Initial Witness Pay & Number of Witnesses

[monsterer]

So I agree with Riverhead that Bitcoin security is very weak.

The bitcoin network spends 25 BTC every 10 minutes on security, that's $33k per day. How much does it cost to elect 17 witnesses? ...Once one person does that, it's goodbye chain.

[liondani]

I have no idea how "the market" will perceive 17 witnesses but I personally see no problem with it.

I see a big problem! I would think that the control of the chain is from a single family (the Larimer's?)
It's better we have the illusion of more witnesses even controlled from the same entities...
I think the market will perceived it very negative. It doesn't inspire any trust or a feel of decentralization (even fake/virtually). We were accused even for the 101 delegates from the crypto community, don't forget that!

[Empirical1.2]

You may laugh at the Bitcoin crowd, but they know what it means to be under attack.

And they are OK with this? https://blockchain.info/pools

Seems like 6 "witnesses" sign more than 50% of BTC blocks.

If the pool goes down miners can still go to other pool to secure the network .

If 17 witness is down , I don't think the network is capable of electing new witness to keep the network running . (unless there is something magical I don't know about . )

In the case, where a pool Ghash.io gained close to 50%,  they (Ghash.io) actually had to stop accepting new miners as the miners continued to go to the most profitable pool even though it seriously threatened Bitcoin security.

So I agree with Riverhead that Bitcoin security is very weak.

However the rest of spectral's answer is spot on.

Edit: I agree with you that if active witnesses go down it's a problem. So I also think besides having more witnesses we should also pay standby witnesses a smaller amount so BTS has witnesses waiting and ready to go if needed and so there is a bit of competition/incentive there. (I imagine the active witnesses will be quite 'sticky' if they do a good job, meaning stand-by's will lose interest, thereby compromising security in case new witnesses are needed.)

Edit 2: Where Bytemaster makes a good point is that there is a limit on the amount of  witnesses, voters can vet.

[btswildpig]

You may laugh at the Bitcoin crowd, but they know what it means to be under attack.

And they are OK with this? https://blockchain.info/pools

Seems like 6 "witnesses" sign more than 50% of BTC blocks.

If the pool goes down miners can still go to other pool to secure the network .

If 17 witness is down , I don't think the network is capable of electing new witness to keep the network running . (unless there is something magical I don't know about . )

[Empirical1.2]

I'm somewhat puzzled by this discussion...

How important is securing the network which is the witness role? How big of a hit would BitShares take if security was compromised? Isn't it better to aim too high than too low?

Bribery, persecution of individuals, DDOS attacks? Do you guys really believe that 17 witnesses is enough decentralization for fighting big government? You may laugh at the Bitcoin crowd, but they know what it means to be under attack. BitShares is still too small to have felt the spotlight, and we're the ones going for the throat of the banks.

Is a potential 100 million dollar financial network really going to run on 17 VPSes costing $40 each? Are we really going to limit the cost of the security for this network to $8160 a year? That sounds kind of absurd to me.

Bitcoin is currently spending more than 200 million dollars on security. Considering the market cap, that is of course not a fair comparison, but even if BitShares brought that number down to 100K (~0.67% of current market cap) it would still be an incredible improvement over Bitcoin.

When I envision the BitShares 2.0 network 2 years into the future, I see 300 high-power witnesses running the network from all corners of the world. The difficulty and cost of attacking the BitShares Network is absolutely mind-boggling, as it should be.

Also, denominating witness pay in BTS is more useful than USD. A higher BTS price means more adoption, higher volume, which in turn places higher requirements on the witnesses. Also, witnesses will participate who actually believe in BTS, not just to make a quick buck.

To sum it up, 17 is way too low. It is natural that the number of witnesses will increase as we grow, not decrease. Start out at 67 witnesses with the option of increasing the number with no upper limit. Keep in mind, letting people run a witness also educates people about BitShares (like me). If you later want to increase from 17 to 35, how long will it take to get 18 qualified admins to jump in?

My 2 BTS. Help me understand why I'm wrong.

(Excluding the notion that Bitcoin security is satisfactory)    https://www.youtube.com/watch?v=6f-2LExTvBw

[Riverhead]

You may laugh at the Bitcoin crowd, but they know what it means to be under attack.

And they are OK with this? https://blockchain.info/pools

Seems like 6 "witnesses" sign more than 50% of BTC blocks.

[Spectral]

I'm somewhat puzzled by this discussion...

How important is securing the network which is the witness role? How big of a hit would BitShares take if security was compromised? Isn't it better to aim too high than too low?

Bribery, persecution of individuals, DDOS attacks? Do you guys really believe that 17 witnesses is enough decentralization for fighting big government? You may laugh at the Bitcoin crowd, but they know what it means to be under attack. BitShares is still too small to have felt the spotlight, and we're the ones going for the throat of the banks.

Is a potential 100 million dollar financial network really going to run on 17 VPSes costing $40 each? Are we really going to limit the cost of the security for this network to $8160 a year? That sounds kind of absurd to me.

Bitcoin is currently spending more than 200 million dollars on security. Considering the market cap, that is of course not a fair comparison, but even if BitShares brought that number down to 100K (~0.67% of current market cap) it would still be an incredible improvement over Bitcoin.

When I envision the BitShares 2.0 network 2 years into the future, I see 300 high-power witnesses running the network from all corners of the world. The difficulty and cost of attacking the BitShares Network is absolutely mind-boggling, as it should be.

Also, denominating witness pay in BTS is more useful than USD. A higher BTS price means more adoption, higher volume, which in turn places higher requirements on the witnesses. Also, witnesses will participate who actually believe in BTS, not just to make a quick buck.

To sum it up, 17 is way too low. It is natural that the number of witnesses will increase as we grow, not decrease. Start out at 67 witnesses with the option of increasing the number with no upper limit. Keep in mind, letting people run a witness also educates people about BitShares (like me). If you later want to increase from 17 to 35, how long will it take to get 18 qualified admins to jump in?

My 2 BTS. Help me understand why I'm wrong.

[monsterer]

Would any of these Else conditions help? https://bitsharestalk.org/index.php/topic,18549.msg238114.html#msg238114

Unfortunately, not. It is trivial for any one single entity to control VPSs anywhere in the world.

[kenCode]

The witnesses should be all in individual countries to be safe. With rising marketcap, the number can be increased.

How do you propose to prove their location?

Would any of these Else conditions help? https://bitsharestalk.org/index.php/topic,18549.msg238114.html#msg238114

[Riverhead]

@bytemaster the biggest issue with having only 17 witnesses are the price feeds for bitassets .. the median of 17 can be manipulated easier that for 31 ...
would it be possible to have yet another role on-chain? Those that publish feeds for bitassets? Not sure how to pay the though ..

Interesting - I hadn't considered price feeds. Perhaps workers could be used?

[xeroc]

@bytemaster the biggest issue with having only 17 witnesses are the price feeds for bitassets .. the median of 17 can be manipulated easier that for 31 ...
would it be possible to have yet another role on-chain? Those that publish feeds for bitassets? Not sure how to pay the though ..

[twitter]

dummy-proof package may work in current early stage but no a long term solution and BM /bts2  would like blockchain to pay the cost not any live company

[btswildpig]

First I think the technical cost for maintaining a server could be reduced by providing a dummy-proof package .

Namely , even a dummy can install , reinstall , resync blockchain ...etc with simple command on a fresh VPS without even asking anyone how to do it .

And secondly , since BitShares is important to Cryptononmex as a show-and-tell project  , could this technical cost paid with shares of  cryptonomax  ? Guess some people would do the job for that . Then the cost to the network could be only 40 USD  . and you can support more number of witness .

[JWF]

I'm thinking that it will be best to have a minimum of 1/4 and max of 1/3 of the original 101, so 25 to 30 witnesses. At this level it will allow for movement in both directions. If we start out at too low of a point, then we have no way to lower the numbers if needed.

Perception is going to play more into adoption once the speculators' influence on the price has diminished and the "product" is solid and in the wild.

[monsterer]

The witnesses should be all in individual countries to be safe. With rising marketcap, the number can be increased.

How do you propose to prove their location?

Reducing the witness count to 17 from 101, reduces the cost of a sybil attack majority.