Author Topic: Proposed Future DAC Delegate Pay Model  (Read 24726 times)

0 Members and 1 Guest are viewing this topic.

Offline GaltReport

I have to admit that I somewhat dislike the proposed system.
Are there any advantages, that being a delegate and getting funding is so closely entangled? I think it would be better to leave the delegate pay system as it is in BitsharesX right now and implement a more or less independent funding feature. Your proposed system has multiple disadvantages:

1. You need to be a delegate to potentially get funding. Especially in the beginning it is a big hurdle to set up and maintain a delegate node.
2. There is a limited number of delegates. If there are many smaller projects this will probably be a bottleneck. Moreover this might have the effect that projects that aren't finished get voted out just because of the shortage of active delegate positions.
3. There are competing criteria when selecting delegates. It is not only about the trusworthyness and ability of an individual to maintain a delegat node but also independently the plans for projects that need funding. That might lead to an overall decreased blockchain security and stability.
4. Once high payed delegates get voted in they are not incentivised to finish on minimal budget and time. Getting voted out again is the only threat and in my opinion not sufficient to motivate the delegate. This is a general problem of every solution that isn't timely restricted.

I tend to agree with this.  Not sure why being a delegate (paid miner) is related to funding and DAC development.  I'm guessing so that the Music DAC can more easily raise funds to actually make their blockchain a Music DAC?  I'm probably confused.
« Last Edit: September 29, 2014, 01:25:40 pm by GaltReport »

Offline blahblah7up

  • Full Member
  • ***
  • Posts: 192
    • View Profile
This proposal would harm early adopters and AGS donors the most.  And frankly, I don't think there would have been many AGS donors if this model had been revealed as the desired model early on.  It certainly moves the goalpost.

Dilution is also the most cited complaint I have heard from people who chose not to donate to Etherium.

Offline Empirical1.1

  • Hero Member
  • *****
  • Posts: 886
    • View Profile
I don't like dilution but if there was dilution at least 51 delegates should be required to approve it or the dilution request should be made well in advance and follow a clear set of procedures and be properly advertised in the client so shareholders can take a position in advance on it.

Example: Delegate.Max issues 14 day notice of a request for 0.01% dilution paid via his delegate for the following program 'Billboard advertising (link)'

Then you have some kind of voting procedure, even if it's as simple as  shareholders removeing approval from Max and getting him out of the 101 delegates. (He will see his approval going lower and will adjust and re-issue his proposal or remove it if there is a negative response.) 

However I don't think dilution should ever be necessary. The initial business plan should set aside a pool of shares that can be used to pay for programmes and expenses using a shareholder voting system. This should fund the development of the DAC until such time as fees are sufficient to cover it.

For BTSX

I think delegates should be paid from  fees but I don't think we should be immediately burning the remaining fees not going to delegates. They should go into a separate pool, then every 4 months we have a 'shareholder meeting' & vote on how to spend those fees. Shareholders can either vote

Burn
Roll-over (Keep fees in pool and roll them over to next 4 month shareholder meeting.)
Fund (Elect to fund 1 or more of the initiatives put down at least 14 days in advance of the shareholder meeting by keeping the delegate proposing it in office.)   

Edit - For AGS

Not exactly related, but off the top of my head I think AGS needs to be made liquid for one month a year, say January to bring in new money and widen it's shareholder base and prevent it from being a closed club.

I think a pool of shares should have been set aside initially for the long term development and promotion of AGS but if they weren't, then it is a case where dilution may be needed and leading up to the liquid month, wecould perhaps vote on a dilution %. So perhaps the majority will agree on a 5% dilution to be given to I3 to promote and develop AGS for the next year for example based on the proposal they give shareholders on what they think they will be able to achieve etc.
« Last Edit: September 29, 2014, 12:29:33 pm by Empirical1.1 »

Offline Frodo

  • Sr. Member
  • ****
  • Posts: 351
    • View Profile
  • BitShares: frodo
I have to admit that I somewhat dislike the proposed system.
Are there any advantages, that being a delegate and getting funding is so closely entangled? I think it would be better to leave the delegate pay system as it is in BitsharesX right now and implement a more or less independent funding feature. Your proposed system has multiple disadvantages:

1. You need to be a delegate to potentially get funding. Especially in the beginning it is a big hurdle to set up and maintain a delegate node.
2. There is a limited number of delegates. If there are many smaller projects this will probably be a bottleneck. Moreover this might have the effect that projects that aren't finished get voted out just because of the shortage of active delegate positions.
3. There are competing criteria when selecting delegates. It is not only about the trusworthyness and ability of an individual to maintain a delegat node but also independently the plans for projects that need funding. That might lead to an overall decreased blockchain security and stability.
4. Once high payed delegates get voted in they are not incentivised to finish on minimal budget and time. Getting voted out again is the only threat and in my opinion not sufficient to motivate the delegate. This is a general problem of every solution that isn't timely restricted.

Offline svk

So if I understand this correctly:
  • Transaction fees remain as they are now but will be separate from delegate fees
  • 100% of transaction fees will be burnt
  • Delegates will only be paid through dilution
While I like the idea I too feel like arhag that it's open to abuse, but this can probably be avoided through the tuning of certain parameters. For starters you can probably set the registration multiplier even higher, x300 and ~3.3 days to break even doesn't seem unreasonable to me, the ROI is still incredible.

While this may be more complicated, how about adding a time duration to the increased delegate pay? Say I want to develop a new website for bitshares Music, I propose an increased pay rate for 3 months to fund my development, after which my payrate returns to a standard rate. This would require active voting on the proposal however.
Worker: dev.bitsharesblocks

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
I have been working through the delegate pay model that I recomend for future DACs including VOTE, MUSIC, DNS, etc.

It works off of the principle that shareholders have a right to dilute to raise capital and that the ability to dilute to raise capital ultimately maximizes shareholder value by allowing rapid growth through many rounds of funding.

1) 100% of all fees are burned. 
2) Delegates can set any fee per block they like measured in XTS (BTSX, VOTES, etc)
3) The cost to register a delegate with a "high fee" is equal to 100x the fee-per-block they want to get paid. 

With this system a delegate must produce at least 100 blocks to break even.  It also naturally limit potential abuses because the cost of registering a HIGH PAY delegate will be VERY HIGH while the probability of getting elected will be very low.  The person registering the HIGH PAY delegate will also lose out on the ability of using their own shares to vote for themselves. 

For starters no one would be able to register a delegate that would pay 2 billion BTSX per block because the cost would be 200 billion BTSX (which don't exist).   This puts a limit to what one could attempt to register to be 20 million BTSX per block... assuming they owned all 2billion BTSX... if we assume the largest stakeholder owns 10% of BTSX and the use 100% of their stake to register a high paying delegate then the maximum pay per block would be 2 million and they would have no stake left to vote for themselves.   

Assume someone used 1% of the BTSX to register their delegate at a pay rate of 200K BTSX per block it would take them over 28 hours to break even during which the rest of the shareholders could coordinate their votes to get them ousted if they didn't consent.

With this setup the DACs can fund development in a decentralized manner for their entire lifetime. 

One thing I would like to mention is that there is a "HARD" limit on the share supply based upon limits of 64 bit numbers and the precision.

My gripe with dilution is it equates to taking a loan. It's growth which relies on debt.

Sometimes corporations need to do this but I think in order to do this there should be a vote and shareholders should be able to set the dilution rate. I don't think dilution should happen by default because it makes it seem like the developers had no business plan in the first place if they didn't factor in paying for continued development.

I still prefer simply raising the transaction fees. I think inflation opens the door to collusion, corruption, and the same sort of problems associated with mining. So now instead of paying miners via dilution/inflation we'll be paying delegates?

Given a choice to support DAC A which has no dilution but all the same features and DAC B which has dilution, it would be rational for me to choose DAC A.

The only time it would be rational for me to choose DAC B is if the developers lay out  comprehensive roadmap, business plan, and can convince me that dilution can increase the utility value of the asset. I think in the case of Bitshares Music, Bitshares Vote and others like it, there is no guarantee that these DACs will ever catch on. Asking for dilution before we have any statistics showing us the rate of growth would in my opinion not go down well with investors.

But if it can be shown that steady growth is happening, adoption is happening, that the idea is sound, then it definitely could make sense to improve on the software as a way to protect the utility value of the DAC from competitors who will make clones/forks. It only makes sense though if adoption is steady.

Right now we don't know the rate that users are adopting Bitshares X. Looking at Bitcoin the adoption rate seems to be a problem. It's not yet shown that people will adopt Bitshares X or the whole Bitshares paradigm.
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline zhao150

  • Hero Member
  • *****
  • Posts: 606
  • 老子早就不想当代表了
    • View Profile
不能任由代表设置自己出块的收费,否则票多的大代表,没人能阻值。
关于具体如何限制代表设置的出块收费,我提一个建议:
每个代表设置的收费不能大于所有代表设置的收费的平均值的2倍。
这样可以减轻代表的贫富差距。
老子早就不想当代表了

Offline xfund

  • Full Member
  • ***
  • Posts: 148
    • View Profile
    • FUND投票基金
  • BitShares: xfund
I think this is a good idea to help the delegate to dig coin in dns .  But don't too bitter.
Asset:FUND

Offline arhag

  • Hero Member
  • *****
  • Posts: 1214
    • View Profile
    • My posts on Steem
  • BitShares: arhag
  • GitHub: arhag
Based on current fee revenue, 0.22% of a 2 billion BTSX share supply would be burned yearly. At current prices, that is $1400 per year for each of the 101 delegates. Let's say the delegates are okay with that level of pay for now (they are actually okay with less currently since the average pay rate is 85%). In this case, all the fees would go to the delegates to pay for expenses. Now let's pretend one delegate wants to make a lot of money. If they put down x BTSX as the registration fee, and they get enough approval to become and stay an active delegate for a year, they could earn up to 312*x BTSX per year. Define p = 100 * x / (2 billion), the percentage of the share supply the delegate is willing to spend as a fee. Then the delegate can earn up to (3.12*p)% of the 2 billion BTSX share supply every year. Ignoring the small effects, such as 0.00217% of the delegate's earning being double counted and ignoring the increase in fees due to putting down the large delegate registration fee, it is fair to say that the inflation of the share supply is approximately (3.1*p)% per year (at least for the first few years) due to this delegate getting elected. So what does it take for this delegate to get elected? Well that is complicated to answer properly since the fact that this delegate will be inflating the share supply will likely reduce voter apathy and rally people to start voting for other delegates. But assuming no changes in voter behavior and the current delegate approval ratings, currently it only takes more than 5.25% of the stake to vote one delegate in. So, if a colluding group had s% of the stake, they could burn all but 5.25% of the stake to get their delegate in and inflate the BTSX supply by (3.1*(s-5.25))% per year (unless the shareholders reacted and raised the approval rating bar for delegates). Under this proposal, a colluding group with 8.5% of the stake could (under current conditions) possibly inflate the BTSX supply by 10% per year and reap the rewards of that inflation. Even if the minimum delegate approval was 20%, a colluding group with 25% of the stake could still inflate the share supply by 15% per year. The only way to avoid the situation where a minority of shareholders are able to inflate the BTSX share supply against the wishes of the majority is if the minimum approval rating of a delegate is above 50%. Perhaps this proposal is smart as a way of killing voter apathy and getting shareholders involved for the sake of increasing delegate approval ratings and thus improving blockchain security, but I worry it could have the reverse effect where the minimum delegate approval rating stays low and a minority of stakeholders have the ability to dilute the shares against the wishes of the majority.

My view is that we should always require at least 50% approval by the shareholders when determining the maximum rate of inflation of the BTSX supply. Under the system I propose, which is delegate proposal ratified by shareholder vote, we can have different levels of shareholders approval needed to ratify the rules of a proposal depending on what rules the proposal intends to change. For example, we could require at least 50% shareholder approval to change the per block inflation rate cap. This cap would limit the effect of any other proposals so that ((amount of BTSX created to pay workers in a block) - (amount of BTSX allocated for destruction for that block))/(amount of BTSX in existence) is always less than or equal to (per block inflation rate cap). Then other proposals could have less strict requirements for ratification. For example, a proposal that changed the pay rate of the delegates or other workers (denominated in BitUSD) could only require approval by min(50%, max(delegate approval rating)) percent of shareholder vote, or maybe just a fixed percentage like 20%. You could even set up rules such as at least 15% of shares need to have voted AND the net approval, ((shares approving of proposal) - (shares disapproving of proposal)), would have to be larger than 15% * (100% - (percentage of shares voting)).
« Last Edit: September 29, 2014, 03:19:53 am by arhag »

Offline BTSdac

  • Hero Member
  • *****
  • Posts: 1219
    • View Profile
  • BitShares: K1
it is like a decentralized company , each delegate is the staff of this company , if consider the pay fee as salary to delegate. it like every staff can offer him salary by himself propose. and all stakeholders can accept or reject it proposed salary
but there are two questions
1. 28 hours is too short ,maybe 90 days is enough, it is like company give financial report per 90 days.
2. if DAC is a company , works delegates do in chain (create block ,broadcast,transfer block etc) is`t the whole works of this company.

« Last Edit: September 29, 2014, 03:21:36 am by BTSdac »
github.com :pureland
BTS2.0 API :ws://139.196.37.179:8091
BTS2.0 API 数据源ws://139.196.37.179:8091

Offline Overthetop

I cant really understand the  meaning of the new solution.

Could anybody or BM help answering the following questions :

1. what the purpose of this New Model of Delegate? (the problem want to be solved)

2. How does the New Model work?(how to solve)

3. What is the desired effect of this New Model? (the advantage of the new solution)

Thanks.

« Last Edit: September 29, 2014, 02:47:04 am by Overthetop »
个人微博账号: Overthetop_万里晴空
“块链创新与创业”交流群: 330378613

Offline BTSdac

  • Hero Member
  • *****
  • Posts: 1219
    • View Profile
  • BitShares: K1
I have a suggestion
1.if resume  max translations per block is N,   fee per translation is 0.1,  exclude the translation fee ,the delegate also can get some fee for creating block ,but this fee is calculation by formulation
addtional fee =(N-N_tx)*R*0.1
N: max translations per block
N_tx: translations of this block
R: a ratio   0<R<1
0.1 : this is equal as fee per translation

so the total fee = N_tx*0.1+(N-N_tx)*R*0.1

I reume R=0.25, N=100
1. if it is a empty block (no translation)  the fee giving to delegate = 100*0.25*0.1=2.5
2. if it is a full block( include Max translation) the fee giving to delegate =100*0.1=10

you can see  delegate can gain some fee even it is a empty block, and if the delegate want to gain more fee ,they must try to include more translations
as the following picture show ,the horizontal mean how many translations in the block , and the vertical mean fees delegate gain.
there are five sulation ,they are R=0.1 , R=0.25 ,R=0.5 , R=0.75, R=1  .

« Last Edit: September 29, 2014, 02:00:25 am by BTSdac »
github.com :pureland
BTS2.0 API :ws://139.196.37.179:8091
BTS2.0 API 数据源ws://139.196.37.179:8091

Offline BTSdac

  • Hero Member
  • *****
  • Posts: 1219
    • View Profile
  • BitShares: K1
pay fee to delegate directly , how to motivate delegate include more translation ?

Votes and approval.
how to know a translation should been included by a delegate.   time-stamp of this translation , when translation been broadcast ? there have network delay.
I mean the stakes cannot know if this delegate is a positive one
github.com :pureland
BTS2.0 API :ws://139.196.37.179:8091
BTS2.0 API 数据源ws://139.196.37.179:8091

Offline muse-umum

  • Hero Member
  • *****
  • Posts: 717
  • BitShares everything
    • View Profile
Assume someone used 1% of the BTSX to register their delegate at a pay rate of 200K BTSX per block it would take them over 28 hours to break even during which the rest of the shareholders could coordinate their votes to get them ousted if they didn't consent.

Image this scene:

Several big whales together control huge amounts of shares (far more than 1%), which makes it difficult to vote them out. After burning 10,000,000 XTS to register a delegate with 100,000 XTS per block paid they probably still have enough shares to vote that delegate in.

If that happens the 100,000 XTS would be the dilution putting investors off.

This is just a simple case. We may face many other problems with this pay model.
« Last Edit: September 29, 2014, 01:53:02 am by heyD »

Offline zhao150

  • Hero Member
  • *****
  • Posts: 606
  • 老子早就不想当代表了
    • View Profile
can I set 1000000btsx per block fee?
then I will give a lot of btsx to everyone.
I will be a philanthropist. :)
老子早就不想当代表了