Waqas Hussain wrote:
I suspect the CPU cost would be lower with this proposal. Saved bandwidth was more appealing to me than saved CPU time, but the missing unsubscribed issue makes this optimization invalid, so no point in attempting tests.
hashes or counters make it possible to check for missing unsubscribed. I've already proposed how to modify repeaters so that they maintain the sleek, uncomplicated protocol and additionally check consistence through hashes. But I guess in this community one has to formally submit such a proposal to edi...@.
