#bitcoin-wizards

/

      • tjader joined the channel
      • mrkent joined the channel
      • AvatarA joined the channel
      • laurentmt joined the channel
      • hashtag_ joined the channel
      • conner joined the channel
      • GGuyZ has quit
      • zookolaptop has quit
      • shovel_boss joined the channel
      • laurentmt has quit
      • bsm117532
        bramc: It's definitely a whiny ragequit.
      • voxelot joined the channel
      • fewfef joined the channel
      • binary01 has quit
      • tjader has quit
      • epscy joined the channel
      • mrkent_ joined the channel
      • tjader joined the channel
      • supasonic joined the channel
      • adnn joined the channel
      • zookolaptop joined the channel
      • smooth
        it is, but calling it that is behaving no better
      • brg444
        smooth Sorry Core devs can't pay for ny times feature
      • bsm117532
        To the points at the end of bramc's post: there is NO RIGHT ANSWER to "what is the right block size". Bitcoin uses an inappropriate data structure (a linked list). The solution is some deeper re-engineering to remove the problem entirely. We need to buy time to do that engineering. So a near-term bump is a reasonable option. Segwit is good for a whole host of other reasons that have nothing to do with block size
      • copumpkin joined the channel
      • GGuyZ joined the channel
      • smooth
        there clearly is a right answer for any given state of technology and marketplace. Identifying it may be difficult.
      • and other solutions may be better still
      • brg444
        bsm117532: No one is disputing this AFAIK but as you mention needing appropriate time it's important for everyone to acknowledge this is a manufactured "crisis" that aims to precipitate the due process
      • bsm117532
        brg444: I think the "next solution" has not emerged yet, so people are being stupid. I'm working as fast as I can, and I'm not the only one.
      • ThomasV joined the channel
      • People who don't do software engineering can't even imagine another solution...
      • brg444
        bsm117532: change a constant! "Off with their head!"
      • Tiraspol has quit
      • Tiraspol joined the channel
      • Tiraspol has quit
      • Tiraspol joined the channel
      • bramc
        The blocksize bump in segwit is clearly the pragmatic thing to right now given current technical and political realities
      • bsm117532
        The blocksize bump is necessary no matter what, because of current transaction volume realities. Segwit is very useful far beyond an effective block size increase, but is (probably) too far out to solve the near-term problem of hitting full blocks on the regular.
      • brg444
        is it?
      • gmaxwell
        bsm117532: you're working from an offensively wrong premise.
      • brg444
        I'd rather have full blocks then precipitate a hard fork that leaves people behind.
      • adnn
        ...aka the crux of the debate
      • bsm117532
        gmaxwell: elaborate?
      • have I offended you?
      • gmaxwell
        There is nothing wrong with full blocks, and blocks have been "full" relative to what miners would produce for _years_. Full blocks is the natural state of the system: The demand for externalized-cost highly replicated external storage at price zero is effectively infinite.
      • bsm117532: A little, you've been around here long enough to have absorbed this view-- even if you don't agree with it, stating something 180 degress off as a simple fact is a bit frustrating to deal with :-/
      • zfsoja has quit
      • alpalp
        the only reason we dont consistently have "full" blocks is relay limits and miners too lazy to change things
      • bsm117532
        gmaxwell: We must solve this problem while the fee profit of miners is ~0.2%, and the effectively are not exerting an opinion. If fee profits were higher, they would lobby to keep them, at the expense of small blocks, and zero growth of the system.
      • dgenr8
        alpalp: here i thought it was because not that many people used bitcoin yet
      • bsm117532
        We are far too young in bitcoin's history to have this kind of pressure...
      • alpalp
        dgenr8, using bitcoin could be using the blockchain to backup your hard drive if there was no fee pressure.
      • bsm117532, pure FUD, what is the worst that happens?
      • bsm117532
        gmaxwell: Are you arguing "fee pressure is good" and therefore small blocks and zero growth are desirable?
      • midnightmagic
        :-/ Is rehashing old pointless, already-rehashed discussion with lots of bald assertions really what you're going over, dude..
      • gmaxwell
        bsm117532: Stop with the absolutist language, shades of Hearn. "We must" -- on what objective basis?
      • midnightmagic
        in -wizards
      • bsm117532
        Uh. confused.
      • dgenr8
        alpalp: and yet even that is not happening
      • bsm117532
        alpalp: What is FUD?
      • frankenmint joined the channel
      • alpalp
        bsm117532, that "we are far too young". What is worst case scenario? fees go up a bit, and people use alternatives for a little?
      • dgenr8, thats because we have minimum relay fees and a block size limit!
      • gmaxwell
        bsm117532: fee pressure is an intentional part of the system design and to the best of the current understanding essential for the system's long term survial. So, uh, yes. It's good.
      • dgenr8
        limit was hit exactly once prior to September 2014
      • midnightmagic
        dgenr8: Also false.
      • dgenr8
        midnightmagic: perhaps bsm117532 means that everything good that's happened for bitcoin's
      • bsm117532
        Fee pressure is an artificial imposition on the system.
      • dgenr8
        midnightmagic: sorry ... here you go http://i.imgur.com/5Gfh9CW.png
      • bsm117532
        It is not caused by lack of bandwith or computation. It's is a constant in a file and we can do better.
      • copumpkin has quit
      • alpalp
        bsm117532, stupid satoshi putting fees in the original design
      • midnightmagic
        geh, so not -wizard material.
      • shovel_boss_ joined the channel
      • bsm117532
        I'm taking this to #bitcoin
      • dEBRUYNE joined the channel
      • shovel_boss has quit
      • ThomasV has quit
      • shovel_boss joined the channel
      • shovel_boss_ has quit
      • bsm117532 disengages from the toxic block size debate, and goes back to braids, which might be a better long-term solution...
      • zookolaptop
        Yay braids
      • LeMiner has quit
      • PeterR joined the channel
      • PeterR
        Here is a visual of the empirical block size distribution since Block 0: http://imgur.com/uPneQ5N,izo03ie
      • hashtag has quit
      • PeterR has quit
      • melvster joined the channel
      • bsm117532
        There is no number, or algorithm WRT said number, that is the "right" solution. The correct solution is to remove the number from the problem.
      • brg444
        anybody find these drive bys hilarious
      • bsm117532
        PeterR that's a good plot.
      • PeterR wants my braids paper. This is why I'm at the computer on a Saturday night...
      • zookolaptop
        bsm117532: hold on just let me get ops power and I'll ban you from -wizards until 24 hours from now.
      • Thank me later!
      • Actually could you ban me from twitter?
      • raedah has quit
      • rustyn has quit
      • rustyn joined the channel
      • bsm117532
        OTOH I found a new algorithm to find cohorts, which is O(N). (cohort is kind-of the analog of a block -- you don't want it to take a long time to find which subset of incoming data is the right one to analyze...)
      • PeterR joined the channel
      • binaryFate has quit
      • PeterR
        bsm117532: Thanks! I like those two animations because they illustrate how the block size is empirically different than the block reward.
      • There have been "obstructions" to the block size in the past at 250kb, 350kb, 750kb and now at 1MB, but "Bitcoin" has always broken them down.
      • bsm117532: Yes, looking forward to the paper. And I believe Chris will be in touch shortly about one for you to review. Thanks :)
      • kanzure
        bsm117532: bitcoin itself is an arbitrary imposition, i think your statement about fee pressure is a non-statement (re: "arbitrary imposition").
      • bsm117532
        PeterR: I received a review request today and confirmed on the website, I'm happy to review!
      • PeterR
        Perfect!
      • tjader has quit
      • bsm117532
        kanzure: I'm referring to the block size as a number, disconnected from the bandwidth. In principle the actual restriction should be the bandwidth, but it's not because of a number in the protocol, which is arbitrary.
      • Also the fact that a "block" is a very spiky bandwidth distribution...
      • kanzure
        well you can't measure bandwidth in a decentralized system anyway
      • so i'm not sure why you would mention the bandwidth here? all we can do is just try our best to make the bandwidth requirements to go down.
      • bsm117532
        My point is: 1MB/10min=1.6kb/s is not a restriction that anyone is concerned about.
      • _rht joined the channel
      • kanzure
        the actual restriction should be a protofcol-level rule to avoid the impossibility of measuring bandwidth
      • raedah joined the channel
      • bsm117532
        The actual bandwidth requirement is (block size)/(block verification time) which is the SPV mining window.
      • We'd be better off spreading the load over 10 min, if it didn't cause orphans...
      • tjader joined the channel
      • r0ach
        Surprised virtually nobody mentions the ethical issue of block size involving utility vs greater fool theory: https://bitcointalk.org/index.php?topic=178336....
      • brg444
      • bsm117532
        "It's funny how the dev side has more small blockers, while the speculation side is almost entirely large blockers nowadays. Why is this? Because the dev side doesn't understand markets," -- No, it's because the market side doesn't understand software engineering. This is an artificial restriction. There are other ways to lift it.
      • alpalp
        what makes you think the speculation side is in favor of big blocks?
      • bsm117532
        I'm disengaging from this argument...
      • Not very good at it, I grant you...
      • GGuyZ has quit
      • adnn
        alpalp: It is my understanding that some Chinese miners are concerned that block size dead lock is hurting the price and block size bump will prove that we "can do it"
      • ironically their most recent calls to action have hurt the price further.
      • alpalp
        adnn, maybe that golden egg laying goose is full of gold!
      • lets kill it!
      • adnn
        lol
      • rusty joined the channel
      • pigeons
        ;;topic
      • gribble
        This channel is for discussing theoretical ideas with regard to cryptocurrencies, not about short-term Bitcoin development | http://bitcoin.ninja/ | This channel is logged. | For logs and more information, visit http://bitcoin.ninja
      • nuke1989 has quit
      • IAmNotDorian has quit
      • rusty has left the channel
      • PeterR has quit
      • Giszmo has quit
      • conner has quit
      • dEBRUYNE has quit
      • conner joined the channel
      • moa
        bsm117532 1MB/10min=1.6kB/s = 13.3 kb/s
      • bsm117532
        I've never heard anyone claim they couldn't get a 14.4 kb/s modem in 2016.
      • That's laughable.
      • oneeman
        bsm1175321: where can I read about braids? Just https://scalingbitcoin.org/hongkong2015/present... or is there also a more 'texty' whitepaper or something?
      • bsm117532
        oneeman: Working on it now, paper will be out in ~days...
      • oneeman
        sweet
      • moa
        I was simply pointing out your mixing bits and bytes units nomenclature and the actual bandwidth requirement for full blocks is at least 2xblocksize (some say 8x)