Author Topic: Proxy: xeroc  (Read 78139 times)

0 Members and 1 Guest are viewing this topic.

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
Dear Xeroc, could you please clarify you point here? What is the reason for you to not unvote those witnesses who feed updated prices to bitUSD?
1. You believe applying BSIP42 is safe for bitUSD.
2. You have concerns regarding BSIP42 applied to bitUSD, but you don't want or afraid to make a decision and take actions.
3. Any other?

Thank you very much in advance!

We currently are in a very interesting (and also competitive) time when it comes to "stable" crypto assets.
With that said, we do *need* to improve, or we will not be able to compete for much longer.

The market sees a need for a tighter peg, and the current "solution" provides that quite well.
It is my believe that the way BSIP42 is currently implemented is suboptimal in many ways, but it
achieves its goals. In addition, there is a backend limitation that prevents the witnesses from using
other means for utilizing BSIP42 fully.

In short, the way most witnesses currently deal with BSIP42 should be a short-term solution that is
ideally replaced with a method that is currently highly discussed but only possible *AFTER* fixing
the backend.

With those things in mind, I believe there is more value in keeping BSIP42 for now than to ditch it away
without replacement.

Offline yury

  • Jr. Member
  • **
  • Posts: 42
    • View Profile
    • openledger.info
Dear Xeroc, could you please clarify you point here? What is the reason for you to not unvote those witnesses who feed updated prices to bitUSD?
1. You believe applying BSIP42 is safe for bitUSD.
2. You have concerns regarding BSIP42 applied to bitUSD, but you don't want or afraid to make a decision and take actions.
3. Any other?

Thank you very much in advance!
Yury Cherniawsky
OpenLedger

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
I hereby remove my "threat" of unvoting witnesses if they apply BSIP42 to bitUSD.
Ultimately, threatening them was a mistake to begin with. Also, BSIP42 has been approved explicitly to grant freedom of choice to the witnesses. IMHO proxies shouldn't push a their decision about BSIP42 onto the witnesses.

With that said, sorry for my previous statement and sorry for the mess and uncertainty it may have caused.
Please also note the discussion in this thread: https://bitsharestalk.org/index.php?topic=27203.msg322845#msg322845

Offline bitcrab

  • Committee member
  • Hero Member
  • *
  • Posts: 1928
    • View Profile
  • BitShares: bitcrab
  • GitHub: bitcrab
To further improve bitassets, I would recommend to lose then limits of the "short protection ratio" in case a call position goes
below a collateral ratio of .. say 130% to force the margin call to buy into the entire market not just up to 10%. This would be an
additional penalty to those that don't even bother keeping their collateral ratio sane.

"increasing penalty to debt position owners" is not a key to solve the problem, it always make things even worse.

when market become serious bear, it's not easy for most of the players to add more money to the market. and actually the borrower in BTS is not only the simple borrowers in traditional financial markets, they play the important role of issuing currency! when you increase the cost to issue currency, the result is less users to issue less currency, and lead to more serious currency deflation problem.

My personal stand: I am AGAINST BSIP42 on BitUSD clearly puts BitUSD holders at an disadvantage and effectively wipes out their ability to convert their BitUSD to BTS and a true face value. BitUSD creators have their call price always ABOVE the current market price (black swan if they don't sell?)

bitUSD is designed to peg USD, not to provide premium, as a stable coin, big premium/discount are hindering potential adopters to come in.

force settlement is seldom used for long time, simple to understand, before BSIP42, feed price is already >3% higher than DEX price, considering the 1% force settlement offset, you always to pay more than 4% to convert bitUSD to BTS by forse settlement comparing to just buy in DEX, if you place a buy order with price 2% higher than latest price, your orders will be fulfilled rapidly, right?
Email:bitcrab@qq.com

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


My thinking is quite easy, I (as a proxy) gave an approval to BSIP42 in the premise that it would be limited to bitCNY for now (in fact I wanted it to named explicitly in the BSIP - which unfortunately didn't happen).
The reason is that I can only acknowledge the fact that the Chinese community apparently wants to experiment with bitCNY which is "their native token" - they know their own markets much better than I do
and I have mostly seen support for BSIP42 from the chinese community so far.

bitUSD is a different beast though. There are many opponents (me included) about running BSIP42 on bitUSD. Not only because there is no real on-off-ramp for bitUSD into fiat USD but also because
it is a less liquid market. I am not convinced it would work well there.

Also, BSIP42, to me is a short term "patch" at best. It forces a price onto the markets instead of providing incentives to the market. Knowing how nuBits failed with such an approach
I would argue that we should do better.

That said, i still support the BSIP42 worker because it want the Chinese community to continue their experiment on bitCNY and chose to instead withdraw my vote
from witnesses that push BSIP42 onto bitUSD.

To avoid playing double standards, will you stop voting for witnesses who refuse to appropriately apply BSIP42 on bitCNY? (note: do it "appropriately" may be a bit hard for some people)
I don't mind witnesses running BSIP42 on bitCNY, its up to them. But for bitUSD, more convincing and more factual information is needed.

Offline clockwork

  • Committee member
  • Sr. Member
  • *
  • Posts: 376
    • View Profile
  • BitShares: clockwork
Witnesses (like employees) are expected to do EVERYTHING right.

<fairy tale>
Alice and Bob are running a window cleaning business, and Charlie is their employee. Alice insists that windows must be cleaned with MagicClean(tm), while Bob insists that windows must be cleaned with WindowMagic(tm).

If Charlie uses MagicClean, Bob will fire him. If Charlie uses WindowMagic, Alice will fire him. What should he do?

Alice and Bob's customers don't care how their windows are cleaned, they're just interested in having clean windows. What should they do?

So the customers (and Charlie as well) go to Dave, who is also running a window cleaning business. Charlie, Dave and the customers live happily ever after, while Alice and Bob go broke.
</fairy tale>

If Alice and Bob were clever they'd sort out their differences among themselves, then tell Charlie about their decision. And NO, the BSIP-42 vote is not a decision, because obviously everyone is reading something else into it.

Stop brute-forcing decisions and find consensus!

nice methaphor...Only thing missing is expecting Charlie to know enough advanced chemistry to decide which of the 2 products is better for the environment.

Offline pc

  • Hero Member
  • *****
  • Posts: 1530
    • View Profile
    • Bitcoin - Perspektive oder Risiko?
  • BitShares: cyrano
Witnesses (like employees) are expected to do EVERYTHING right.

<fairy tale>
Alice and Bob are running a window cleaning business, and Charlie is their employee. Alice insists that windows must be cleaned with MagicClean(tm), while Bob insists that windows must be cleaned with WindowMagic(tm).

If Charlie uses MagicClean, Bob will fire him. If Charlie uses WindowMagic, Alice will fire him. What should he do?

Alice and Bob's customers don't care how their windows are cleaned, they're just interested in having clean windows. What should they do?

So the customers (and Charlie as well) go to Dave, who is also running a window cleaning business. Charlie, Dave and the customers live happily ever after, while Alice and Bob go broke.
</fairy tale>

If Alice and Bob were clever they'd sort out their differences among themselves, then tell Charlie about their decision. And NO, the BSIP-42 vote is not a decision, because obviously everyone is reading something else into it.

Stop brute-forcing decisions and find consensus!
Bitcoin - Perspektive oder Risiko? ISBN 978-3-8442-6568-2 http://bitcoin.quisquis.de

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 4664
    • View Profile
    • Abit's Hive Blog
  • BitShares: abit
  • GitHub: abitmore
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


I started to feed BSIP42 on BitUSD mostly because majority of witnesses started to do the same. Few hours ago in the witness chat Bitcrab told everyone that they will start removing votes for witnesses that do not support BSIP42 starting with spring-team, but in reality their votes do not much the BSIP42 BitUSD witnesses.

Guys, your fighting is endangering the stability of the blockchain! Keep in mind that the witness job is not only about the price feed.

Please acknowledge the fact that there are disagreements about BSIP-42 and do not vote out witnesses who have been providing reliable services just because they disagree with you.

Witnesses (like employees) are expected to do EVERYTHING right.

Voters are expected to make rational decisions. However, no matter what decisions are made, voters have the final say. We're not like the PoW chains.
BitShares committee member: abit
BitShares witness: in.abit

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 4664
    • View Profile
    • Abit's Hive Blog
  • BitShares: abit
  • GitHub: abitmore
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


My thinking is quite easy, I (as a proxy) gave an approval to BSIP42 in the premise that it would be limited to bitCNY for now (in fact I wanted it to named explicitly in the BSIP - which unfortunately didn't happen).
The reason is that I can only acknowledge the fact that the Chinese community apparently wants to experiment with bitCNY which is "their native token" - they know their own markets much better than I do
and I have mostly seen support for BSIP42 from the chinese community so far.

bitUSD is a different beast though. There are many opponents (me included) about running BSIP42 on bitUSD. Not only because there is no real on-off-ramp for bitUSD into fiat USD but also because
it is a less liquid market. I am not convinced it would work well there.

Also, BSIP42, to me is a short term "patch" at best. It forces a price onto the markets instead of providing incentives to the market. Knowing how nuBits failed with such an approach
I would argue that we should do better.

That said, i still support the BSIP42 worker because it want the Chinese community to continue their experiment on bitCNY and chose to instead withdraw my vote
from witnesses that push BSIP42 onto bitUSD.

To avoid playing double standards, will you stop voting for witnesses who refuse to appropriately apply BSIP42 on bitCNY? (note: do it "appropriately" may be a bit hard for some people)
BitShares committee member: abit
BitShares witness: in.abit

Offline Thom

At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


I started to feed BSIP42 on BitUSD mostly because majority of witnesses started to do the same. Few hours ago in the witness chat Bitcrab told everyone that they will start removing votes for witnesses that do not support BSIP42 starting with spring-team, but in reality their votes do not much the BSIP42 BitUSD witnesses.

Guys, your fighting is endangering the stability of the blockchain! Keep in mind that the witness job is not only about the price feed.

Please acknowledge the fact that there are disagreements about BSIP-42 and do not vote out witnesses who have been providing reliable services just because they disagree with you.

Too late, the trend of voting out witnesses started already with BSIP42 while experimenting on BitCNY. Bitcrab and Abit both withdrew their votes from verbaltech2 witness dropping the rank from 3rd down to 19th, simply for asking questions and disagreeing with the way BSIP42 experimentation was being conducted.

I concur, that the health of BitShares will be best with a spirit of cooperation rather than competition & fighting. I do agree with your sentiment PC, but we also shouldn't over react and let voting become constrained either.

I see the biggest issue with trying to move too fast and with too narrow of a view. I would like to see a safer, more conservative, engineering with published analysis.

And thank you PC for reminding everyone there is FAR MORE to the role of witness than price feeds. There seem to be many that disagree, and it's an all or nothing proposition for them to base their votes on only 1 aspect of the witness role, and only on 1 market rather than thinking of the ecosystem as tool for the world.
Injustice anywhere is a threat to justice everywhere - MLK |  Verbaltech2 Witness Reports: https://bitsharestalk.org/index.php/topic,23902.0.html

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


I started to feed BSIP42 on BitUSD mostly because majority of witnesses started to do the same. Few hours ago in the witness chat Bitcrab told everyone that they will start removing votes for witnesses that do not support BSIP42 starting with spring-team, but in reality their votes do not much the BSIP42 BitUSD witnesses.

Guys, your fighting is endangering the stability of the blockchain! Keep in mind that the witness job is not only about the price feed.

Please acknowledge the fact that there are disagreements about BSIP-42 and do not vote out witnesses who have been providing reliable services just because they disagree with you.

Agreed.

Offline pc

  • Hero Member
  • *****
  • Posts: 1530
    • View Profile
    • Bitcoin - Perspektive oder Risiko?
  • BitShares: cyrano
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


I started to feed BSIP42 on BitUSD mostly because majority of witnesses started to do the same. Few hours ago in the witness chat Bitcrab told everyone that they will start removing votes for witnesses that do not support BSIP42 starting with spring-team, but in reality their votes do not much the BSIP42 BitUSD witnesses.

Guys, your fighting is endangering the stability of the blockchain! Keep in mind that the witness job is not only about the price feed.

Please acknowledge the fact that there are disagreements about BSIP-42 and do not vote out witnesses who have been providing reliable services just because they disagree with you.
Bitcoin - Perspektive oder Risiko? ISBN 978-3-8442-6568-2 http://bitcoin.quisquis.de

Offline abit

  • Committee member
  • Hero Member
  • *
  • Posts: 4664
    • View Profile
    • Abit's Hive Blog
  • BitShares: abit
  • GitHub: abitmore
At this point, I decided to withdraw my support from witnesses to that feed BSIP42 to bitUSD. This will happen later this week so there is sufficient time for witnesses to
evaluate their individual situation.


My thinking is quite easy, I (as a proxy) gave an approval to BSIP42 in the premise that it would be limited to bitCNY for now (in fact I wanted it to named explicitly in the BSIP - which unfortunately didn't happen).
The reason is that I can only acknowledge the fact that the Chinese community apparently wants to experiment with bitCNY which is "their native token" - they know their own markets much better than I do
and I have mostly seen support for BSIP42 from the chinese community so far.

bitUSD is a different beast though. There are many opponents (me included) about running BSIP42 on bitUSD. Not only because there is no real on-off-ramp for bitUSD into fiat USD but also because
it is a less liquid market. I am not convinced it would work well there.

Also, BSIP42, to me is a short term "patch" at best. It forces a price onto the markets instead of providing incentives to the market. Knowing how nuBits failed with such an approach
I would argue that we should do better.

NuBits has failed due to its pure botting framework which highly exposes to risks when trend changes. We're far different.

BSIP42 DOES provide incentives to the market. When bitUSD is at premium, settlement price is adjusted to higher (note: all price in my comment is X bitUSD per BTS) than trading price, borrowers are encouraged to borrow more bitUSD and sell into existence thus helps the peg, because it's less likely to be margin called. It's not setting trading price, instead, trading is free.


Quote
That said, i still support the BSIP42 worker because it want the Chinese community to continue their experiment on bitCNY and chose to instead withdraw my vote
from witnesses that push BSIP42 onto bitUSD.

How to go forward (IMHO):
While BSIP42 is a rather harsh approach that seems to work well (for bitCNY), I would try another experiment and allow witnesses to leverage the "maintainance
collateral ratio" (MCR). In my (and other people's) mind, when there is a premium, we see too little supply and reducing the MCR would allow
"cheaper" creation of bitUSD. While if there is a negative premium, there is oversupply and the MCR could be raised to reduce the supply (potentially through
margin calls).
The only drawbacks I currently see are:
- shorters need to watch their collateral more actively because their position could be called even if the price is not moving
- there is a backend issue that currently prevents this approaach from being effective/fair

The fact that MCR cannot be lower than 100.1% is _NOT_ a drawback because bitassets are supposed to always be backed
by at least 100% reserves. Removing settlement would not make sense here. Removing black swan/global settlement would
not be required (make sense) either.

When talking about "make sense" you only provided conclusions but no valid argument in regards to why it makes sense or not.

Quote

To further improve bitassets, I would recommend to lose then limits of the "short protection ratio" in case a call position goes
below a collateral ratio of .. say 130% to force the margin call to buy into the entire market not just up to 10%. This would be an
additional penalty to those that don't even bother keeping their collateral ratio sane.

As mentioned in BSIP42, applying a modifier to settlement price is effectively same as applying a modifier to MCR, MSSR and force_settlement_offset at same time.

When the modifier is greater than 10%, MSSR is effectively adjusted to less than 100% thus margin calls is avoided at all, thus reduced sell pressure (of collateral), thus reduced bitUSD premium, because selling BTS is buying bitUSD which pushes up the premium. Your proposal about increasing MSSR will lead to more consumption to market depth thus IMHO doesn't help the peg.

On the other hand, reducing MCR does effectively increase of possibility of black swan, it's no difference than increasing settlement price.

The only difference I can see is adjusting MCR and etc provides more transparency than adjusting settlement price, while effectively it's the same. The cons are also obvious: they can not be adjusted to less than 100% when needed as (IMHO) proven on bitCNY.

Quote
The MSSR is meant to protect shorts from paying to much in the
case they get margin called. This limit is supposed to be effective in illiquid markets.
bitCNY:BTS is far from illiquit and MSSR should IMHO not be the major problem.

Margin calls not protected by MSSR DID greatly suppress liquidity. Increasing MSSR in a downtrend DOES harm the peg. It IS one of the major problems.
« Last Edit: October 01, 2018, 10:27:55 am by abit »
BitShares committee member: abit
BitShares witness: in.abit

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
Provided that the UI team is looking for help with UX/UI and they now have a leason, I am in support for the Rossul worker: 1.14.124 https://www.bitshares.foundation/workers/2018-09-rossul-ui

Offline binggo

  • Hero Member
  • *****
  • Posts: 2374
  • 世间太多瘪犊子
    • View Profile
Quote
I disagree, or at least, I cannot agree. No one ever tried experimenting with the MCR, so we simply do not know
if it will work or not. I believe it is worth a try and can help reduce the premium.

The MSSR lead the premium,not the MCR,if you change MSSR to 105% or 103%, you will find a different premium.

Quote
It is only dangerous to those that are short and do not monitor their position, as it should be.
It is dangerous to the whole bitasset, not only the short, it's not a experiment.

Quote
Not sure what you mean with "flexible". If you are proposing to add another "price" just for settlement, then
I believe you haven't thought through the consequences with respect to arbitrage that could exploit two
different price feeds. If you mean something else, please elaborate.
no, the feed price didn't change , just reaction the fair price. we can design a quick force settlement fuction to instead of the margin call for to the shorter which below 175%, and make a dynamically force settlement offset(like min -2%) to the shorter which below 175%.

Quote
I respectfully disagree. The MSSR is meant to protect shorts from paying to much in the
case they get margin called. This limit is supposed to be effective in illiquid markets.
bitCNY:BTS is far from illiquit and MSSR should IMHO not be the major problem. Those
not maintaining their collateral are the problem and they risk paying a premium.
In crypto, paying a premium of 10% pay not be "sufficient" incentive to maintain the ratio
hence why I propose to remove the MSSR entirely for positions that literally *fail* to
maintain the ratio @ less than 150% (or so).

The 110% MSSR is destroy the depth of the market, and Serious squeeze the market,make the premium up to 15%.

i don't want to argue about it,  as it happened again and again in the market.


Quote
I don't see how changing MSSR affects anything ..
you can find the relation of the price of BITNCY/bts and the bitcny feed price/MSSR, and something will clearly.

Quote
We do have BSIP18 - preventing black swan is not an option unless you accept running fractional reserver, which i do not.
BSIP18 is not a good method, you can check the BITMEX Perpetual Contracts.
 https://www.bitmex.com/app/perpetualContractsGuide

If we didn't have a new method to change MSSR, the DEX price can't be as a feed price, beacause the dex price always been squeezed by the MSSR.

and the The feed BSIP42 has some effect,but it adjust dynamically too high.
« Last Edit: October 01, 2018, 10:08:57 am by binggo »