Deviding sixlowpan into 2 seperate modules makes sense from an architectural standpoint. But additionally regarding space constraints I wonder if it would be even more plausible to devide them further up with a dependency graph as follows (in git-log --graph like syntax ;-)):
Deviding sixlowpan into 2 seperate modules makes sense from an architectural standpoint. But additionally regarding space constraints I wonder if it would be even more plausible to devide them further up with a dependency graph as follows (in
git-log --graph
like syntax ;-)):I guess this way we can save a lot of space (no forwarding table, neighbor cache size == 1, less code) and overhead in general in the network stack.