51200+ entries in 0.023s
mircea_popescu: then the "10" can be configurable ; or else it can be calculated by the node, (such as "117% of average time it took me to check the past 144 blocks" for instance)
mircea_popescu: how about "pushgetblock to ONE peer randomly selected from the list of peers advertising higher block tip than you have upon every block verification (whether successful or not) and every 10 minutes henceforth"
mircea_popescu: "every 10 minutes pushgetblocks to all peers that advertise higher height than yours" is still wasteful, you will ask 114 people for the same item with your new cuts.
mircea_popescu: trinque i just got to watch like 4 or 5 different fireworks displays.
mircea_popescu: it's in the channel as an ad-hoc thing, "this person is voiced atm, if that's worth anything to you". not intended to prove anything to archeologists.
mircea_popescu: separate items. i think the sha was in the ml name to prevent collisions ;
mircea_popescu: danielpbarron yeah, not too crazy about that behaviour either.
mircea_popescu: for teh first time in many weeks, i dunno what to say.
mircea_popescu: nobody will note nor should they. why does it need renaming ?
mircea_popescu: fuck 'em, the salvation of the drowning is work for the hands of the self-same drowning. god may mercifully send you a sign, but that's already asking for too much.
mircea_popescu: hardest thing to look at is how you wasted your life, halfway in. so, yeah.
mircea_popescu: i'm curious what flavour of muh feelings gets pulled this time.
mircea_popescu: ben_vulpes if you (or anybody) has a contact for this gregory pery tard (i dun imagine totallynotrobots@protonmail.com actually works ?), maybe point out to him that
http://btcbase.org/log/2017-12-24#1758148 and it's really ridoinculous to be wasting his weeks to freely support the usg pile of evil ?
☝︎ mircea_popescu: "What’s weird though is that the OAuth2 token doesn’t ever change, and I can’t find anything in the Uber developer documentation that deals with token expiration;" << word.
mircea_popescu: that's just because you're lazy. real men always work with tools designed to work counterproductivelty and maim them in the process.
mircea_popescu: it'll be cool when they "upgrade" it to ipv6 and the 000ffff mask will have to be updated to correctly handle unicode.
mircea_popescu: it's one thing to have it in the press, esp considering the history of how that came about. but as obvious knob as it gets.
mircea_popescu: myeah. anyway, stopgap of the "if i make it what i think it should be -- it crashes" line.
mircea_popescu: recall we had a discussion re star breaking nat crapolade. but you need the stars and a system specifically designed to use them, as there contemplate
mircea_popescu: this is the problem here : you get a bunch of "nodes" which only connect out, and a few nodes who get hammered. and since no serious thought was given to
mircea_popescu: ben_vulpes a node behind a NAT is not capable of being addressed. so it receives no inbound, merely connects out. so it does not participate in lessening the load.
mircea_popescu: if that limit is not there, you can readily encounter a situation where one of the 10 is taken down every other hour by 90 inbound connections. or at least -- more readily.
mircea_popescu: with the limit of 8, you have the 10 receiving no more than 720 connections, ie 7.2 on average, with a linear distrib max of 15 ? so you can live a standard of "and can handle 16 conns" for a "proper" box
mircea_popescu: ben_vulpes suppose your relay network is 100 nodes ; suppose 10 of those are real nodes and 90 of those bullshit nodes behind a nat in a zek's house somewhere.
mircea_popescu: (if node is behind nat -- it is a drag and naught else)
mircea_popescu: it was important, to limit the derp nodes from choking the network.
mircea_popescu: when you worship your betters you ensure your place in the world, ~such as it is~. when you worship clitler you get nice words and utter doom.
mircea_popescu: 2020 "young writer for the ny times" ain't gonna be 1950s ginsburg, as it ain't gonna be 1970s mia farrow. it's gonna be fungusbrain bigmomma.
mircea_popescu: but the same drive to ever cheapen the soviet's offerings to the zek ("only political training contributes to forwarding teh revolution!! everything else contributes to costs!!!" said teh world famous elena ceausescu-zigglar) will cheapen the very rug from under your feet also.
mircea_popescu: i know it seemed all great and fabulous back when deskilling brought about abominables a la condolezza rice as "secretary of state" notwithstanding strict incapacity of any sort of thinking
mircea_popescu: and mind all ye of the "aspirational" demographic, all ye dumb pantsuited hussies : the sword keeps right on cutting after it cut the slice you wanted.
mircea_popescu: (for bonus lulz, 12795/30/24 = 17.770833333 or roughly speaking 10 to the bitcent.)
mircea_popescu: or that weird geek from early pedo-pedia days, what's his name
mircea_popescu: asciilifeform what's that prove tho ? winklevoss tards periodically go into their basement/den where they "have" "their own" copy of fb.
mircea_popescu: an evident issue would be of course that if ownership is implemented as-to-specified-coin as seen in bitcoin, then E has a ready avenue -- "all txn including spends from my own addresses are mine". this is a somewhat weakening of the premise, but perhaps sufficient.
mircea_popescu: "especially because it agrees with logotron time oracles". sure...
mircea_popescu: but in point of fact the hook to hang by is there : at the present time how would someone who is hostile (ie, does not trust what we say) discern who originated mpi ?
mircea_popescu: asciilifeform do you propose E's "process of elimination" is simple bit-comparison of P and P' ?
mircea_popescu: much like "you can't repudiate signatures -- can't see whole space" discussion in that thread.
mircea_popescu: see, the SELECTION of .sigs you see is not promised to cover the whole space!
mircea_popescu: at t1, E perceives P', signed by l1...lk ; which is not an isomorphism of l1...lm, and might wel lnot even include E
mircea_popescu: at t0, E produces P and signs it. E shares P signed by E with l1...lm.
mircea_popescu: at this point, how would E discern which party of l1..ln originated P ?
mircea_popescu: let observer E know l1...li of this wot. let observer E observe a P signed by a subset l1...lm of his wot.
mircea_popescu: asciilifeform let's model this. let "patch" be a bitfield ; let wot be comprised of l1...ln.
mircea_popescu: ie, it's leveraged to fuck over the miner collusion properties of current pow.
mircea_popescu: this is a sort of gossipd-meets-the original "can't have signature repudiation" problem you encountered in comments there, but REVERSED.
mircea_popescu: nobody can know WHO truly authored a patch. just what set of the signatories signed it that they're willing to share with him