Proof of capacity Flawed??



  • I was having a discussion on FB about Burst Coin.
    Here was the man''s comment
    Proof of Capacity is flawed. It suffers from the "nothing at stake" problem inherent in Proof of Stake, as well as being subject to slow disk access times. Unless you have a RAM disk or, at the very least, an extremely fast SSD, you'd risk slowing down the network.

    Was wondering what you all think about this comment?



  • @PETENICE
    Could you explain the second statement?
    As the disk access time does not affect the 4 minute block time. As it is accounted for at difficulty adjustments.



  • @PETENICE is this his opinion or is there new research on proof-of-capacity?



  • @PETENICE said in Proof of capacity Flawed??:

    I was having a discussion on FB about Burst Coin.
    Here was the man''s comment
    Proof of Capacity is flawed. It suffers from the "nothing at stake" problem inherent in Proof of Stake, as well as being subject to slow disk access times. Unless you have a RAM disk or, at the very least, an extremely fast SSD, you'd risk slowing down the network.

    Was wondering what you all think about this comment?

    Sounds like your dealing with someone that is not aware of how the coin works.

    The size of your HDD or plots and the speed at which they are read (either with CPU or GPU) affects nothing other than your ability to mine for the best deadline in an expedited manor. The block chain still generates a block at a target time of 4 minutes.

    Nothing at stake? I'm not sure what that means. Only a few coins truly have something at stake and I believe the best example would be Sia Coin. You generate contracts to hold data and are paid for fulfilling that contract how ever if you break the contract the block chain reaches into your wallet and takes coins out.

    -IceBurst



  • @IceBurst
    The first statement is true. You can mine on all the chains you like. With 10 TB you can mine with full 10 TB on 2 forking chains.

    You risk nothing of your mining capacity if the network forks.

    This can only be solved with POW or similar concepts, as you literally work for every nonce you send every round.

    To give you an excample, let's say Burst forks to abc and xyz. They are both equality accepted and people are working to improve their favourite chain.

    As a miner, you want to earn as much as possible. Because of that, you choose to mine on both chains. You lose nothing, but just doubled your efficient capacity.



  • @FrilledShark said in Proof of capacity Flawed??:

    @IceBurst
    The first statement is true. You can mine on all the chains you like. With 10 TB you can mine with full 10 TB on 2 forking chains.

    You risk nothing of your mining capacity if the network forks.

    This can only be solved with POW or similar concepts, as you literally work for every nonce you send every round.

    To give you an excample, let's say Burst forks to abc and xyz. They are both equality accepted and people are working to improve their favourite chain.

    As a miner, you want to earn as much as possible. Because of that, you choose to mine on both chains. You lose nothing, but just doubled your efficient capacity.

    Interesting concept. However with your user account number built into the plot files and the need to have a valid reward recipient set how could one possibly have the same nounce accepted by two different chains with only a single reward assignment? I guess this works if you make the assumption that the forked chains will never be recombined.

    -IceBurst



  • @IceBurst

    The reward assignment is irrelevant. Nonces are always valid. (Unless there is a network wide ban on a certain set, which is more then unlikely)

    Unless the protocol is changed, the difference between my 10 TB on chain abc is no different from the same 10 TB on xyz.

    Btw, 2 forked chains can/will never be combined. They are spilt for good or bad. One of them will be dropped if when it gets irrelevant or outdated. The losers will lose everything on the forked chain, if it is dropped.



  • @FrilledShark So somebody can mine on few pools with same plots at the same time and get rewarded from both? That is a problem then



  • @Hyzi0 no it is not like this,
    in case there will be BURST and BURST CLASSIC, like ethereum..
    with the current POC you will be able to mine the 2 coins at the same time



  • @ccminer
    No, not 2 POC coins at the same time. 2 clones of Burst, with no change to the mining system or account system, at the same time.

    The account system, has to be the same. Otherwise, you might be giving a random person free coins. Aka, burning coins.

    The poc aspect has to be the same as well, otherwise you just can't mine.



  • @FrilledShark in fact you should change the plot file...



  • So is there a "fix" to the concept which is flawed PoC has been great so far right, can't say I understand the under laying workings but at a basic level - this must have been brought up before


  • admin

    Different PoC will ask for different Scoops at different times or even at the same time. This will lead to doubling the read time, doubling the CPU cycles and the HDD read head going back and forth on your disk, which lowers lifetime and read-speed.

    I can see that two or three chains could be mined on one machine but after a certain amount your CPU or GPU is 100% all the time like a PoW miner.



  • as for the OP i dont think that person fully understands how Burst work's for the second statement thats completly false as HDD speed read times have nothing to do directly with the network speed only for your personal ability to mine blocks within the network. as the network dif auto adjust itself and aims to target a 4 min block forge time. the first statement thou true from a theoretical view placed in a pre determined circumstance. from an applyed view works no different than say Bitcoin when looking at problem / result, Bitcoin and Other's at start of block Hash math equation's and submit them till block is forged, next block they start all over and do it agin. Burst saw this as a waste of energy. and insted Burst Hash math equation's and saves them to HDD, at start of block scan's a % of the saved hash and submit's it block is forged. the only applyed diff is Burst recycles it's Hashed equations to reduce resource's used. and does it in a way that prevents duplications of the Hash's



  • @iKnow0 his opinion.
    This guy is a big advocate for BTC, so that could explain his position.
    There are some discussions about proof of capacity and the nothing at stake concept.
    I am simply trying to understand all of this.



  • @PETENICE Welcome to the world of crypto, some find it hard to be objective if heavily invested in a particular coin.



  • You should mention to your friend that having different coins that require different proofs is good, as they can be combined on a single machine.

    For Example, a person running their own bitcoin pool can run an ASIC miner connecting to that pool, but then use the same pool rig with some high end video cards to GPU Mine some alt coins, have a wallet with coins in it on same machine for POS, PLUS run burstcoin on spare hard drives connected via USB. THe only real restriction here is RAM, because each of the different coins use a completly different protocol for mining.

    My rig for burstpool has a 8 core processor and 16 GB ram, runs burstcoin and CPU mining through nicehash, plus runs the pool quite well, and it is a computer sitting in my house running off of solar power.

    Being able to combine multiple coin mining on a single machine lowers electricity cost and overall rig costs



  • I just found this while researching your question: https://github.com/ethereum/wiki/wiki/Problems

    You'll see the problem stated fairly clearly in section 8. Proof of Stake

    I do need to do a bit more reading on this but initially, it seems that as Burst miners, we do have revenue "at stake," as I believe @daWallet was referring to above. If forked chains begin to produce different block heights, it does effect our mining efficiencies and potential incomes. Especially if one of the chains is dropped by the network at large (wasted time and energy in terms of wear on our hardware).



  • @k.coins What about all the blocks you don't win? Would you consider that "at stake"? Because that is also wasted energy and time... If so that is something all the crypto algos i know until today have xD

    In physics, when in school, you learn about the perfect machine (wich don't have any wasted energy so 100% work) and you learn that is utopic and impossible... The perfect is something that don't exist, is something that we try to reach everyday but it will be always possible to do in a better, faster or more efficient way ;D



  • These sorts of questions about proof happened with Burst Coin. Typically it meant that people were actively scaring off new people causing it to eventually centralize.


Log in to reply
 

Looks like your connection to Burst - Efficient HDD Mining was lost, please wait while we try to reconnect.