Author Topic: RDPOS - Recommended Delegated Proof of Stake  (Read 20031 times)

0 Members and 1 Guest are viewing this topic.

Offline Empirical1

  • Hero Member
  • *****
  • Posts: 884
    • View Profile
RDPOS has really grown on me, now that I get it more, really like it now. 

Also impressed issue was seen so quickly and addressed. 

Far too hard for the average  shareholder to handle a very large selection themselves and considering many shareholders will be customers even more so. So good.  +5%

Offline Empirical1

  • Hero Member
  • *****
  • Posts: 884
    • View Profile
Ten Most Trusted (Kind of a variation of RDPOS)

Up to 101 unique trusted delegates are needed to decentralise & secure the system. However the task of trying to vet, choose and monitor a large number of delegates is an incredible challenge for the average active shareholder never mind the apathetic shareholders who just want the system to work.  This makes it a significant attack vector.

Premise The ten delegates with the most approval have more trust from the shareholders than the delegates with lower approval.

Conclusion They can be trusted with more responsibility than the other delegates.

- What if a normal delegate only got transaction fees but additional fees (& in future equity release) were only distributed to the ten most trusted delegates?

- What if only the top 10 were able to have a slate of up to 20 delegates (But the system doesn't cast the vote for anybody on their slate already in the top 15.)

Then we as a community  can primarily focus on, promote, debate & discuss who are the ten most trusted and best to represent the DAC.  We can make sure we particularly hold them to account and keep negative influences out  (Tracking ten peoples slates, spending & their development/marketing strategies is easier than 101 which is too hard & noisy for us to focus on as individuals let alone as a group.)

The idea is not to give them control of the system but a greater influence and responsibility because they are more trusted and the task of managing/monitoring large numbers of delegates is not realistic for the average shareholder.
 
However regular delegates can still campaign to regular users as well as be incentivised by fees.
(& have a fair shot at becoming a  'Ten Most Trusted' themselves - because positions 1-15 can't be influenced via the slates of the incumbents.)

« Last Edit: July 26, 2014, 06:37:18 pm by Empirical1 »

bitbro

  • Guest
What if delegates have to register to a "type" or "political party"... And then anyone who doesn't want to vote for 101 delegates can choose to vote for a "type" or "politcial party," however this would not be a vote for a set slate, instead a vote for a political party is a vote for any delegates randomly selected from that political party

Edit: the key difference in this proposal is the randomly selected aspect of the slate

Edit: example: if I want 50 of my votes going to the lowest bidding delegates then I choose the Low Pay Party for 50 votes.  50 random Low Pay Delegates then get those votes. 

I also want 51 Above Low Pay Marketing Delegates to get votes.  I vote for that party and 51 randomly selected Above Low Pay Marketing Delegates get a vote.

Edit:  these parties could be maintained and regulated by formula or by a third party





Sent from my iPhone using Tapatalk
« Last Edit: July 26, 2014, 05:36:50 pm by bitbro »

Offline Bitcoinfan

  • Sr. Member
  • ****
  • Posts: 240
    • View Profile
This will follow the progression similar to indexing stocks-- S&P, Russell 2000, Dow Jones, mutual funds, I think indexing of delegates is going to happen by someone else if it isn't done here.   

« Last Edit: July 26, 2014, 04:06:21 pm by Bitcoinfan »

Offline puppies

  • Hero Member
  • *****
  • Posts: 1659
    • View Profile
  • BitShares: puppies
I'm still not sure how I feel about this.  Still working to wrap my head around it.

It seems to me that this is a consequence of allowing variable pay rates for delegates.  It dramatically increases the difficulty of picking delegates.  If we want to make voting easier for the end user, perhaps simplifying the choice would do more than further politicizing it.  If all we had to watch was reliability, and attack vectors, most of that could be automated and enabled in the wallet by default. 

An idea for an alternate way of funding growth of the dac would be a per transaction donation broadcast that adds .01 or whatever amount broadcast to a person you believe will use the funds in a manner you deem worthy.  This could also be enabled by default in wallets.  If its easy enough and spread out enough, people will be willing to do it.  I would suggest a clear sign that you are currently donating when you make transactions, but then you could sign up by clicking a link, just like a voting slate. 

Of course there is the argument that people will not be as willing to spend their own money on a project of their choosing, if they are not able to spend other peoples money at the same time.  Its a bit of a stretch but I think it applies.  Kinda
Quote from: Lysander Spooner
The ostensible supporters of the Constitution, like the ostensible supporters of most other governments, are made up of three classes, viz.: 1. Knaves, a numerous and active class, who see in the government an instrument which they can use for their own aggrandizement or wealth. 2. Dupes – a large class, no doubt – each of whom, because he is allowed one voice out of millions in deciding what he may do with his own person and his own property, and because he is permitted to have the same voice in robbing, enslaving, and murdering others, that others have in robbing, enslaving, and murdering himself, is stupid enough to imagine that he is a “free man,” a “sovereign”; that this is “a free government”; “a government of equal rights,” “the best government on earth,” and such like absurdities. 3. A class who have some appreciation of the evils of government, but either do not see how to get rid of them, or do not choose to so far sacrifice their private interests as to give themselves seriously and earnestly to the work of making a change.

https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline alexkravets

  • Full Member
  • ***
  • Posts: 81
    • View Profile
Perhaps a naive suggestion to address this. 
What about a small area inside the blockchain metadata for this ?

If the blockchain itself records delegate ID's and a few small past-performance parameters ?

The blockchain already has information about the missed blocks. I think this is enough as a start, I'm building some tools to help visualize that.

Enough for which purpose ? If entire security and performance of DPOS depends on behavior of delegates, seems like keeping a trust-free historical audit inside the blockchain would be useful.

One use was already pointed out in this thread, which is to allow defaulting to a sound voting slate of most reliable delegates sporting the smallest pay rate ?

Offline bitmeat

  • Hero Member
  • *****
  • Posts: 1116
    • View Profile
Perhaps a naive suggestion to address this. 
What about a small area inside the blockchain metadata for this ?

If the blockchain itself records delegate ID's and a few small past-performance parameters ?

The blockchain already has information about the missed blocks. I think this is enough as a start, I'm building some tools to help visualize that.

Offline alexkravets

  • Full Member
  • ***
  • Posts: 81
    • View Profile
What happened to the idea of having automatic votes based on past performance as default?

Light weight nodes have no means of independently verifying past performance and even evil nodes can appear to be performing well.

Perhaps a naive suggestion to address this, given my ignorance of availability of the required data storage area in the consensually set common "logical" dabase.

What about adding a small area inside the blockchain header/metadata area for this ?

If the blockchain itself records a table of delegate ID's mapped to a few small cumulative past performance metrics ?

Sorry I'm not familiar with either physical or logical lay out of the BitShares Blockchain, but the corresponding area inside Ripple ledger's header would be where such info would be placed inside Ripple.
https://ripple.com/wiki/Ledger_Format#Header_Format.

This arrangement requires no client trust as it would be subject to the usual DPOS consensus algorithm.




« Last Edit: July 26, 2014, 08:09:34 am by alexkravets »

Offline AdamBLevine

  • Sr. Member
  • ****
  • Posts: 492
    • View Profile
    • Let's Talk Bitcoin!
Well, I won't be voting for any delegate that publishes a "slate" of other delegates to vote for.  Delegates should be campaigning for the approval of shareholders not campaigning for the approval of a cabal of other delegates who all add each other to their slates.

I encourage any shareholder that cares about this stuff for the long term to take the same stance.

Cabals are going to happen with or without slates.

I can see by including this function, it might quickly encourage someone to provide a slate "service".  Someone who does thorough research into delegates and has strict standards of accountability/transparency.  concentrating the effort to find the best, most reliable parties to be made delegates into 1 person/group so that we, lazy users, don't need to bother with it.  Because most of us won't anyway.  It leaves the average user who's even slightly interested in voting with a much smaller burden of inspecting a hand-full of slate servicers, rather than a countless myriad of delegates.  Of course, for those of us who wish to do the research ourselves are free to do so, and might even lean on these slate servicers to help pick and choose our own delegates who are up to our standards/interests.

I don't see how not having slates prevents people from forming their own manual slates anyway if that's what they want/serves their interest.

I would rather say, I won't be supporting a slate/slate provider that's full of members all representing a cabal.  I'm interested in seeing a slate that's highly diverse representing a wide variety of groups from all over the globe with a diverse representation of interests.  All of which can be tracked and varified.  It will be up to these slate servicers to provide a clear, thorough, and professional representation of proposed delegates to earn my trust in using that slate or some protion of it.

Yup.
Email me at adam@letstalkbitcoin.com

Offline gamey

  • Hero Member
  • *****
  • Posts: 2253
    • View Profile

I would rather just have a list of 3-4 trusted slates and have them all combined.  The delegates with the most weight win. 

I still agree with Agent86 about this screwing up incentives.  I've come to believe it may be a necessary evil to keep voter participation up with a meaningful 101 votes.
I speak for myself and only myself.

Xeldal

  • Guest
Well, I won't be voting for any delegate that publishes a "slate" of other delegates to vote for.  Delegates should be campaigning for the approval of shareholders not campaigning for the approval of a cabal of other delegates who all add each other to their slates.

I encourage any shareholder that cares about this stuff for the long term to take the same stance.

Cabals are going to happen with or without slates.

I can see by including this function, it might quickly encourage someone to provide a slate "service".  Someone who does thorough research into delegates and has strict standards of accountability/transparency.  concentrating the effort to find the best, most reliable parties to be made delegates into 1 person/group so that we, lazy users, don't need to bother with it.  Because most of us won't anyway.  It leaves the average user who's even slightly interested in voting with a much smaller burden of inspecting a hand-full of slate servicers, rather than a countless myriad of delegates.  Of course, for those of us who wish to do the research ourselves are free to do so, and might even lean on these slate servicers to help pick and choose our own delegates who are up to our standards/interests.

I don't see how not having slates prevents people from forming their own manual slates anyway if that's what they want/serves their interest.

I would rather say, I won't be supporting a slate/slate provider that's full of members all representing a cabal.  I'm interested in seeing a slate that's highly diverse representing a wide variety of groups from all over the globe with a diverse representation of interests.  All of which can be tracked and varified.  It will be up to these slate servicers to provide a clear, thorough, and professional representation of proposed delegates to earn my trust in using that slate or some protion of it.

« Last Edit: July 25, 2014, 10:54:37 pm by Xeldal »

bitbro

  • Guest
Won't it be best to ultimately have only a few slates with clear intentions? Or would everyone ultimately converge on one majority-having slate?  If that happens, it's pretty clear who the delegates should be paying under the table.  Should the US govt create a slate? Should China create a slate?  So many wow. Questions.


Sent from my iPhone using Tapatalk

Offline Agent86

  • Sr. Member
  • ****
  • Posts: 471
  • BTSX: agent86
    • View Profile
Well, I won't be voting for any delegate that publishes a "slate" of other delegates to vote for.  Delegates should be campaigning for the approval of shareholders not campaigning for the approval of a cabal of other delegates who all add each other to their slates.

I encourage any shareholder that cares about this stuff for the long term to take the same stance.

Offline bitmeat

  • Hero Member
  • *****
  • Posts: 1116
    • View Profile

Offline vikram

So that everyone is aware, in the latest development version, the standard wallet_transfer command currently uses the vote_recommended strategy by default: https://github.com/BitShares/bitshares_toolkit/compare/1ca979a7e6...3cb1d0f820