Quote
What makes a proof of work algo better in this context than a proof of stake algo? POW has more friction / is less efficient / more costly overall...
There are been some interesting innovations in the proof of stake space with the introduction of Nxt and also BitShares adopting PoS. We feel that mining currently best meets the design decisions we've made for Ethereum, but this does need to be explained in depth and with context. There will be a lag between the fundraiser and the launch for mining thus in the gap, we are going to release an article explaining why we have chosen a mining versus a PoS security model. Should data from Nxt or Bitshares (if released within this timeframe) provide a more convincing counter-argument, we'll have the opportunity to switch.
Overall, designing a new cryptocurrency is a game of tradeoffs, economic philosophy and user experience. Ethereum is being developed with the notion of balance in all things from the method of distribution to the security model.
Quote
From when to when will the IPO be?
Current plans are for a 60 day fundraiser, starting from Miami on; however, we are still exploring this and thus will set something in stone closer to the conference.
Quote
Does that mean there will be 2^100 =
1,000,000,000,000,000,000,000,000,000,000 units of Ether!?
Can you comment on this am I reading that wrong, or is half of that decimals?
1,000,000,000,000,000,000,000,000,000,000 units of Ether!?
Can you comment on this am I reading that wrong, or is half of that decimals?
1 bitcoin = 100000000 satoshi
1 ether = 1000000000000000000 Wei
The unit of account of the system will be ether like bitcoin is the unit of account for the bitcoin ecosystem.
Quote
Also, is there an ETA/roadmap for the public fundraising (.0001 BTC = 1 ether)?
Come visit us in Miami. We should have everything ready by then.
Quote
Let's say there will be 2 million USD (roughly 200 000 BTC) invested in the IPO. That would be 2^9 ether. Then there are 2^91 (respectively 2^119) to be minded. That is a lot of inflation!
Is there anyway to figure out whether the price will drop after the IPO or not because a lot will be mined at the beginning... It is about a 60% inflation rate per year acc. to your white paper. Right?
Is there anyway to figure out whether the price will drop after the IPO or not because a lot will be mined at the beginning... It is about a 60% inflation rate per year acc. to your white paper. Right?
No, the rate of inflation is always decreasing and comparable with bitcoin. The Total supply formula is as follows TS(t) = 1.5x + 0.5x*t with x being the initial supply and t being time in years. At genesis block TS(0) = 1.5x. Year 1 TS = 2.0x. You'll notice rate of inflation slows over time and we even made this graph to compare it to bitcoin:
The core idea is that the rate of inflation tends towards year and that supply is growing fastest while demand is scaling non-linearly. In terms of ROI, this should be reflected with a positive ROI.
Quote
do you mean, if we apply the above example, that 2^9 * 0,25 goes to the founders and so on...
No, the founder shares will be given only once like founder stock in a startup and the percentage founder ownership of the total amount of ether in supply will be 6.25% in year 5, which is a little over half of what Satoshi has of bitcoin. We really wanted to model supply like a startup to big company. The benchmark I used was Bill Gate's holdings of Microsoft going from 64% initially (which is considerably larger than our holdings) to roughly 5 percent or so today. Overall, both gates and Satoshi percentagewise are doing better. Finally, founder shares cannot be spent for a locked period of time that's hard coded into Ethereum. Thus, we can't sell our Ether until long after the ecosystem is bootstrapped.
Quote
Our technical team has evaluated the Ethereum proposal and design and have concluded the following:
1) Mining means the DAC will be operating at a loss or break-even at best. No dividends.
2) Scripts will require more blockchain space and bandwidth resulting in lower transaction volume for the same level of decentralization.
3) We do not believe the scripts can efficiently implement a BitShares like market matching with automatic margin calls at scale.
4) Merged mining would be required to secure parallel chains... this has its own challenges.
5) Mining will result in centralization one block at a time, something very bad for chains that implement markets.
6) Finding GPU developers is hard enough, defining a new dedicated language for this purpose will be even harder.
7) If you eliminate mining, then the cost of launching a new DAC is near 0 and you can simply use C++ to encode your contracts starting from a 'shell DAC' and launch without having to overload everyone not interested in your contract.
Cool NO 'competitor' thus far is willing to admit that multiple parallel blockchains will be required to handle the order of magnitude greater transaction volume an exchange experiences vs Bitcoin and this is for a SINGLE currency pair. Imagine attempting to have every tradable market on one chain! This will rapidly be centralized into trusted supernodes that can handle the bandwidth requirements.
9) You think bitcoin verification times have trouble scaling, imagine executing an interpreted language!
Conclusion: We believe Ethereum is an interesting computer science project with little compelling advantage in developing new DACs and many drawbacks.
We wish them well and if their scripting language and contract design proves useful as a means for very special purpose contracts then we suspect we will be able to adapt it to a more efficient, profitable, AGS honoring DAC
https://bitsharestalk.org/index.php?topic=1854.msg27154#msg27154
1) Mining means the DAC will be operating at a loss or break-even at best. No dividends.
2) Scripts will require more blockchain space and bandwidth resulting in lower transaction volume for the same level of decentralization.
3) We do not believe the scripts can efficiently implement a BitShares like market matching with automatic margin calls at scale.
4) Merged mining would be required to secure parallel chains... this has its own challenges.
5) Mining will result in centralization one block at a time, something very bad for chains that implement markets.
6) Finding GPU developers is hard enough, defining a new dedicated language for this purpose will be even harder.
7) If you eliminate mining, then the cost of launching a new DAC is near 0 and you can simply use C++ to encode your contracts starting from a 'shell DAC' and launch without having to overload everyone not interested in your contract.
Cool NO 'competitor' thus far is willing to admit that multiple parallel blockchains will be required to handle the order of magnitude greater transaction volume an exchange experiences vs Bitcoin and this is for a SINGLE currency pair. Imagine attempting to have every tradable market on one chain! This will rapidly be centralized into trusted supernodes that can handle the bandwidth requirements.
9) You think bitcoin verification times have trouble scaling, imagine executing an interpreted language!
Conclusion: We believe Ethereum is an interesting computer science project with little compelling advantage in developing new DACs and many drawbacks.
We wish them well and if their scripting language and contract design proves useful as a means for very special purpose contracts then we suspect we will be able to adapt it to a more efficient, profitable, AGS honoring DAC
https://bitsharestalk.org/index.php?topic=1854.msg27154#msg27154
I'm not going to fully address Dan's concerns here. The questions he listed indicate they either didn't read or didn't comprehend the whitepaper. For example, the entire philosophy of Ethereum is to be a base layer for innovation thus the particular economic model of a DAC running on top of Ethereum is beyond the scope of our design. A person could indeed have dividends in a sub-currency, yet this point seems to have been missed or ignored.
As for P2P exchange, we have a close relationship with Open Transactions and combined with a namecoin style contract provided in the whitepaper and bitmessage makes a significantly more efficient distributive exchange than is possible with BitShares. Trust is not required as auditing can be done on Ethereum blockchain and we wouldn't suffer any bloat.
Things like 7 again demonstrate either a lack of comprehension or ignorance of our design, contracts are more than robust enough to launch a proof of stake subcurrency. 6 seems to ignore ethereum script is turing complete and thus you can compile a language like c++ into it (anyone ever used coffeescript to write js)?
On a side note, I am honestly curious how Invictus intends on building DACs without a turing complete language? It seems like you would end in an infinite inductive process of having to build a bigger feature set for the next set of DACs. I guess they have a different philosophy and this is fine. I wish them well and hope they find success for the market's benefit.
NO 'competitor' thus far is willing to admit that multiple parallel blockchains will be required to handle the order of magnitude greater transaction volume an exchange experiences vs Bitcoin and this is for a SINGLE currency pair. Imagine attempting to have every tradable market on one chain! This will rapidly be centralized into trusted supernodes that can handle the bandwidth requirements. 
