Next Committee Proposal: Witness Pay

[8ejM]

Hi,

I new to Bitshares, so I may fully informed about everything.  I started buying BTS a few days ago, mainly because of bitUSD. I read through the white papers and were especially excited about the pegged assets. I went on an bought a bunch of BTS. Some days later I wanted to buy some bitUSD and bitCNY. I was and still am very confused about market prices below settlement price since I red in the white paper that pegged assets are expected to be traded at at least the price of the underlying asset.
My excitement for bts took a hit, since it didn't seem to behave as outlined in the white paper and it seemed that a lot more market knowledge was needed to buy at a reasonable price and don't get cheated.
Then I learned about the page where pricefeeds where shown and was shocked that feeds are not updated in manners of seconds but minutes, hours, days and even weeks. I saw forced settlements and witnessed a change in settlement price just a minute later. This made me very suspicious.
Then a learned about complaints about witnesses that don't update the price feed and are still official sources of price data.

At the end I stopped trading and will wait till I gathered enough information about the market. I simply don't really trust the data at the moment.

Why am I telling you all this? Because I think its a typical example for someone new to bts. And some will just leave when they feel that something is weird. It is very important that problems like proper price feeds are fixed. When the cryptomarket sees a correction people will give BTS as a possibility for cashing out in fiat a try. If by then these problems are not fixed. We will loose a huge amount of potential users, that never will come back.

If the problem can be solved by throwing more money at it, take both hands a throw as lot money as needed.

A more modular approach might also be a way to achieve better results. I don't know about all tasks of a witness. But would price feed data a task of its own, one could easily drop everybody with poor data quality from the list. Removing bad quality data sources is something a algorithm can do well, you would not have to start voting campaigns on the forum for that. 

[pc]

Thanks for the explanation. I didn't read that far. :-/

That means steem pays 25*0.189 STEEM per 63 seconds (one round), or 25*0.189 / 63 * 86400 * 30 * 0.562 = 109,252.80 USD/month.

[abit]

Thanks for starting this discussion.

One remark - I think steem witnesses are being paid like 0.94 STEEM/block, not 0.189.

The backup witnesses in Steem get paid 5x per block in comparison to the top witnesses. They produce much less blocks per day though.

[pc]

Thanks for starting this discussion.

One remark - I think steem witnesses are being paid like 0.94 STEEM/block, not 0.189.

To answer your questions:

1- Should we consider how much the witness "ideal" guy is worth on the real world? (ie. how much a (even Junior) SysAdmin gets paid)

Yes, BUT running a witness is far from a full-time job. When I was a witness I was happy with the payment, despite the much lower rewards and BTS price. I had a few night shifts in the early days of the chain, but since then the node was mostly running on its own.

Admittedly, the work load on witnesses is somewhat higher today, not so much from running a node but mostly from the price feeds. Apparently feed sources are unreliable at times, and there is a certain lack of consensus on what the "correct" price for certain assets is.

2- Should we consider the 24/7/365 requirement aspect of the witness job? (you can sleep, have week-ends and holidays, but you still have to "go to the office" in a timely manner no matter what the day and time is)

The problem here is that we (the shareholders) cannot enfore what rules we impose on witnesses. We can *expect* a witness to respond in time, but we cannot claim damages if he doesn't. It is therefore unreasonable to pay them as if it was a regular 24/7 job.

It has always been my understanding that market rules apply here - offer a reasonable payment, then let witnesses apply for the job and fire those who deliver bad service. Increase payment offer if you don't get a sufficient number of acceptable quality candidates.

3- Should we consider how much other blockchains spend (invest?) for their own security? (ie. how much BTC, ETH, etc spend, compared to market-cap) (even other dpos like STEEM and soon PEERPLAYS)

Yes, BUT the fact that the cost for security is much lower in DPOS than in POW chains has always been highlighted as a big pro for BitShares. So don't let us drop that.

4- Should we consider the current profitability of the DEX DAC?

No. The chain is still running far below its saturation level. At this point it is important to attract more users. Sacrificing service quality for a minor increase in profitability is a bad idea IMO. (Remember that witness pay is a relatively small part of the current spendings).

Maybe more tough (and subjective?) questions:
5- How much is the whole blockchain task, security and decentralization worth? (this is what a witness really do, from a pure blockchain pov) (looking at point 3 may help)

IMO the best judge for this is the market. :-) Obviously, the service provided by witnesses is essential for the existence of the blockchain. But that doesn't mean we should all hand over our holdings to the witnesses.

Remember that BitShares is a DAC, and whatever the C stands for, it will try to get the best possible service from witnesses for the least possible cost.

6- Should 4, and particularly 4b, be able to overcome 5?
n/a

[Thom]

If instead you want to decide what a witness pay should be without consider what his job is really about and what it is worth for the blockchain etc, then I am not interested in the discussion...

Thanks Bhuz for taking the time to post this. I sure hope we see more discussion and input from a wider audience.

So far I don't see any informed comments. No appreciation for the Witness job, and only the "Politics of Envy". Lot's of emotionally charged opinion but not many facts.

I only hope the shareholder collective will realize just how important the Witness role is, and make informed decisions concerning Witness pay.

I could reiterate many of the points I've raised in other posts concerning pay, but I will not repeat myself now, but rather I will wait to chime in after more informed comments appear here, if they do.

I am not here to defend the current pay level of Witnesses, despite being a Witness. I believe the pay is higher than it needs to be. I will withhold my opinion as to what I believe a reasonable pay level should be. It will be more than pure altruism however, I will say that! I had enough of that for almost an entire year!

I gotta say, some of the remarks here infuriate my sensibilities!

[Bhuz]

Expense ratio has no relationship to market cap indeed. No matter how much the price is, reward per block per month / total supply is almost fixed in other crypto projects. But in BitShares we're able to adjust.

Code: [Select]

monthly_expense_ratio
= monthly_expense_rewards / market_cap
= (tokens_per_block * blocks_per_month * token_price) / (total_supply * token_price)
= tokens_per_block * blocks_per_month / total_supply


You are right.
I still wanted to show USD values tho, since that was the actual thing that started the complaints about the witness pay


Inviato dal mio Nexus 5X utilizzando Tapatalk

[abit]

Since the last Committee Proposal went through (fee-adjustement), @xeroc remembered and asked the Committee tu discuss about the witness pay, and more precisely what is a rational payment for a witness job
The following is the reply I wrote on the telegram committee group and, as suggested by @abit soon after, I am now posting it here too, for a wider discussion with the community

******
These are just inputs... there surely are more questions that may be worth asking and answering... please comment and discuss them, modify or add your own

If instead you want to decide what a witness pay should be without consider what his job is really about and what it is worth for the blockchain etc, then I am not interested in the discussion... I will vote for whatever the committee or shareholders decide, but probably without giving any meaningful input or personal thought


Questions we should answer:
1- Should we consider how much the witness "ideal" guy is worth on the real world? (ie. how much a (even Junior) SysAdmin gets paid)
2- Should we consider the 24/7/365 requirement aspect of the witness job? (you can sleep, have week-ends and holidays, but you still have to "go to the office" in a timely manner no matter what the day and time is)
3- Should we consider how much other blockchains spend (invest?) for their own security? (ie. how much BTC, ETH, etc spend, compared to market-cap) (even other dpos like STEEM and soon PEERPLAYS)
4- Should we consider the current profitability of the DEX DAC?
   4a- If so, should it be able to greatly increase what the points 1,2,3 would suggest?
   4b- If so, should it be able to greatly decrease (or even reset) what the points 1,2,3 would suggest? (ie, no profit for the DEX == no profit for the witnesses --> witness pay = costs of running a witness)

Maybe more tough (and subjective?) questions:
5- How much is the whole blockchain task, security and decentralization worth? (this is what a witness really do, from a pure blockchain pov) (looking at point 3 may help)
6- Should 4, and particularly 4b, be able to overcome 5?

Plus:
IMHO we shouldn't start from a number to later check if that is sustainable for a witness, but rather start from what an adequate pay for the job is, and only later set the pay...
So, witness pay = job pay + costs (adequate/deserved witness net income + costs that the job itself requires)
******


PS:
BTW, it's basically one year that I pay my BTS witness costs thanks to STEEM... if it was for BTS itself I wouldn't have earned nothing, or worst I would be at loss...
I hope this will be read and understood by all those people that were jumping and screaming around as soon as bts (finally) rose in value, to point the finger to the witnesses that, suddenly, were "earning too much" thanks to their "unbelievable, unaccettable and absurd pay"
I also hope this will be considered for choosing an adequate BTS witness pay


_ _ _ _ _ _ _ _ _ _ _ _ _ _
Some data about point 3 at the time of this post, from coinmarketcap, considering current rewards
BTC:
Market-cap: $25,617,582,661
Price: $1570.05
Monthly blockchain expenses: 12.5(btc/block) * 144(blocks/day) * 30(days/month) * 1570(usd/btc) = $84.780.000
BTC expenses / BTC market-cap --> 84.780.000 / 25.617.000.000 = 0.0033

STEEM:
Market-cap: $132,000.000
Price: $0.562
Monthly blockchain expenses: 0.189 (steem/block) * 864000(block/month) * 0.562 (usd/steem) = $91.800
STEEM expenses / STEEM market-cap --> 91.800 / 132.000.000 = 0.0007

BTS:
Market-cap: $95,908,469
Price: $0.036894
Monthly blockchain expenses: 3600 (bts/hour) * 24(hours/day) * 30(days/month) * 0.037(usd/bts) = $95.900
BTS expenses / BTS market-cap --> 95.900 / 95.900.000 = 0.001

This means, compared to their market-cap:
BTS current expenses is less than 1/3 of BTC's expenses
STEEM current expenses is about 30% lower than BTS's current expenses
_ _ _ _ _ _ _ _ _ _ _ _ _ _


If interested in the matter, please comment, discuss, and share your thoughts. Thanks

Expense ratio has no relationship to market cap indeed. No matter how much the price is, reward per block per month / total supply is almost fixed in other crypto projects. But in BitShares we're able to adjust.

Code: [Select]

monthly_expense_ratio
= monthly_expense_rewards / market_cap
= (tokens_per_block * blocks_per_month * token_price) / (total_supply * token_price)
= tokens_per_block * blocks_per_month / total_supply


[CoinHoarder]

Witnesses should be paid nothing, zilch... nada.

In such a scenario, they will be compensated from the value rising due to the amount of BTS in circulation being lowered from burned fees and the token value rising due to demand staying the same (or increasing due to deflation).

$95,900 a month is ridiculous for  23 delegates. That is what... ~$4169 a month for doing nothing? Give me a break..

1- Should we consider how much the witness "ideal" guy is worth on the real world? (ie. how much a (even Junior) SysAdmin gets paid)

No, running a witness node is stupidly easy. Junior level SysAdmin skills are not a necessity. I could write directions easy enough to follow that my grandma could do it.

2- Should we consider the 24/7/365 requirement aspect of the witness job? (you can sleep, have week-ends and holidays, but you still have to "go to the office" in a timely manner no matter what the day and time is)

No. When was the last time witnesses had to do anything? They haven't even needed to update their nodes in what... over a year? lol. There is zero maintenance required due to the lack of non-GUI bitshares development.

3- Should we consider how much other blockchains spend (invest?) for their own security? (ie. how much BTC, ETH, etc spend, compared to market-cap) (even other dpos like STEEM and soon PEERPLAYS)

Since when has Bitshares cared about what other blockchains spend securing their blockchain? Especially when 99% of them are doing it wrong. That is one of the biggest selling points for dPoS... it is cost efficient.

4- Should we consider the current profitability of the DEX DAC?
   4a- If so, should it be able to greatly increase what the points 1,2,3 would suggest?
   4b- If so, should it be able to greatly decrease (or even reset) what the points 1,2,3 would suggest? (ie, no profit for the DEX == no profit for the witnesses --> witness pay = costs of running a witness)

No, it requires the same amount of work for witnesses no matter the profitability of the DEX.

Maybe more tough (and subjective?) questions:
5- How much is the whole blockchain task, security and decentralization worth? (this is what a witness really do, from a pure blockchain pov) (looking at point 3 may help)

Raising witness pay does not increase the decentralization of the dPoS consensus algo. Neither does lowering it, as you will always have the same amount of delegates.

Security with dPoS is obtainable without delegates leeching off of the ecosystem. I plan to release a dPoS-based coin to prove this point (that is, if no one beats me to it). dPoS is the ideal consensus algorithm for a deflationary asset, which is what Bitshares was initially championed as. Unfortunately, Bitshares has wavered many times from that path with much dilution through mergers etc.

[Bhuz]

Since the last Committee Proposal went through (fee-adjustement), @xeroc remembered and asked the Committee tu discuss about the witness pay, and more precisely what is a rational payment for a witness job
The following is the reply I wrote on the telegram committee group and, as suggested by @abit soon after, I am now posting it here too, for a wider discussion with the community

******
These are just inputs... there surely are more questions that may be worth asking and answering... please comment and discuss them, modify or add your own

If instead you want to decide what a witness pay should be without consider what his job is really about and what it is worth for the blockchain etc, then I am not interested in the discussion... I will vote for whatever the committee or shareholders decide, but probably without giving any meaningful input or personal thought


Questions we should answer:
1- Should we consider how much the witness "ideal" guy is worth on the real world? (ie. how much a (even Junior) SysAdmin gets paid)
2- Should we consider the 24/7/365 requirement aspect of the witness job? (you can sleep, have week-ends and holidays, but you still have to "go to the office" in a timely manner no matter what the day and time is)
3- Should we consider how much other blockchains spend (invest?) for their own security? (ie. how much BTC, ETH, etc spend, compared to market-cap) (even other dpos like STEEM and soon PEERPLAYS)
4- Should we consider the current profitability of the DEX DAC?
   4a- If so, should it be able to greatly increase what the points 1,2,3 would suggest?
   4b- If so, should it be able to greatly decrease (or even reset) what the points 1,2,3 would suggest? (ie, no profit for the DEX == no profit for the witnesses --> witness pay = costs of running a witness)

Maybe more tough (and subjective?) questions:
5- How much is the whole blockchain task, security and decentralization worth? (this is what a witness really do, from a pure blockchain pov) (looking at point 3 may help)
6- Should 4, and particularly 4b, be able to overcome 5?

Plus:
IMHO we shouldn't start from a number to later check if that is sustainable for a witness, but rather start from what an adequate pay for the job is, and only later set the pay...
So, witness pay = job pay + costs (adequate/deserved witness net income + costs that the job itself requires)
******


PS:
BTW, it's basically one year that I pay my BTS witness costs thanks to STEEM... if it was for BTS itself I wouldn't have earned nothing, or worst I would be at loss...
I hope this will be read and understood by all those people that were jumping and screaming around as soon as bts (finally) rose in value, to point the finger to the witnesses that, suddenly, were "earning too much" thanks to their "unbelievable, unaccettable and absurd pay"
I also hope this will be considered for choosing an adequate BTS witness pay


_ _ _ _ _ _ _ _ _ _ _ _ _ _
Some data about point 3 at the time of this post, from coinmarketcap, considering current rewards
BTC:
Market-cap: $25,617,582,661
Price: $1570.05
Monthly blockchain expenses: 12.5(btc/block) * 144(blocks/day) * 30(days/month) * 1570(usd/btc) = $84,780,000
BTC expenses / BTC market-cap --> 84,780,000 / 25,617,000,000 = 0.0033

STEEM:
Market-cap: $132,000.000
Price: $0.562
Monthly blockchain expenses: 25 * 0.189 / 63 * 86400 * 30 * 0.562 = $109,252.80 (edited to consider the x5 runner-up cost)
STEEM expenses / STEEM market-cap --> 91,800 / 132,000,000 = 0.00083

BTS:
Market-cap: $95,908,469
Price: $0.036894
Monthly blockchain expenses: 3600 (bts/hour) * 24(hours/day) * 30(days/month) * 0.037(usd/bts) = $95,900
BTS expenses / BTS market-cap --> 95,900 / 95,900,000 = 0.001

This means, compared to their market-cap:
BTS current expenses is less than 1/3 of BTC's expenses
STEEM current expenses is about 15% lower than BTS's current expenses
_ _ _ _ _ _ _ _ _ _ _ _ _ _


If interested in the matter, please comment, discuss, and share your thoughts. Thanks