Conversation

Notices

1. Embed this notice
   GhostOnTheHalfShell (ghostonthehalfshell@masto.ai)'s status on Friday, 27-Dec-2024 08:12:49 JST GhostOnTheHalfShell

   in reply to

   • Esther Payne :bisexual_flag:
   • Brett Sheffield (he/him)
   • Librecast
   • Daniel Lakeland

   @dentangle @dlakelan @librecast @onepict

   In ages past, in another life, I used to know more an out the innards of these things, and know something of the innards of routing.

   • Embed this notice
     Brett Sheffield (he/him) (dentangle@chaos.social)'s status on Friday, 27-Dec-2024 08:12:50 JST Brett Sheffield (he/him)

     • Esther Payne :bisexual_flag:
     • Librecast
     • Daniel Lakeland

     @GhostOnTheHalfShell @dlakelan @librecast @onepict Ghost: There is additional latency with IP multicast depending on the number of hops (routers), as the MLD join has to propagate. There are also delays on parting a group. MLD is not designed to be a low latency protocol and performs very poorly when there are lots of joins and parts.

     IP multicast is very broken and there are no forms of IP multicast remaining that even meet Librecast's definition of "multicast" 😉 eg. SSM is *not* multicast.