Hi, On Jun 11, 2025, at 11:47, ruediger.geib@xxxxxxxxxx wrote: > Apart from these, Med and Lars had minor comments and nits, which should be addressed by this version too. Lars raised concerns with the term "traditional", which aren't addressed by this version. I also said > Major issues: > > I'm kinda sad to see it's 2025 and we still invent new control protocols, > especially over UDP. Most of the complexity in this doc is due to that. I can't > see a requirement here that the control protocol (also) needs to run over UDP. > Can we not simply exchange JSON blobs over HTTPS to configure the subsequent > UDP testing flows? That should eliminate much of the complexity here, and would > probably be more extensible as well. I appreciate that this probably isn't actionable at this time, but I do strongly believe that building these sorts of protocols from scratch is not something we should do anymore in 2025. Thanks, Lars -- last-call mailing list -- last-call@xxxxxxxx To unsubscribe send an email to last-call-leave@xxxxxxxx
