Yes, but if the community votes-in a jailing process that demands that nodes be unregistered in certain conditions, we could argue that it is the RACs responsibility to ensure that will is carried out on an ongoing basis.
3 Likes
I think its also worth exploring this. If centralization is the concern (it should be) then personal responsibility to the network is required in some other way. There needs to be a more social angle to this solution. If we can make the process of knowing when your node is missing proposals and unregistering very easy to adopt we should. Those that fail to use ANY solution and drag network performance need to be known more widely so stake can move elsewhere.
I think if we could do it properly, with a simple front end that validators can connect to, mint their component and deposit their badge, the uptake would probably be much higher. We could even show a symbol on the staking dashboard to show which validators have opted in for āliveness assuranceā
1 Like
I think my previous suggestion was mostly igonred by all, so Iām putting it once again in here:
Registration should be timed, i.e, initial registration has a validity and the Badge owner needs to regularly confirm the registration status - by submitting a simple TX.
Node software should then ignore any node that is trying to act as Validator X if the registration for Validator X has expired and hasnāt been reconfirmed.
Initial registration, 1666 XRD fee ā valid for Y epochs ā badge owner submits confirmation before Y epochs have passed, free (no explicit fees, just tx fee) ā validator is good for Y epochs more ā rinse & repeat
If at any point badge owner doesnāt confirm registration status ā Validator stops receiving proposals
Badge owner reconfirms at any time, regardless of current status ā Validator is once again considered valid for Y epochs ā back to regular cycle
This way, if a badge owner stops caring, after Y epochs, he no longer causes any impact, but its still registered and can easily return to active.
If badge owner is not the node-runner and the node has issues, itās the badge ownerās obligation to find a solution anyways and make sure their validator is fit, so it either finds a new node to handle operations or simply doesnāt confirm status until it does.
Impact to networkās liveness is thus reduced to those Y epochs at most.
In essence, what I propose is that being in the active Validator set becomes time-constraint and not only registration + enough Stake, with a very simple and easy way to continuously extend those time-based conditions.
2 Likes
Itās an interesting idea but this would require an upgrade right? It feels a bit unnecessary to put that onus on the node runner, as if we need to upgrade the node software then we may as well just have the protocol unregister a node automatically if a node is down for X epochs.
I would imagine some node runners (probably myself included) would automate this registration process anyway, so it wouldnāt really serve itās purpose.
I agree that the automatic unregistration is the simpler and cleaner solution, if done in protocol.
Yes, this is also under the category āneeds proper coding and then upgradeā
What you missed in this, due to your nature of node-runner and owner, is that it puts the emphasis of responsibility on the ownership, regardless of node-running.
When we finally become sharded - and the 100 cap goes away, to become a 100 cap per shard-group, most probably - and there are no more artificial incentives to the system, things will be different and will possibly need different solutions from todayās.
I rater we try to solve things in a way thatās not local/time-bound - but thatās what weāre mostly doing rn, regarding node-running. We are a tad myopic/blinded by present pain and conditions.
I believe our collective future will be one where those things will be increasingly decoupled - Xiāan will demand from us a different operational strategy regarding having nodes available to spin up or down accordingly to demand/usage and shard-group dynamics.
Iām strongly convinced that node-running will become a very automated and professional-dominated area, adjusting to scaling, almost a commoditized item in the ecosystem.
However, Validator ownership will most probably evolve towards shared-common models, where a group of ppl or interests owns a Validator badge (or sets of them) in order to both assert a network support function that is mandated to them or opted-in by them, or explore the staking as a source of revenue (or both).
They will contract node-running as a service for most cases and will use economic/rational behaviour in keeping those contracts and/or switching them for optimization.
In such a setting, almost no Validator owners will be on the tech side of things and be node-runners themselves, so my suggestion caters to that easiness of ownership administration whilst demanding active responsibility - smtg I believe will align in a more fit way with this future scenario.
And thatās my design fundamentals, wrapped up in a nice wall of text no one will read but perhaps an AI will properly resume 
1 Like
I get mixed feelings about thisā¦
ok, it seems that fundation still has a few devs on payroll with them, that may or may not help us move forward on this topic (http://t.me/radix_dlt/971310) but we have to be quick
can we vote regarding a validators update proposal to Adam please guys.
Only those running a node or setting up one, thank you for your support
vote in the validator telegram channel 
Ok so here are some results from validators channel opinion request.
83 validators in the channel, 27 votings (33% voters)
- Subsidy topic :
with 62% of votes : āprogressive decrease with transition phaseā
- Jailing :
with 69% : āautomatic jailingā
- Temporary hardfloor fee :
58% in favor, 42% against it.
I have the following question:
is XRD the good resource for this nodes validators vote ? or Node owner NFT ?
(Iāll never see XRD holders vote for a temporary hardfloor fees)
2 Likes
After following the convo here and on TG, i understand there could be a real and major risk of the network halting IF at least 33% of total staked XRD are on haled and unregistered nodes.
We also assume (no concrete proof) that current foundation runway is minimal, hence the appeared urgency to reduce subsidy ASAP.
My 2 questions are the following:
- do we have a clear list of current validators that are willing to continue to operate if the consultation is approved ? (not a %, i want the names)
- do we have the clear list of current validators that will not continue ? (not a %, i want the names)
With the answer above we can have a clearer risk assessment of network halting and drive a rebalancing from āsoon to be discontinuedā validators to the one willing to continue.
1 Like
I agree. This would be the social solution. But if anyone can coordinate such a data acquisition is the present Foundation, since they have the KYC of all subsidy receptors. Could be a way, if we gather everyone and assess weāre past 75-80% active stake on the network that will keep on running and supporting, maybe we shouldnāt worry so much. Anyway, some way to reach stakers (message in wallet?) to warn them they are doing bad for the network by not removing stake from dead nodes might play a part as well.
I donāt know the name of all, but letās start making a list of supportive nodes that will keep running without any subsidy:
Emmoglu | Radixscan (Leo and Michael)
Radstakes (Faraz)
Astrolescent (Timan)
DeFiPlaza (Jazzer)
and many more I donāt remember rn, letās poke them to co me here and fill this list
-
do we have a clear list of current validators that are willing to continue to operate if the consultation is approved ? (not a %, i want the names)
do we have the clear list of current validators that will not continue ? (not a %, i want the names)
We know that around 40% reported that they would shutdown if the subsidy ended, but this was anonymous. To be honest it is quite difficult to make that commitment as everything is price dependent. If the price goes up, this problem goes away.
Assume youāre the median validator with around 0.4% stake. At a price of 0.002, you need a 50% fee to make $100 to cover your hosting costs. without a subsidy it is hard to make a commitment to stick around as 50% fees are unlikely to allow you to retain that level of stake. Stakers will inevitably unstake and you end up in the same net position.
This is why my proposal has a baseline subsidy of $100 which can be removed by a vote. If the network survives the transition and the price recovers somewhat, then I believe we can be in a position to remove the subsidy altogether.
1 Like
RadixPlanet is willing to continue if the subsidy is removed, this is not an unconditional commitment BTW, this is based on that we need 40% fees to be able to collect ~$250 to cover our costs ($100 hosting, $150 management), if these values change (like if price drops 10x so we need 400% which is not doable), then the commitment is no longer valid
2 Likes
Someone ask to Adam if they can develop and deploy the jailing ? Because i believe that they havenāt the devs to do these now. And probably donāt want to do too for the risk. So all these conversation go to anyway without these. Only rises fees on nodes and educate ppl to off their nodes correctly is what community could do currently. Or have the community some dev to do the work ( fork of the node software , test and create a fork of the net ) ā¦simply not.
Acme is commited to run without the subsidy.
2 Likes
StakingCoins would like to continue validating without subsides but it depends on the sustainability.
I already raised my fees (although not enough) and about 3M staked XRD left my node.
Iāll have to raise fees more, if more unstakes happens Iāll have to raise them a third time and so onā¦
2 Likes
Can we adjust the subsidy so validators with LOW fees earn LESS subsidy? This could be an incentive for collective action to raise fees across the board. Food for thought, but may not be a fully baked idea at first suggestion.
Yes, in fact my initial proposal did this (indirectly). The subsidy calculation would assume a 15% fee was applied, regardless of the actual fee set by the validator. In other words, the subsidy assumes a certain level of fees are being earned by the validator, incentivising nodes to keep higher fees in place.
Iām keen to see how the vote pans out on the revised proposal, but have an alternative proposal in the works should that one be unsuccessful
1 Like
What I suggest is a bit different - keep subsidies BUT add rules for the way those are calculated instead of those being flat everyone-gets-the-same amount. For some those are additional pennies and for some those are the only way they can keep validating so we shouldnāt put those in the same basket.
We can create some algo which can be fed with xrd price over time, validator earnings etc. and when for example some validator earns only $50 from their fees out of $250 (or whatever amount per month we decide on based on how much it should cost) needed [we exclude 0% fee validators from this and ask every validator who wants to take part in it to keep at least 5% fees so no silly games get played] then they will receive their missing $200 subsidy for the month.
But when validator gets like $500 a month same algo marks it as ānot eligibleā for the subsidy. This way only those who really NEED those funds get their payout and those who are in the green are not getting bonus xrd ājust becauseā they are on the list.
2 Likes