Authorization Sample Selection Algorithm about graft-ng HOT 4 OPEN

A more reliably consistent/fairer reward distribution is desirable and makes sense.

Potential SN operators would be much more likely to join the network if there was some sort of uniformity to rewards.

Especially if it encourages a more decentralised network and more SNs on the network.

The least complicated ways of achieving this should be seriously considered.

Regarding network assessed SN performance requirements - I do think this has value and could be used due to the fact that RTA is dependant on SNs response time and consistent up time especially if placed in a queue. As the Real Time Auth response time would obviously be a factor as it would be desired to be as short as possible or within some sort SLA. And SN performance requirements should reflect this but also take into account geographical differences to try promote an even distribution in location as well

from graft-ng.

Very interesting thoughts, I share your view that a more consistent reward system is needed.
I think however that delisting SNs due to poor performance is a bit harsh, especially if the que will be weeks long. Poor performing SNs could be shifted back one or a few steps in the que each time another SN has performed an auth and drops to the bottom of the que.

from graft-ng.

Require that 0.5% of all RTA payments be burned, then remint some fraction

Thinking about this some more, this really won't fly while keeping RTA amounts secret. (But on that note: a percentage-based fee for RTA payments doesn't allow for keeping RTA amounts secret in the first place).

from graft-ng.

Dropping a few steps in the que (for each newly processed block) would be a better incentive to get the SN online again asap. If you're immediately delisted, the offline-time doesn't really matter.

from graft-ng.