Why doesn't DagKnight use a fixed k parameter like GHOSTDAG does?

DagKnight removes the hardcoded k parameter that GHOSTDAG requires, allowing it to adapt dynamically to whatever k the network is actually producing. In GHOSTDAG, k is a number chosen in advance that shapes how the protocol handles blocks created in parallel — but a fixed k is always tuned to an assumed network state that may not match reality. DagKnight instead measures and responds to the "real" k as it naturally emerges from live network conditions. For users and developers, this matters because the protocol becomes self-calibrating — it stays well-tuned as the Kaspa network grows and changes, without needing manual parameter updates.