Quote from: alt on December 26, 2018, 09:31:16 am

I will support gdex-witness if they agree to give the right feed price. feed price is the base requirment, not the sufficient.
And I still have to wait those witness who give wrong feed price be voted out.

we agree, kindly support.

I have supportted, but seems the feed price from gdex-witness still 3%-4% higher than market price.

proxy bitcrab will vote these witnesses out tomorrow morning if they do not change, give them some time to response.

thank you.
I have to wait until all the wrong feed price be voted out or be fixed.
I need to make sure the majority voter have got the consensus to provide the real feed price, then I will support to reduce MSSR ASAP.

it's clear now that bsip42 is given up by the community.

the key flaw of bsip42 is that it may make the feed price much higher than market price and make the margin call orders just stay there without being eaten. this can lead to risk accumulation.

even though, we have learned that the negative feedback mechanism really works, it's possible that we update the solution, one of the solution update plan is here: https://bitsharestalk.org/index.php?topic=27522.0

in the plan list the 4th -  reduce bitCNY force settlement offset to 2% - is finished, the 1st and 2rd involve big complex and need much time to reach consensus and implement.

the 3rd - change MSSR to 5% - is drafted as bsip41 and is now in voting process, in the voting process we can see it get big support from the community, however, some whales vote to oppose it as they do not want to reduce MSSR when some witnesses feed a price which is much higher than the market price.

as bsip42 is voted out, I think witnesses now need to feed real market price, right?

https://bitsharestalk.org/index.php?topic=27672.msg326693;topicseen#msg326693 here we can see that @roelandp @xman @xn-delegate and @cyazybit now feed a price which is 5%+ higher than the market price, how about to switch to feed real market price? in order to make the whales to support bsip41?

you are wrong to request permit from these witness.
the witness didn't have the authority to make the decide.
it's the voter's fault to keep these unresponsible witness acitve.

Quote from: alt on December 21, 2018, 09:18:08 am

我说的很清楚是爆仓价，在MSSR 10个点的时候，喂价最高不能高于市场价8个点

虽然目前见证人的喂价呈现出两极分化，但喂价中间值基本上与CEX市场价持平，内外盘差价依然9%左右，要消除这个溢价，还是得试验 MSSR 调整为1.05观察溢价变化，如果溢价减小到5%左右，可以尝试MSSR调整为 1.03、1.02

现在 crazybit 喂价超过市场价30%，仍然有很多人投票给他，
在这种垃圾喂价都能被大多数 proxy 支持，没有取得放弃喂价操纵共识的情况下，我肯定不会支持调低MSSR

Quote from: bitcrab on December 21, 2018, 08:27:23 am

Quote from: Yao on December 21, 2018, 05:31:08 am

反弹期就是消化爆仓单的最好时机，而按照本帖帖主alt的建议，见证人应该让喂价比市场价低2%，这样才能让爆仓单尽可能处于“卖一”的位置，买盘可以直接吃爆仓单而不是吃其他正常的卖单。对于已经爆仓的人来说是很痛苦，但是在反弹的时候在相对高的价格被吃掉然后平仓要比创新低之后在更低的价格被吃掉要损失小吧？！除非，你们真的以为牛市来了。

如果你们认为负反馈喂价是价格操纵，那有为何不认为喂比市场价低2%的价格是价格操纵。

确实也是价格操纵，达成共识（所有见证人都不再喂负反馈，同时MSSR减少至5%或者3%、2%）之前我也不会喂比市场价低2%，没有意义。

市场没有反转之前，bitCNY的黑天鹅风险一直存在。如果把目前大量的爆仓单比作银行坏账的话，喂价比市场价低2%可以算是一种“处理坏账”的方式，但也需要社区达成共识（当然，让处于爆仓状态的抵押者接受这种处理坏账的方式，心理上有些不可接受，仓位决定想法）。达成共识之前，witness.yao 会一直喂市场真实价格。

不知道我哪句话让你们认为我要求喂价低于市场价2%的，
我说的很清楚是爆仓价，在MSSR 10个点的时候，喂价最高不能高于市场价8个点

Quote from: alt on December 20, 2018, 07:07:43 am

Quote from: xeroc on December 20, 2018, 07:01:12 am

Quote from: alt on December 20, 2018, 06:57:33 am

Quote from: clockwork on December 20, 2018, 04:51:54 am

Which version client are you guys using?

I have tried almost 10 different "stable" versions recently published,
but almost no one can be used in fact
I wonder who are the manage of the GUI development worker, too bad.

Bug reports can go here:
https://github.com/bitshares/bitshares-ui/issues/

no thanks.
I thought we should pay for a professional worker.
I wonder who can tell me which version is really usable.

you can either keep dissing the worker or help make it better, your call

If you don't provide details/bug report it can't be fixed

(fwiw, mine runs with 15 different accounts loaded and appears to work fine)

There are 2 possible reasons for your problem:

a) There was a change recently on the subscription behaviour of bitshares-core. Copying from teh release notes:

"Operators of API nodes that service client software that expect to receive updates about all data on the blockchain by default (such as BitShares Reference Wallet before 180401), as opposed to narrowly subscribed data, should ensure that their API nodes are configured with the enable-subscribe-to-all set to true due to Pull Request 1049."

So maybe you're connecting to an API node with the subscribe-to-all setting set to true

or

b) This issue recently brought up by abit: https://github.com/bitshares/bitshares-core/issues/1472

Interestingly, neither of those is a UI issue but core API issues identified and fixed (for (a)...application depends on node operator)  or in the process of being fixed (for (b))

So please don't spread false accusations.

can you tell me which stable version do you use?
when I met a problem, I try to change another version, almost every stable version has different problems after I try less than  several minutes.
I think there are no need to report bug for such unstable software, they need to test themself, it's a job with payment, not a hobit job.

for the issue brought by abit, it's also the GUI's job of course, they used a wrong design, used a wrong api.
when I tell the wallet that my account is alt, it just need to subscribe infomation about alt, I never told it to subscribe other account with the same key.
eventually, it even bring those junk account to my default list of transfer function. such stupid.

Which version client are you guys using?

I have tried almost 10 different "stable" versions recently published,
but almost no one can be used in fact
I wonder who are the manage of the GUI development worker, too bad.

as in the Poll of bsip41 - worker proposals  1.14.144 and 1.14.155, far more voting power support to implement bsip41, gdex-witness plan to feed 1.05 as MSSR for bitCNY in 2 days.

bsip41: https://github.com/bitshares/bsips/blob/master/bsip-0041.md

Sorry, I will vote for no until these wrong feed price been voted out.
I don't think witness should execute it until it become an active worker.