk.
Maybe the real test for Dusk is not whether Kadcast works when conditions are ideal, in the Dusk network. It is whether the shortcuts still hold when the Dusk network becomes crowded, uneven or unpredictable. #Dusk @Dusk $DUSK
$ACE
$XOS.US