Author Topic: Replace delegate fee of two weeks pay with a two week no-pay probationary period  (Read 1742 times)

0 Members and 1 Guest are viewing this topic.

Offline arhag

  • Hero Member
  • *****
  • Posts: 1214
    • View Profile
    • My posts on Steem
  • BitShares: arhag
  • GitHub: arhag
A delegate should secure their votes prior to registering and have a high degree of confidence that they will be elected and earn their funds back.

If someone cannot build a coalition prior to registering then that is a good sign that they are not a good candidate.

And how exactly are we expected to measure that without any off-chain voting functionality? Just because some accounts on a forum say they support a diluting delegate's candidacy doesn't mean it will translate to the delegate getting enough votes from BTS stake.

I would suggest that all dilution based delegates should run as non diluting delegates first.

This lets stakeholders know that they are good block signers, but it doesn't tell them anything about if they are a good worker or if they are worth the pay they eventually intend to ask for. This again seems to all stem from the mistake (IMHO) of mixing block signers and DAC employees together into one entity, a delegate.

$2000 is proof of commitment not just proof you are not an attacker.

So apparently only in the DAC world do we expect the employee (who would only get paid after stakeholders have decided the employee is worthy of being hired BTW) to actually pay an upfront fee for the privilege of contributing their value to the DAC? This is like the reverse of what occurs in traditional companies where they have signing bonuses and they pay for relocating costs.

Offline bytemaster

A delegate should secure their votes prior to registering and have a high degree of confidence that they will be elected and earn their funds back. 

If someone cannot build a coalition prior to registering then that is a good sign that they are not a good candidate.

I would suggest that all dilution based delegates should run as non diluting delegates first. 

$2000 is proof of commitment not just proof you are not an attacker. 

For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

Offline arhag

  • Hero Member
  • *****
  • Posts: 1214
    • View Profile
    • My posts on Steem
  • BitShares: arhag
  • GitHub: arhag
I believe the two week fee to register as a delegate is to allow stakeholders enough time to kick out (by voting enough for other honest delegates) a malicious delegate that votes themselves into the top 101 ranks and collects diluted pay without contributing anything to the ecosystem, is that correct?

First, I don't really think that would be necessary if we separated delegates from workers so that delegates only got a standard pay, and if the approval percantage of the 101th delegate was high enough. The potential gain from voting oneself into the 101th position to collect a low/medium pay (just enough for server costs that they would not bother actually paying for) for a long enough time before people voted the delegate out would be negligible compared to the amount of stake necessary to vote the delegate into the 101th position in the first place. And if they are doing a good job to not get voted out, then they are a good delegate and there is no issue. The separate workers would have their own quorum requirements to get hired, and those could be sufficiently high enough that we shouldn't realistically have to worry about a rich attacker using that as a method to get "free money" for no work.

I also think that the two week fee wouldn't really be necessary even with the coupled delegate/worker system we have today, if we simply changed the pay mechanics for the delegates such that they needed substantial approval (meaning more than the percentage of stake to take over the network, i.e. the approval percentage of the 51st delegate) in order to receive a pay substantially larger than the basic salary necessary to just run a block signing node. I gave an example of such a delegate pay system in this thread.

But this proposal is not to argue for either of the above two systems but rather a much simpler system. I think that there should only be a small fixed fee to register as a delegate, which should not depend on the requested pay rate. However, the delegate would be required to produce their first 1200 blocks (corresponds to a period of time approximately equal to 2 weeks) without getting any reward. Actually, I would prefer if the number was instead 600 blocks (or 1 week) since I think 2 weeks is excessive. These blocks would have to be produced over a period of time in which the delegate remained an active delegate. If the delegate ever dipped below rank 101, the count would be reset. After the necessary number of blocks were produced by the delegate, their probationary period would end and they would then be able to collect the requested reward for each valid block they produce from that point forward. A slightly alternative system would be to pay the delegate a basic reward, e.g. 3% max pay, instead of no pay during the probationary period. Then after the probationary period, they would get their full requested pay.

The main purpose of this proposal is to prevent delegate candidates from losing substantial funds if they fail to ever get enough approval to become an active delegate. This lowers the barriers to entry for potentially great workers who want to earn dilution pay as a delegate but don't want to risk it since they are not sure if they can get enough stakeholder support. However, because of the probationary period, it also means that anyone with a lot of stake cannot profit by voting their 100% pay delegate into the top 101 to collect some "free" money before the stakeholders realize what is happening and vote for enough honest delegates to kick out the bad delegate. The probationary period means that there isn't any dilution damage done during the 2 week (or even 1 week) period that stakeholders need to react to such attacks. And ideally, experiences from such attacks would teach stakeholders to keep the 101th delegate approval percentage high enough so that such attacks are not even possible in the future.

BTW, on a different but somewhat related note, I don't think RDPOS should vote for a >3% pay delegate because they are in the delegate slate of someone you gave the thumbs up to unless you also explicitly gave a thumbs up to that >3% pay delegate. Of course, depending on a hard-coded fixed number like 3% is not a good idea. So a smarter implementation would be necessary. The easiest to implement solution would be to allow the user to just adjust that number in the preferences of their local client.