#bitcoin-wizards

/

      • snakesandbarrels joined the channel
      • fabianfabian has quit
      • metamarc has quit
      • cypher__ joined the channel
      • snakesandbarrels has quit
      • drexel joined the channel
      • metamarc joined the channel
      • drexel has quit
      • metamarc has quit
      • metamarc joined the channel
      • metamarc has quit
      • metamarc joined the channel
      • cypher__ has quit
      • hashtagg_ joined the channel
      • drexel joined the channel
      • agorist000 joined the channel
      • agorist000 has quit
      • agorist000 joined the channel
      • metamarc has quit
      • drexel has quit
      • metamarc joined the channel
      • metamarc has quit
      • metamarc joined the channel
      • alpalp joined the channel
      • agorist000 has quit
      • hearn_ joined the channel
      • hearn has quit
      • hashtag_ joined the channel
      • hashtagg_ has quit
      • hearn joined the channel
      • hearn_ has quit
      • hearn has quit
      • hearn_ joined the channel
      • melvster joined the channel
      • kanzure
        if you have multiple proof-of-work algorithm variations, how would you judge the relative merit between their outputs? i can only think of arbitrary answers.
      • gmaxwell
        For what application?
      • kanzure
        eh, just your typical "reset pow to fiddle with decentralization" (loosely- nothing specific)
      • gmaxwell
        I know nothing but $arbritary for that. (For hashcash you can just blinding increment the cost when the rate is high in order to rate limit requests; esp if you can distinguish bad requests after the fact.)
      • kanzure
        i have heard the idea to "flip the switch" and go to a different pow algorithm before, but it seems more natural to have a geometric drop-off of one pow algorithm while switching to another, or keeping sha256 pow and then adding another one instead of just a hard switch
      • and by "adding" i mean "blocks can be verified using this second verification method" i guess
      • gmaxwell
        kanzure: if you're thinking about concurrent-- you're aware that the whole argument for "Multiple algorithims to improve decentralization" is probably totally busted, right? Needing more circut area increases R&D costs and one time mfg costs, but not operating; and thus probably is a centeralizing force (fewer parties can make hardware).
      • Oh a transistion, well for that you can have a programmed ratio that changes over time.
      • kanzure
        is a ratio equally busted, or less busted?
      • gmaxwell
        I think as a transition mechenism that could work. The argument I was giving above basically only holds for the system that has multiple ones in the steady state.
      • kanzure
        right, also i want to clarify i definitely don't mean "sha256 AND scrypt == 1 hash"
      • nsh
        hmm
      • kanzure
        also i'm curious how controversial alternative-pow proposals have been in the past.
      • obv. it pisses off miners but... ratios seem less likely to, since they already factor hardware depreciation into their plans anyway?
      • *transitory ratios
      • alpalp
        Changing the algorithm has pretty big incentives on any investment on the new algorithm since what's to stop you from changing it again?
      • hearn joined the channel
      • kanzure
        well, probably nothing
      • but allocating 1% of blocks to sha256-next mig-- actually you don't need to allocate BTC supply to it, you could allocate some chunk of the transaction fees.
      • gmaxwell
        I tried to get litecoin to change their POW to make a best effort attempt to make good on their social contract of <s>GPU</s> <s>FPGA</s> ASIC proofness but they did not decide to do so.
      • nsh
        who's you? the miner signs away some of their transaction fees, or some fraction of blocks get mined under a secondary hash by difficulty equivocation?
      • kanzure
        well i think this would just be a systemic detail, not something miners would be able to control, although a voluntary mechanism would also be curious
      • nsh
        voluntary mechanism seems nontrivial
      • winning a lottery is easy to do for yourself, hard to do for someone else
      • well, actually i guess miners could include the lowest hash presented by a connected node
      • so you can voluntarily objectively reward other participants
      • but it's still voluntary, which is an issue
      • a problem with allowing more than one type of hash to mine a block is that each block measures the global hashrate only for a single hash type, so you have a harder block period convergence problem
      • bendavenport has quit
      • hearn joined the channel
      • jaekwon joined the channel
      • sparetire_ joined the channel
      • StephenM347 has quit
      • kanzure
        re: relative merit, that was some thought along the lines of "BTC has to be allocated between them in some split"
      • btcdrak has quit
      • dEBRUYNE joined the channel
      • hearn joined the channel
      • hearn_ joined the channel
      • hashtagg_ joined the channel
      • jnpn has quit
      • AaronvanW has quit
      • jtimon joined the channel
      • RedEmerald_ has quit
      • hashtag_ joined the channel
      • hashtagg_ has quit
      • RedEmerald joined the channel
      • hashtagg_ joined the channel
      • Oizopower has quit
      • hearn joined the channel
      • dEBRUYNE has quit
      • cluckj joined the channel
      • hashtag_ joined the channel
      • tromp___ joined the channel
      • tromp has quit
      • hashtagg_ has quit
      • jnpn joined the channel
      • c-cex-yuriy has quit
      • ryanxcharles joined the channel
      • msdkxla joined the channel
      • hashtagg_ joined the channel
      • msdkxla has quit
      • blackwraith has quit
      • hashtag_ joined the channel
      • hashtagg_ has quit
      • mrkent has quit
      • hashtagg_ joined the channel
      • hashtag_ has quit
      • Giszmo has quit
      • jaekwon joined the channel
      • hashtag_ joined the channel
      • hashtagg_ has quit
      • hearn joined the channel
      • hashtagg_ joined the channel
      • hashtag_ joined the channel
      • hearn_ joined the channel
      • hashtagg_ joined the channel
      • lmatteis has quit
      • hashtag_ has quit
      • alpalp has quit
      • TheSeven joined the channel
      • chmod755 joined the channel
      • wallet42 joined the channel
      • ThomasV joined the channel
      • hearn joined the channel
      • hearn has quit
      • hearn__ joined the channel
      • hearn_ has quit
      • wallet42 has quit
      • chmod755 has quit
      • bendavenport joined the channel
      • bramc joined the channel
      • ThomasV has quit
      • hearn joined the channel
      • hashtag_ joined the channel
      • copumpkin has quit
      • tromp___ has quit
      • ThomasV joined the channel
      • NewLiberty has quit
      • drexel joined the channel
      • drexel has quit
      • hashtagg_ joined the channel
      • ryanxcharles joined the channel
      • hearn joined the channel
      • btcdrak joined the channel
      • hashtag_ joined the channel
      • tromp joined the channel
      • jnpn has quit
      • tromp has quit
      • fabianfabian joined the channel
      • fabianfabian has quit
      • bendavenport has quit
      • hashtagg_ joined the channel
      • mjerr joined the channel
      • davi joined the channel
      • go1111111 has quit
      • priidu joined the channel