yeah, those notes are not so promising... but i have to wonder how current or specific the test conditions were...
the thing with most dynamic routing protocols specifically complex link-state types... is load as a metric... this means that multiple paths through the mesh are used and load optimised... getting you a performance bang for your buck... rather than just failover pathways.
there is also olsr(link-state routing protocol)... but in any event... we both learned something... and there is a reason network admins get paid good money ( ideally lol. )...
what would be my disadvantage of just using 802.11s? instead of using batman-adv?
my setup is simple.
Interface is only LAN (i deleted WAN and WAN6). I dont use any mesh interface.
on backhaul band i added the 2nd mode 802.11s with a unique mesh id.
bind it on LAN.
encryption = none
the mesh is running. (no remove and install of package no editing)
I will be just doing ordinary mesh with a single bridge to lan.
As even with batman its not properly tested or no one actually has tried setting up a mesh with 2 lan bridge.
I don't understand why 802.11s is all the time mentioned by you.
This is a standard for wireless lan only and has never been intended for anything like ethernet.