1. do you think this is a flaw? or it's a right logic?

I think it is the right logic. When a black swan happens, the peg is broken, and holders of the bitasset can settle their holdings at the price of the least collateralized short. This is the defined behaviour of our smart coins, see for example https://bitshares.org/technology/price-stable-cryptocurrencies/ :

If this revaluation happens faster than the short positions can be forced to cover, then all SmartCoins are liquidated at the exchange rate of the least collateralized short position. This is similar to an insolvent bank converting its deposits to equity.

The fact that the backing collateral increased significantly in value since the black swan is irrelevant. With the black swan, the collateral put into the settlement fund effectively belongs to the holders of the smartcoins.

2. did you descipe all the rules changed because of BSIP18 really clearly? did you descipe this rule?where have you discuss these?

Yes, see BSIP-18 (emphasis added):

When a market-pegged asset undergoes a global settlement, [,,,] the outstanding debt (the BitAsset long positions) are still collateralized through the settlement pool at the fixed settlement price. It is even possible that the value of the collateral exceeds the nominal value of the MPA significantly.

Bugfix: MPAs that have seen a global settlement cannot be settled after the price feed expires

It has turned out that force-settling an MPA requires a valid price feed even when the MPA has a settlement_price set. This is clearly a bug, since in that case the settlement price is independent from the price feed. Furthermore, publishing price feeds is no longer possible after a global settlement, so the time when settlement is possible at all is limited to the expiration period of the price feed of the MPA.

This bug will be fixed. See https://github.com/cryptonomex/graphene/issues/664#issuecomment-254056746 for a discussion.

BSIP-18 was discussed here: https://bitsharestalk.org/index.php/topic,24322.0.html .

3. did the share holders really know these rule? and who reviewed these rules? who make the decition to change rule like this?

The shareholders have approved of these rules in the usual process, i. e. through a worker proposal. I can only assume that everyone involved did their due diligence.

4. who will take responsibility for the fault?
5. how do you avoid next mess up?

There is neither fault nor mess up.

SOLCERT HOLDERS (Special Early Contributors)
During the months of March to May 2016 we had our earliest contributors to the project donate BTC, ETH and BTS (Bitshares) to get the project started at the application layer. We are very grateful to these individuals who allowed us to make huge progress in a big way. A token was created on the Bitshares blockchain (SOLCERT) which was given to donators AFTER the Pre-Sale had ended. This token is suppose to give access to the core token of the official chain at a set rate per solcert.

This is still being thought through and a conversation will be had with our legal advisors on how best to compensate while avoiding regulatory landmines. As such for each solcert an individual holds we expect to exchange that token for the core token Sollar Bond at a specific rate. Again that rate has not been chosen as of yet but will most likely be 3-5 Sollar Bonds per SoLCert. Right now about 8 Million SoLCerts exist in the market. Four Million were sold. The rest of the 4 Million supply of SoLCerts will be burned.

SOLCERT TRADING (Ending)
SoLCert holder have until the end of Saturday December 30th 2017 at 12pm CST to sell and or buy the token. After this date and time the markets will be locked and trading will no longer be permitted. At this date whoever contains the supply in their hands will be the ones to receive the exchange rate when the blockchain launches. This is a fair warning. The following message will go out to all SoLCert holders during the next week:

Attention All SoLCert Holders. All SoLcert trading and markets will be frozen December 30th 2017 at 12pm CST. This is in preparation for future funding rounds for 2018. Please see this link for more information and updates: (link)

Erm... I "donated" several BTC IIRC, but I don't have any SOLCERT tokens in my wallet. How do I claim them?

How did it happen guys that you didn't work out the fork such that it does not piss off anybody? Alt was a very good witness and he shut down btsbots together with his witness node, which is a lot of pity...

We came up with a detailed description of the proposed changes 6 months ago. The BSIP-18 proposal was discussed here in the forum, all questions that came up were answered, all concerns addressed (AFAIK). Then a worker was created and was quickly voted in, meaning the proposal has been approved by a majority of the shareholders. The proposal was implemented as planned, and was extensively tested in testnet. All witnesses were supposed to participate in testing, were repeatedly asked to do so, and had lots of time to do it. Then we created a release and announced the hardfork, 5 weeks before the scheduled time. We even postponed the release for another 10 days after discovering a bug.

AFAICS we did everything exactly right.

Then, a mere 5 days ago, this message appeared:

btsbots.com will not support the fork because of the unreasonable trade logic.

No explanation what was meant with "unreasonable trade logic". Apparently there were discussions in various other channels. I have heard about claims that the release contained "malicious code injection", and that your "funds are not safe on BitShares". Again, without any explanations what exactly the problem might be.

I have been told that several chinese-speaking community members talked to @alt - it seems that the "unreasonable trade logic" is that holders of SILVER can settle their bitassets against the settlement_fund. I don't see how that is unreasonable. Quite the opposite, I think this is the way how things are intended to work. Nobody else who I talked to really understands alt's point of view, or supports his position.

I'd like to point out that at "T - 5 days" you can't simply stop such a hardfork. Almost everybody who runs a full node will most likely already have upgraded at that point. We've announced the release 5 weeks before the planned date, precisely because it takes time to get the information out to an unknown number of people.

How did it happen guys that you didn't work out the fork such that it does not piss off anybody?

Please tell me what we could have done better. Maybe it will help next time.

Your backup node should be kept running in hot standby, so it will keep itself in sync with the rest of the network.
You must configure it to use a different block production key though, to make sure that not both nodes are producing blocks at the same time.

Your server is more than sufficient for running a witness node at this time, and for some time to come.

In order to get voted in, you should provide additional infrastructure, like seed node(s) and/or API node(s). You also need a backup server, preferrably in a different location, that can take over when your main server goes down.

For setting things up and proving yourself it would be advisable to run a testnet witness as well.

* The BBF will redeem these funds on a regular basis and buy up bitUSD from the market (with reasonable premiums).
* For this reason and due to volatility of BTS, the available budget in USD terms might vary over time.

In my understanding, the idea of bitUSD-denominated workers was meant to protect both sides from BTS volatility. In this case this doesn't apply, particularly because the available budget as well as the spent budget will vary anyway.

Therefore, IMO, buying bitUSD is unnecessary in this case, and will only serve to subsidize bitUSD shorters, which I generally oppose.

Apart from this I think a budget worker would be a good idea. Our proposal process is too rigid and too complicated in many cases.

Whenever I read promises like these

Someone tells you that there is a way to increase the value of your coin and simultaneously accumulate more of it. An easy way. You can just stake your coins and get a passive income. Like you can have your own masternode with as little as $1. You might earn 0.01% per day, or 10%. Depends on the day. And you don't even take the value from someone else.

a klaxon goes off  in my head and a big red sign reading "SCAMMER" lights up.

You cannot write a bot that trades with a guaranteed profit.

I think magicwallet can be one reliable source for BTS/CNY price, here an CNY/bitCNY price can be calculated out based on the P2P trading data, combined with the bitCNY/BTS price in DEX, a reliable price can be get.

You shouldn't do that. The CNY/bitCNY price and the bitCNY/BTS price are both influenced by the feed price (and settlement offset). Using these prices to define the feed price creates a feedback loop, which is inherently unstable.