Welcome to the Dash Forum!

Please sign up to discuss the most innovative cryptocurrency!

RC4 issues, bugs & feature requests

Discussion in 'Official Developer Thread' started by flare, Aug 14, 2014.

Thread Status:
Not open for further replies.
  1. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    I'm not sure what you mean about "closed-source RC". As I understand the only closed-source is Darksend. Did you download the stable v.0.9.12.32 from darkcoin.io? Check your version.
     
  2. splawik21

    splawik21 Grizzled Member
    Dash Core Team Foundation Member Dash Support Group Moderator

    Joined:
    Apr 8, 2014
    Messages:
    1,916
    Likes Received:
    1,273
    Trophy Points:
    1,283
    yes I think he is talking abount 9.12.32
     
  3. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    Maybe he had some hidden running processes. Stable v.9.12.32 is running fine on my win7 with 2 Cores. Shouldn't be scared to run it.
     
    • Like Like x 1
  4. darkstrike420

    darkstrike420 Active Member

    Joined:
    Jul 1, 2014
    Messages:
    178
    Likes Received:
    136
    Trophy Points:
    103
    I am.

    moli Closed-source RC is the Darkcoin wallet version with Darksend. I don't have "hidden" processes. The CPU usage was strictly from darkcoin-qt.
     
  5. stonehedge

    stonehedge Well-known Member
    Foundation Member

    Joined:
    Jul 31, 2014
    Messages:
    696
    Likes Received:
    333
    Trophy Points:
    233
    Thanks Evan for closing out Jira-33.
     
  6. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    I think you can trust Evan with his closed source Darksend. Not trying to be critical... but you trust windows :p ... Plus DS will be open source soon.
     
  7. JGCMiner

    JGCMiner Moderator
    Moderator

    Joined:
    Jun 8, 2014
    Messages:
    358
    Likes Received:
    210
    Trophy Points:
    113

    Yeah, but we lost enforcement. That seems like it might be ok for now, but in the future leaving things to the pool's generosity could cause major problems if someone wants to do something malicious. Especially if they attempt such after DRK has gotten greater adoption and marketcap.

    As it is many pools only pay 10%. I just have a bad feeling about his strategy...
     
  8. DrkMiner

    DrkMiner Member

    Joined:
    Jun 7, 2014
    Messages:
    204
    Likes Received:
    63
    Trophy Points:
    88
    We need enforcement!

    In the last week (2 MN running) 3 pools didn't pay fee, 2 paid only 10%
    Thats 4 DRK lost.

    Whats stopping pools from using pre RC3 wallet and pay 0% fee?
     
    • Like Like x 1
  9. TaoOfSatoshi

    TaoOfSatoshi Grizzled Member

    Joined:
    Jul 15, 2014
    Messages:
    2,718
    Likes Received:
    2,613
    Trophy Points:
    1,183
    This. The goal is to have more Masternodes to make things more anonymous with less chance of bad actors disrupting proceedings.

    No enforcement = Less rewards = Less incentive to run MNs = Less MNs = Less security using DRK & more coins on exchanges!

    Bad situation. I will keep an open mind until I hear what Evan has to say about this issue.

    There are a lot of upset people about this.

    Tao
     
  10. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    I think you guys are too worried. It's a good thing that the MN system is over complicated, to discourage any copycats that want to replicate the same system.
     
    • Like Like x 1
  11. DrkMiner

    DrkMiner Member

    Joined:
    Jun 7, 2014
    Messages:
    204
    Likes Received:
    63
    Trophy Points:
    88
    Worried? we should... news/rummer like that need to be clarified ASAP. Its already caused some movement in the price.
     
  12. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    OKAY.. Let's read what Evan said again from site http://jira.darkcoin.qa/browse/DRK-33:

    "It turns out the masternode voting system is over complicated and somewhat risky to the network, so I've removed it (i think he means the bug that was causing "payout code skips blocks and duplicates payments") and am using the RC3 masternode payment system. This means we won't have enforcement (but he doesn't say this is a permanent decision, right?), but it fixes all of these issues and removes the risk to the network."

    Please, if you don't understand or don't know for sure, don't spread rumors and fears on BCT and wonder why the price falls. And please tell your friends not to. I am sure Evan will clarify his statement but right now it's night time in America and people do need to rest, you know?
     
    #252 moli, Aug 28, 2014
    Last edited by a moderator: Aug 28, 2014
  13. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    Evan's answer posted in BCT a few minutes ago:

    https://bitcointalk.org/index.php?topic=421615.msg8563225#msg8563225

    "The RC4 system has inherent issues that were documented in Jira, plus it also has a systematic risk to the network (Sporking is less risky than a hard fork, but there's a risk the network forks wouldn't actually go away when the spork was turned off after a failure). These few issues combined with the fact that we've reached 80-90% payment efficiency tells me that it's not worth the risk to the network to move from the RC3 payment system.

    Also, I have a separate plan I've been considering as an alternative that carries no risk and can be done at RC5's launch. Basically, I would set a minimum protocol version and boot anyone not running RC4 or later off the network. This should get us to 98%+ payments."
     
    • Like Like x 2
  14. stonehedge

    stonehedge Well-known Member
    Foundation Member

    Joined:
    Jul 31, 2014
    Messages:
    696
    Likes Received:
    333
    Trophy Points:
    233
    I have a question. I ask here because it isn't easy to find the answer elsewhere.

    How does the RC3 payment system work? How are masternodes being paid now? Totally randomly?
     
  15. fredairbex

    fredairbex New Member

    Joined:
    Aug 28, 2014
    Messages:
    2
    Likes Received:
    2
    Trophy Points:
    3
    Actually, on macosx, the wallet is crashing at startup when it tries to open the log file, which contains a space in it.
    Anyone else tried on macosx ?
     
  16. darkwing

    darkwing Active Member

    Joined:
    Apr 8, 2014
    Messages:
    149
    Likes Received:
    110
    Trophy Points:
    103
    Working fine for me.

    I made a new wallet and let it sync then sent all coins to it.
     
  17. flare

    flare Administrator
    Dash Core Team Moderator

    Joined:
    May 18, 2014
    Messages:
    2,287
    Likes Received:
    2,406
    Trophy Points:
    1,183
    RC3 payment system was live for 1.5 month and was working 95% good all the time - and beside some network propagation problems, where in rare occasions the masternode list was different on certain miner nodes, we saw a good random distribution of payments.

    Evan hoped to fix the network propagation issue by switching the payment system to a "Top 20 rank approach", but it seems this introduced more flaws, which led to

    http://jira.darkcoin.qa/browse/DRK-33 and
    http://jira.darkcoin.qa/browse/DRK-24

    So the rollback to RC3 payment system is very logical for me, as it worked most of the time.

    As for me, i am still searching for a new approach to make the payments more network intrinsic - MNs should actually mine their coins by proof of service, not rely on good acting miners to share their mined coins.

    If we can somehow create a model where miners (PoW) and masternodes (PoSVC) mine independently from eachother, we will even have a solution for further payments for "DarkTor" relays and exit nodes. The current approach (divide PoW reward) seems not flexible enough to achieve this.
     
    • Like Like x 4
  18. stonehedge

    stonehedge Well-known Member
    Foundation Member

    Joined:
    Jul 31, 2014
    Messages:
    696
    Likes Received:
    333
    Trophy Points:
    233
    Seems like sound logic to me. Cheers flare. Masternode owners do need to know precisely how payments are being run currently though. I'm assuming total random selection which is fair enough.
     
  19. moli

    moli Grizzled Member

    Joined:
    Aug 5, 2014
    Messages:
    3,261
    Likes Received:
    1,837
    Trophy Points:
    1,183
    I'm guessing the current MN payment is still based on this system that Evan explained here:

    https://darkcointalk.org/threads/masternode-payment-variance.1636/

    It seems to me the MN system is needed for a very important task but at the same time the plan is not to encourage every darkcoin to be stashed away in a masternode, hence, the random selection for payment is implemented. A very smart financial thinking.
     
    #259 moli, Aug 28, 2014
    Last edited by a moderator: Aug 28, 2014
    • Like Like x 1
  20. Probe

    Probe New Member

    Joined:
    May 28, 2014
    Messages:
    25
    Likes Received:
    3
    Trophy Points:
    3

    Nothing is happening.
    After 20 hours of waiting NOTHING .
    Here is the debuglog
     

    Attached Files:

    #260 Probe, Aug 28, 2014
    Last edited by a moderator: Aug 28, 2014
  21. thelonecrouton

    thelonecrouton Well-known Member
    Foundation Member

    Joined:
    Apr 15, 2014
    Messages:
    1,135
    Likes Received:
    813
    Trophy Points:
    283
    I had a similar notion earlier that I posted on BCT:

    One possible solution that occurs to me would be a slight change to the X11 algo to go with the update for the rest of us, but that's a big undertaking. Or maybe it isn't, I don't know? If the algo changes then pools/p2pool would have to update for their mined blocks to be considered valid by the network, and the algo itself could include a MN fee check... but at this point I am speculating way above my pay grade... [​IMG]
     
  22. flare

    flare Administrator
    Dash Core Team Moderator

    Joined:
    May 18, 2014
    Messages:
    2,287
    Likes Received:
    2,406
    Trophy Points:
    1,183
    Yep, saw that - same idea: intrinsic masternode payouts.

    During one of my long jogs i thought about something like your modified X11 algo, where miners are dependant on a token from the masternodes to finalize their Proof the Work. But i came to the conclusion that it would introduce another attack/spoof vector.

    The current solution (miners share their reward) has the benefit to be simple, but is dependant on good actors. And as you wrote: Changing this is a big undertaking, regardess wether you change the algorithm, add another "Proof-of"-Layer or other changes to the Darkcoin blockchain logic.

    I read the TorCoin/TorPath paper, and really like the approach for a Proof-of-Bandwidth algorithm for paying incentives to the exit nodes. The difference between TorCoin and Darkcoin is that Darkcoin is already live, and the coins are not mined by the masternodes, but the miners.

    Nevertheless we will need to think about a model which enables us for more complex payout model when we turn to "DarkTor".

    Lots of things to consider to find the right trade off between technically possible and economically necessary :)
     
  23. thelonecrouton

    thelonecrouton Well-known Member
    Foundation Member

    Joined:
    Apr 15, 2014
    Messages:
    1,135
    Likes Received:
    813
    Trophy Points:
    283
    What about having some form of PoService p2pool running on the Masternodes, with each MN 'mining' on it?

    edit: I have absolutely no idea how this might be integrated with regular mining. :eek:
     
  24. flare

    flare Administrator
    Dash Core Team Moderator

    Joined:
    May 18, 2014
    Messages:
    2,287
    Likes Received:
    2,406
    Trophy Points:
    1,183
    Yep, my first thought as well - will not satisfy our requirements:

    a) ~ 20% of PoW coin supply distributed to active masternodes
    b) independent of no. of masternodes
    c) minor effect on maximum total coin supply
    d) fork proof
     
  25. crowning

    crowning Well-known Member

    Joined:
    May 29, 2014
    Messages:
    1,428
    Likes Received:
    2,005
    Trophy Points:
    183
    Proof of Stake for Masternodes.

    POS is an existing and well tested technology used by dozens of cryptocurrencies.

    Only Masternodes are allowed to do it with a fixed stake of 1000 DRK.

    Define the interest percentage in such a way that each masternode gets 1 DRK (or whatever else seems reasonable) per day.
     
    • Like Like x 1
  26. thelonecrouton

    thelonecrouton Well-known Member
    Foundation Member

    Joined:
    Apr 15, 2014
    Messages:
    1,135
    Likes Received:
    813
    Trophy Points:
    283
    I think the headache lies in the integration with the existing PoW scheme.

    If the mining software (pools and p2pool) could be modified to require some MN stamp of approval before being committed to the blockchain that would do it, but I don't know enough to venture how that might be accomplished.

    One (truly terrible) thought I had was that MNs could mine/stake/PoS a different coin altogether, perhaps MNDRK, which would be required to purchase services from the MN network. Set a 1:1 exchange rate somehow between MNDRK and DRK and require clients to pay in MNDRK for anonymisation of coins, messages, bandwidth, hosting etc.
     
  27. iHeartSmartArt

    Joined:
    May 2, 2014
    Messages:
    151
    Likes Received:
    59
    Trophy Points:
    88
    thelonecrouton I dont know if we want to divide DRK in that manner.
     
  28. thelonecrouton

    thelonecrouton Well-known Member
    Foundation Member

    Joined:
    Apr 15, 2014
    Messages:
    1,135
    Likes Received:
    813
    Trophy Points:
    283
    Neither do I, I did say it was a truly terrible idea. ;)
     
    • Like Like x 1
  29. innergy

    innergy Member

    Joined:
    Jun 27, 2014
    Messages:
    46
    Likes Received:
    13
    Trophy Points:
    48
    What if we split the MN voting system from mining. I mean, couldn't we use only one and the same address to collect every 20% block reward and then (and later) to send the funds from this address to selected MN on a regular time basis, randomly or by some algorithm? For example - every time when funds exceed 1DRK, we'll send exactly 1DRK to a MN. I don't know if there is a way to measure the masternode's provided proof-of service, but if there is, we can make a MN reward "waiting list" based on that.
     
    #269 innergy, Aug 30, 2014
    Last edited by a moderator: Aug 30, 2014
  30. David

    David Well-known Member
    Dash Support Group

    Joined:
    Jun 21, 2014
    Messages:
    618
    Likes Received:
    628
    Trophy Points:
    163
    That would require centralization--one single point of failure.
     
Thread Status:
Not open for further replies.