> Devs need to know udp is collateral damage in the ddos war Is it? Where's the citation to support the claim that UDP is widely rate-limited? Even if some n

> Devs need to know udp is collateral damage in the ddos war Is it? Where's the citation to support the claim that UDP is widely rate-limited? Even if some network operators are daft enough to be indiscriminately rate-limiting all of UDP and not just the ports that are abused (DNS, SNMP, NTP, etc.), how is it productive to encourage more network operators to do that, as this draft does? (Section 4, "Recommendations for Network Operators" point 3.)