Quote

The current performance goal for OBC is to achieve 100,000 transactions per second in a standard production environment of approximately 15 validating nodes running in close proximity.

Quote

The initial OBC release package will offer three consensus implementations for users to select from: 1) No-op (consensus ignored); 2) Classic PBFT; and 3) SIEVE (an enhanced version of classic PBFT).

Quote

Does OBC have native currency?
No. However, if you really need a native currency for your chain network, you can develop your own native currency with chaincode.

Quote

This software is made available under the Apache License Version 2.0.

Looks like a strong competitor to Graphene toolkit if CNX wants to push Graphene to potential private chain owners.

What I am trying to solve with the collateral sponsors proposal is - the need for the bridges to come up with 2x in BTS for collateral when a customer wants to go from BTC into bitBTC. Isn't this the main reason why no bridge - you or anybody else even thinks about providing BTC->bitBTC bridge? The proposal effectively locks as much BTS as if you sold BTS for the BTC received at the current price and nothing more. Isn't this a good deal. This is effectively equal to what happen if it was a bridge for  BTS<->BTC.

Just comes one idea. If we can provide CFD style 100% collateral backed asset, why we need 200%? In any case except a black swan event, the rest 100% is not changed thus unnecessary. With 100% collateral you can afford 50% price drop, with 200% collateral you can afford 75% price drop, both are not 100% safe.

bump. Is it better to move this post to Committee Proposal board?

Quote from: abit on February 17, 2016, 12:44:00 am

好吧，我提点意见。
个人感觉vector不行，可能根本没必要加从数据库取数据存到vector那一步。速度慢，吃内存。用BTS资产做个分红测试，截个CPU图看看？

我注意到了，
但我先预先申请足够的内存。所以vector和数组效率相当。
性能还可以，我在本地测试的
给1000个账号分红，用时少于1s
给11000个账号分红，用时间大于3s

这个感觉优化空间还很大。处理时间大于500ms时witness就不会出块了，如果每个witness都处理不过来，等于网络挂掉。只要有一个 witness够强，能处理过来，就还好，会丢块，但至少不会挂掉。

可以预先计算下，有多少个账户分红，如果超过一定数量，直接拒绝处理，需要使用优化过版本。这个数量由一个性能测试给出来，写死在代码里。但这个预先计算也要先做好性能测试和优化，别直接跑挂了。

另外可以加个规则：每个块最多跑一个分红交易，多的不受理。甚至，发现有分红交易，其他交易都不处理。或者只在“维护块”时才处理分红交易，这样处理时间就相对充裕一点。

考虑 A股股东一般在几万左右，
初期的BTS的资产 股东一般在 几百左右，性能够用了

将来如果一个资产真在几万甚至几十万个股东，有两个方便可以改进
1.witness当让需要更好的电脑，
2.执行操作时候采取异步， 先adjust 发分红者的余额，就可以打包交易广播了，然后在慢慢调整股东余额。这样不会影响打包速度。

异步操作的思路可以，但是需要在处理下个块，或者本块下一个操作前完成，否则，如果下个操作需要引用分红操作的结果数据，就傻了，会导致卡死或者数据不一致（分叉）。这个可能涉及整个架构的调整，方向没问题，但预计工程量不小。

以上供参考。

Quote from: abit on February 17, 2016, 02:07:08 am

Quote from: fuzzy on February 16, 2016, 11:52:06 pm

Quote from: abit on February 16, 2016, 04:09:56 pm

@fuzzy can you make the topic longer when posting a new thread next time? I always ignore the posts if nothing can be get by reading the topic alone. Time is money, my friend.

//Update: the idea in OP is interesting.

what precisely are you meaning?  i intentionally left it a little focused and concise but will gladly explain further if you want.

Some people would think: when I see the title of a new post, if I don't know what it is talking about, I click in and see what it is.
Some other people would think: when I see the title of a new post, if I don't know what it is talking about, I just ignore it since it will likely waste my time.

I think I'm the latter. I don't like the title of this thread. Makes sense?

thanks for the feedback.

Thanks for update.

Quote from: tonyk on February 16, 2016, 07:28:11 pm

Important points:
In case of liquidation the collateral provided by the bridge is used to cover the losses (if any) first [before the collateral from the 'collateral sponsor'] .
The loss for the bridge is fictional BTW. That loss is exactly offset by the same gain of holding BTC received by the customer requesting the bridge from BTC to bitBTC.

I don't think the loss is fictional - If a customer sells us 1 bitBTC, and then the price of BTS drops, we start to incur liquidation risk as the value of the collateral drops, whereas the customer does not, since we sent him 1 genuine BTC.

In this case you earn some profit indeed, since the 1bitBTC you bought can buy more shares now if you haven't bought earlier. It's the one who provided this 1 bitBTC bear the risk, if it's you, you can easily close the position without loss.

No responses whatsoever?

Tony, I replied earlier. Did you see it?

Quote from: bytemaster on February 16, 2016, 08:56:11 pm

https://github.com/bitshares/bitshares-2/releases

Windows binaries will be posted shortly.

You should use here:

https://bitsharestalk.org/index.php/topic,7067.0.html

That thread has been abandoned for a long time.
Check this: https://bitsharestalk.org/index.php/topic,19040.0.html

Quote from: abit on February 16, 2016, 11:52:53 pm

Quote from: lil_jay890 on February 16, 2016, 04:57:32 pm

Quote from: noisy on February 16, 2016, 04:37:33 pm

prices of the cryptocoins by nature are very volatile. This is the reason why we should use stable bitAssets as payment for workers: https://bitsharestalk.org/index.php/topic,21449.msg278949.html#msg278949

Every worker should get exactly this amount of money which he need to provide a good service. For USA citizen this will be X bitUSD, for Chinese citizen this will be Y bitCNY.

The reserve fund needs to hold an assortment of bitassets.  It reduces the reserve fund risk and allows developers to be paid in their preferred currency.  It also promotes the entire BTS ecosystem, instead of just the bts token.

The reserve fund should act like sovereign wealth fund.  Constantly rebalancing it's holdings.  For starters, the reserve fund should hold 70% bts and split the other 30% between bitUSD, bitCNY, bitEUR bitBTC and bitGLD.  This would promote 5 bts currencies and improve liquidity while attracting dev's who may not want to get paid in a constantly fluctuating asset.  Fluctuating assets are nightmares to report for taxes.

You mean dump 30% of reserve fund to the market? It's about 300M. I don't even believe there are so many bit assets exist.

I never said dump... If a sovereign wealth fund wants to change allocations, they don't go and buy or sell the entire position over the next trading period.  It's is done over a period of months in order to make the transition smooth and not generate weird price distortions.

Right now the reserve fund is in an extremely risky scenario. Remember, apple doesn't hold its reserves in apple stocks. Governments don't hold all their reserves in their national bonds.  It's prudent to diversify, not to mention promotion of the bts ecosystem.

If you don't create bit assets and sell into market, but buy bit assets from market, you're consuming but not providing liquidity.