Author Topic: [BLOG POST] - Casper Review  (Read 5255 times)

0 Members and 1 Guest are viewing this topic.

Offline lil_jay890

  • Hero Member
  • *****
  • Posts: 1197
    • View Profile
.. This constant dumping isn't going to make any newbie want to come in and buy anything.
Not sure we want "newbies" that don't understand the tech to buy into BTS .. do we?
I'd rather have investors that are well informed and understand the implications and possibilities ..
IMHO it is this way that will make an organic growth ... </opinion>

Newbies = new money.  New investors can be classified as newbies at some point.

If the hottest project in crypto is cut by 2/3 from the opening price with days, that won't inspire confidence in any investors.  Rising or stable market caps will gain outside coverage.  So far there has only been a total of 4billion dollars floating around crypto.  That number hasn't changed much over this year.

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
.. This constant dumping isn't going to make any newbie want to come in and buy anything.
Not sure we want "newbies" that don't understand the tech to buy into BTS .. do we?
I'd rather have investors that are well informed and understand the implications and possibilities ..
IMHO it is this way that will make an organic growth ... </opinion>

Offline lil_jay890

  • Hero Member
  • *****
  • Posts: 1197
    • View Profile
Wow money is just fleeing crypto... I believe a strong ethereum would be more helpful to bts.  This constant dumping isn't going to make any newbie want to come in and buy anything.

Offline BunkerChainLabs-DataSecurityNode

Wow! ethereum market cap dropped from $160M to $80M after your post! A 50% decline in one day!  :P
I hope we witness the same increase next time you post something... for bitshares   :D



that  would be awesome;)
http://bytemaster.github.io/2015/08/08/Review-of-Casper-Ethereums-proposed-Proof-of-Stake-Algorithm/

It's down almost another 50% from the 50% yesterday.

Still hasn't hit bottom.
+-+-+-+-+-+-+-+-+-+-+
www.Peerplays.com | Decentralized Gaming Built with Graphene - Now with BookiePro and Sweeps!
+-+-+-+-+-+-+-+-+-+-+

Offline twitter

  • Sr. Member
  • ****
  • Posts: 279
    • View Profile
Wow! ethereum market cap dropped from $160M to $80M after your post! A 50% decline in one day!  :P
I hope we witness the same increase next time you post something... for bitshares   :D



that  would be awesome;)
http://bytemaster.github.io/2015/08/08/Review-of-Casper-Ethereums-proposed-Proof-of-Stake-Algorithm/
witness:

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
Market cap has nothing to do with tech.  See doge.   


Sent from my iPhone using Tapatalk
Nice one.


@ghost: no one knows how to do distributed consensus correctly. Or, stated differently, all consensus schemes achieve a different consensus with different criteria.

Thats why we also state that bts is not democratic and we can live with it bexlcause it need not be as a business

Offline sittingduck

  • Sr. Member
  • ****
  • Posts: 246
    • View Profile
Market cap has nothing to do with tech.  See doge.   


Sent from my iPhone using Tapatalk

Offline ghost.teller

  • Newbie
  • *
  • Posts: 2
    • View Profile
Quote
Only Delegated Proof of Stake ensures that all stakeholders have their opinion included in an economically viable manner.
He can cast his opinion but the message sounded too assertive to me.
Judging by the market cap, the consensus algorithm of Bitshares is not superior to others'.
So I thought it is arrogant.
Or the assertion really is correct and the market cap ... ??

Offline mint chocolate chip

Quote
Casper the friendly ghost
The friendliest ghost you know
Though grown ups might look in anger and fright
The children all love him so

Casper the friendly ghost
He couldn't be bad or mean
He'll romp and play, sing and dance all day
The friendliest ghost you've seen

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
Anyone worried about how arrogant Bitshares' Bytemaster will be conceived of by public?
s/bytemaster/vitalik/?!

why shouldnt bm cast his opinion? Everyone doea and should do so!
if i had the time and knowledge it would cast my opinion more publically too

Offline ghost.teller

  • Newbie
  • *
  • Posts: 2
    • View Profile
Anyone worried about how arrogant Bitshares' Bytemaster will be conceived of by public?

Offline Pheonike


Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
You cannot really build an asic that runs a turing complete algorithm as efficiently as hashing asics .. you NEED a general-purpose architecture

I dont know what the possibility is but, you can build a 256bit general purpose turing machine with additional casper specific instruction set.
Yes .. but it will never be as efficient as a "stupid" that only computes hashes .. You need more instructions to be Turing complete ..
That's all I wanted to say

Offline lakerta06

You cannot really build an asic that runs a turing complete algorithm as efficiently as hashing asics .. you NEED a general-purpose architecture

I dont know what the possibility is but, you can build a 256bit general purpose turing machine with additional casper specific instruction set.

Offline xeroc

  • Board Moderator
  • Hero Member
  • *****
  • Posts: 12922
  • ChainSquad GmbH
    • View Profile
    • ChainSquad GmbH
  • BitShares: xeroc
  • GitHub: xeroc
You cannot really build an asic that runs a turing complete algorithm as efficiently as hashing asics .. you NEED a general-purpose architecture

Offline Pheonike

You're not addressing the incentives of those who develop new full node tech, which is what really matters in the long run in my opinion. They will likely be incentivized to keep their competetive advantages proprietary and launch in competition with incumbents rather than collaborate. As far as I can tell, the result of the design is that ROI on node R&D is maximized, meaning TPS optimization, which is what I'd prefer from a common computation infrastructure.

Also, ignoring that, lets not forget that centralization of block production at worst only exposes people to what is essentially very complex and ridiculously expensive phishing attacks.

They're a theoretical problem, but easily solved since economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations). With that in mind Ethereum block validation is ultimately the transaction processing and anti-spam layer, not explicitly a trust layer. For that function casper seems to be the best algorithm invented yet.
Could you explain what you mean by "full node tech" and how the pishing attacks would work you mentioned?

With full node tech I mean hardware and software for processing the largest amount of EVM computations at the lowest possible cost, increasing marginal income and reducing marginal cost. The EVM has some exotic properties, such as a 256 bit word size, that means that right now it is ridiculously slow compared to a real computer, but with dedicated hardware it could actually become very fast.... when someone invents that in the future.

Oh and the phishing attack was just a metaphor for double spends, since they're both very similar types of digital fraud.
So you are saying there will custom hardware built for it like asics which will centralise the raw achievable performance to a handful of manufactures.

Offline liondani

  • Hero Member
  • *****
  • Posts: 3737
  • Inch by inch, play by play
    • View Profile
    • My detailed info
  • BitShares: liondani
  • GitHub: liondani
Wow! ethereum market cap dropped from $160M to $80M after your post! A 50% decline in one day!  :P
I hope we witness the same increase next time you post something... for bitshares   :D





« Last Edit: August 08, 2015, 01:55:21 pm by liondani »

Offline bytemaster

EVM = Ethereum Virtual Machine
For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

Offline btswildpig

  • Hero Member
  • *****
  • Posts: 1424
    • View Profile
You're not addressing the incentives of those who develop new full node tech, which is what really matters in the long run in my opinion. They will likely be incentivized to keep their competetive advantages proprietary and launch in competition with incumbents rather than collaborate. As far as I can tell, the result of the design is that ROI on node R&D is maximized, meaning TPS optimization, which is what I'd prefer from a common computation infrastructure.

Also, ignoring that, lets not forget that centralization of block production at worst only exposes people to what is essentially very complex and ridiculously expensive phishing attacks.

They're a theoretical problem, but easily solved since economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations). With that in mind Ethereum block validation is ultimately the transaction processing and anti-spam layer, not explicitly a trust layer. For that function casper seems to be the best algorithm invented yet.
Could you explain what you mean by "full node tech" and how the pishing attacks would work you mentioned?

With full node tech I mean hardware and software for processing the largest amount of EVM computations at the lowest possible cost, increasing marginal income and reducing marginal cost. The EVM has some exotic properties, such as a 256 bit word size, that means that right now it is ridiculously slow compared to a real computer, but with dedicated hardware it could actually become very fast.... when someone invents that in the future.

Oh and the phishing attack was just a metaphor for double spends, since they're both very similar types of digital fraud.
What is EVM stands for ?
这个是私人账号,表达的一切言论均不代表任何团队和任何人。This is my personal account , anything I said with this account will be my opinion alone and has nothing to do with any group.

Offline btswildpig

  • Hero Member
  • *****
  • Posts: 1424
    • View Profile
If we believe that 2.0 is all about innovation , then any "flaw" -----If there is any ----- can be solved by innovation .

 It's not like Bitcoin who can not develop a new version ...... It maybe totally different even 6 months later . It already happened with BitShares .

这个是私人账号,表达的一切言论均不代表任何团队和任何人。This is my personal account , anything I said with this account will be my opinion alone and has nothing to do with any group.

Offline Rune

  • Hero Member
  • *****
  • Posts: 1120
    • View Profile
You're not addressing the incentives of those who develop new full node tech, which is what really matters in the long run in my opinion. They will likely be incentivized to keep their competetive advantages proprietary and launch in competition with incumbents rather than collaborate. As far as I can tell, the result of the design is that ROI on node R&D is maximized, meaning TPS optimization, which is what I'd prefer from a common computation infrastructure.

Also, ignoring that, lets not forget that centralization of block production at worst only exposes people to what is essentially very complex and ridiculously expensive phishing attacks.

They're a theoretical problem, but easily solved since economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations). With that in mind Ethereum block validation is ultimately the transaction processing and anti-spam layer, not explicitly a trust layer. For that function casper seems to be the best algorithm invented yet.
Could you explain what you mean by "full node tech" and how the pishing attacks would work you mentioned?

With full node tech I mean hardware and software for processing the largest amount of EVM computations at the lowest possible cost, increasing marginal income and reducing marginal cost. The EVM has some exotic properties, such as a 256 bit word size, that means that right now it is ridiculously slow compared to a real computer, but with dedicated hardware it could actually become very fast.... when someone invents that in the future.

Oh and the phishing attack was just a metaphor for double spends, since they're both very similar types of digital fraud.


Offline santaclause102

  • Hero Member
  • *****
  • Posts: 2486
    • View Profile
You're not addressing the incentives of those who develop new full node tech, which is what really matters in the long run in my opinion. They will likely be incentivized to keep their competetive advantages proprietary and launch in competition with incumbents rather than collaborate. As far as I can tell, the result of the design is that ROI on node R&D is maximized, meaning TPS optimization, which is what I'd prefer from a common computation infrastructure.

Also, ignoring that, lets not forget that centralization of block production at worst only exposes people to what is essentially very complex and ridiculously expensive phishing attacks.

They're a theoretical problem, but easily solved since economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations). With that in mind Ethereum block validation is ultimately the transaction processing and anti-spam layer, not explicitly a trust layer. For that function casper seems to be the best algorithm invented yet.
Could you explain what you mean by "full node tech" and how the pishing attacks would work you mentioned?

Offline Rune

  • Hero Member
  • *****
  • Posts: 1120
    • View Profile

economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations).

Could you explain this a bit more. Do you mean a DPOS-like consensus amongst the issuers of various types of collateral for pegged assets?

I mean a consensus layer built into the EVM that takes a portion of the income of one or more dapps, and in return validates transactions related to the dapp in "validation transactions" (which I guess could be described as mini-blocks - still need to think of a better name for it). I.e. privatized, modular, customizable consensus validation systems that can be rapidly iterated on.

Offline CLains

  • Hero Member
  • *****
  • Posts: 2606
    • View Profile
  • BitShares: clains

julian1

  • Guest

economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations).

Could you explain this a bit more. Do you mean a DPOS-like consensus amongst the issuers of various types of collateral for pegged assets?

Offline Rune

  • Hero Member
  • *****
  • Posts: 1120
    • View Profile
You're not addressing the incentives of those who develop new full node tech, which is what really matters in the long run in my opinion. They will likely be incentivized to keep their competetive advantages proprietary and launch in competition with incumbents rather than collaborate. As far as I can tell, the result of the design is that ROI on node R&D is maximized, meaning TPS optimization, which is what I'd prefer from a common computation infrastructure.

Also, ignoring that, lets not forget that centralization of block production at worst only exposes people to what is essentially very complex and ridiculously expensive phishing attacks.

They're a theoretical problem, but easily solved since economic dapps will run their own secondary consensus algorithms in the virtual machine to validate in a way that is tailored to their own security preferences (DPOS is probably gonna be a first choice in many situations). With that in mind Ethereum block validation is ultimately the transaction processing and anti-spam layer, not explicitly a trust layer. For that function casper seems to be the best algorithm invented yet.

Offline liondani

  • Hero Member
  • *****
  • Posts: 3737
  • Inch by inch, play by play
    • View Profile
    • My detailed info
  • BitShares: liondani
  • GitHub: liondani
"Once it is clear who the “deciders” are it becomes safe for everyone else on the network to simply follow the leader and place bets without actually performing any validation."

I think the "deciders" are on this list  :)   
https://etherchain.org/accounts

Offline Ben Mason

  • Hero Member
  • *****
  • Posts: 1070
  • Integrity & Innovation, powered by Bitshares
    • View Profile
  • BitShares: benjojo
What worries me is that some of these other projects don't seem to view a systemic centralised outcome as a fatal flaw.....

Nice review

Offline bytemaster

For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.