Show Idle (>14 d.) Chans

← 2017-12-20 | 2017-12-22 →
trinque: << I have been fighting stuck/slow transactions more frequently in the past month, leading to the last week where I have been completely unable to get connected nodes to take my txn.
a111: Logged on 2017-12-21 02:17 ben_vulpes: phf: well i'd like to get something out of trb that i can dump into a txrelayulator
trinque: the version number appears to be a factor which ends up isolating trb nodes, hypothesis being that the version number being set high invites nodes to insult the malleus patch.
trinque: my own nodes have begun lagging far behind the chain tip, other folks report the same.
trinque: for now, I can report that setting the version number to 50400, a trb node will catch back up, txns will unstick.
trinque: not a proposal for action, but those are my observations.
mircea_popescu: there was a lot of spam pre block 500k.
mircea_popescu: may alsoi be a factor.
trinque: was on 50000x, within 10 or so of top now, after version change and restart.
trinque: prior to that the lag was increasing every time I checked.
trinque will be comparing multiple test subjects.
trinque: and there's 500358
trinque: ah, and 9. topped out.
BingoBoingo: Spanish test a success. Was a wonderful opportunity to get forced into a prolongued stretch of not head translating. Thankfully the test was very patient and ergonomic.
BingoBoingo may have cheated a bit on the test leaning on the whole "animal magnetism" applied Trilema business.
BingoBoingo: The accent on the test though, the single most challenging encountered so far.
BingoBoingo: Cada dia, hacer algo por su Trilema. Lo trabaja cuando vos trabajas
diana_coman: congrats BingoBoingo
BingoBoingo: !~ticker --market all
jhvh1: BingoBoingo: Bitstamp BTCUSD last: 16329.54, vol: 21602.60607456 | Bitfinex BTCUSD last: 16248.0, vol: 59133.22186753 | Kraken BTCUSD last: 16022.0, vol: 4129.91578854 | Volume-weighted last average: 16257.7579482
BingoBoingo also this morning returned to the consultora to sign more forms on the road to having a bank account.
BingoBoingo: And the data center has sent draft contracts to review
BingoBoingo: Y en otra inglesas: "In other news, The Times reported that, for the first time in modern history, no Royal Navy fighting ships – destroyers, frigates, amphibious assault ships or carriers – would be operating away from Britain over Christmas and New Year."
BingoBoingo: Inglaterra es un pais pobre
asciilifeform: << i dug into this item last yr, and found that the version thing is mostly a red herring : when you set 50400, the same derps drop ~you~, a millisecond or so sooner than you would've dropped ~them~ via malleus. the real culprit is always .
a111: Logged on 2017-12-21 05:20 trinque: the version number appears to be a factor which ends up isolating trb nodes, hypothesis being that the version number being set high invites nodes to insult the malleus patch.
a111: Logged on 2017-08-23 13:09 asciilifeform: the reason why mike_c's node appears 'stalled' is that it stopped fetching blocks. trb's block sync behaviour is unspeakably moronic, it will attempt 'long sync' ONCE, on warmup
asciilifeform: aside from brief period at bootup, a trb node DOES NOT TRY TO SYNC AT ALL, passively waits for someone to come and feed it. forever.
a111: Logged on 2017-03-04 00:52 asciilifeform: and are the only times a trb node asks for blocks explicitly from peer
asciilifeform: + , and really whole thread worth reviewing.
a111: Logged on 2017-03-04 00:55 asciilifeform: so mircea_popescu , as you can probably tell, if node misses the window when $block was being actively thrown at it, then it has only these two knobs for attempting to get it
asciilifeform: so trinque , yer node synced on account of being reset. as in more or less every previous case of 'my node stalled, i changed knob K, and reset, and bam! , synced , knob K must be reason' -- it's the reset.
asciilifeform: i'ma put this in the l0gz, ftr :
asciilifeform: the ONLY 2 times PushGetBlocks is called ( i.e. to explicitly ASK a peer for new blocks ) is :
asciilifeform: 1) inside if (strCommand == "version") { .... that is , command processor for peer's commands :
asciilifeform: // Ask the first connected node for block updates
asciilifeform: static int nAskedForBlocks;
asciilifeform: if (!pfrom->fClient &&
asciilifeform: (pfrom->nVersion < 32000 || pfrom->nVersion >= 32400) &&
asciilifeform: (nAskedForBlocks < 1 || vNodes.size() <= 1))
asciilifeform: nAskedForBlocks++;
asciilifeform: pfrom->PushGetBlocks(pindexBest, uint256(0));
asciilifeform: the very FIRST connected peer who issues 'version' gets asked to help sync. and when THAT connection breaks -- that's it.
asciilifeform: 2) inside the bastardry handler, in ProcessBlock(...) :
asciilifeform: // If don't already have its previous block, throw it out!
asciilifeform: if (!mapBlockIndex.count(pblock->hashPrevBlock))
asciilifeform: printf("ProcessBlock: BASTARD BLOCK, prev=%s, DISCARDED\n", pblock->hashPrevBlock.ToString().substr(0,20).c_str());
asciilifeform: // Ask this guy to fill in what we're missing
asciilifeform: if (pfrom)
asciilifeform: pfrom->PushGetBlocks(pindexBest, pblock->hashPrevBlock);
asciilifeform: return true;
a111: Logged on 2017-03-04 01:11 asciilifeform: note that ProcessBlock() only ever asks ~the bastard-supplying peer~ for the bastard's prevblock.
asciilifeform: apologies for the clutter, but this item really ought to live in the permarecord.
asciilifeform: because it is imho a major open problem .
asciilifeform: much of what the 'power rangers' did to their bitcoin, was an elaborate dance around this problem, with a dozen pseudosolutions that guzzled memory, and -- more importantly -- destroyed the integrity of their sync ( the orphanage bullshit, the headers-first bullshit , etc )
shinohai: !!up andreiislost
deedbot: andreiislost voiced for 30 minutes.
BingoBoingo: !!Up freetlas
deedbot: freetlas voiced for 30 minutes.
BingoBoingo: freetlas: Habla! Quien es tu papa y Que lo hace el?
freetlas: BingoBoingo: Mark? Is it you?
BingoBoingo: Quien es Mark?
freetlas: Zuckerberg
BingoBoingo: Yo soy El senor Bingo de los Republica.
BingoBoingo: !!up mquander
deedbot: mquander voiced for 30 minutes.
BingoBoingo: freetlas: Quien es vos?
freetlas: I have not the least doubt it
BingoBoingo: Y mquander, Quien es vos?
freetlas: I am an individual who sometimes reads trilema
mircea_popescu: ello freetlas
freetlas: i mircea_popescu
mircea_popescu: no, i mircea_popescu ; you freetlas
freetlas: make sense
BingoBoingo: Te registra una llave con deedbot de el senor trinque
mircea_popescu: registra-te
mircea_popescu: imperative reverts
BingoBoingo: Ah, gracias
BingoBoingo: That would explain a bit
deedbot: << Contravex: A blog by Pete Dushenski - Bitcoin Year-In-Review 2017
freetlas: I wish I knew how Spanish grammar works
BingoBoingo: The only way to really learn is to move, live among the locals, then report back to one's betters
mod6: <+asciilifeform> apologies for the clutter, but this item really ought to live in the permarecord. << thanks for posting this.
freetlas: BingoBoingo: I do speak Spanish, but I do not know rules. Do you have an Uruguayan accent?
mircea_popescu: or rather an uruguashan accent.
BingoBoingo: freetlas: Not yet. But I am starting to get a hang of ll y Y sound like Shhhh
freetlas: Damn
mircea_popescu: it's funny how ll works as j[olly]j[oker] : it's sh in the south, moves to g in colombia, it's y normally...
freetlas: mircea_popescu: You have an Argentine accent, don't you?
BingoBoingo: When Colombians pass through the hostel though, very easy accent for my ears to process.
mircea_popescu: i deliberately avoided adapting to the idiots.
freetlas: mircea_popescu: be honest. DOn't be ashamed lol
mircea_popescu: normally i'm a total phonetic slut, but in this case i made an effort.
BingoBoingo: freetlas: mircea_popescu has a Roman accent
mircea_popescu: can't learn the accent of a despised herd of subhuman filth.
freetlas: BingoBoingo: It is. I mean, there are more than 4 different accents in Colombia.
BingoBoingo: freetlas: I man the accent on the columbians con dinera sufficient por vacacciones en Pocitos, Montevideo
BingoBoingo: *dinero
mircea_popescu: dinero lol. it's not like plata.
freetlas: gonna say the same lol
freetlas: BingoBoingo: that's where you currently living?
BingoBoingo: Correcting a/o is what the Venezualana dentist makes her side gig doing
BingoBoingo: freetlas: Living yes, But in a way that challenges what i considered living before. Much more living aca.
a111: Logged on 2017-12-21 13:30 BingoBoingo may have cheated a bit on the test leaning on the whole "animal magnetism" applied Trilema business.
trinque: << this does *not* explain what I put in the log, though it is also a problem.
a111: Logged on 2017-12-21 14:52 asciilifeform: << i dug into this item last yr, and found that the version thing is mostly a red herring : when you set 50400, the same derps drop ~you~, a millisecond or so sooner than you would've dropped ~them~ via malleus. the real culprit is always .
asciilifeform: trinque: hm plox to expand then ?
mircea_popescu: trinque hey, i'm all ears.
trinque: it is not as though I restarted the trb node once, said "oh ok, worx nao"
trinque: mats also reported to me that his nodes *will not keep up*, ben_vulpes also
asciilifeform: trinque: ( and this is not specific to this case ) plox to describe, in pedantic detail, the events
trinque: I already put it in the log, what
a111: Logged on 2017-12-21 05:23 trinque: for now, I can report that setting the version number to 50400, a trb node will catch back up, txns will unstick.
mircea_popescu: you got a blog, you saw teh fetlife/okcupid/whatever write-ups, knock yourself out in a mead of fermented detail.
BingoBoingo: Well, I just stayed a playful friendly extranjero and let the laughing and the context do the not heavy at all lifting
asciilifeform: but did not see , e.g., 'did 7 resets with 9.999... but only the 8th, with 50400, did the trick'
mircea_popescu: trinque the problem with "mats also reported to me that his nodes *will not keep up*, ben_vulpes also" is the ambiguity attendant. two guys walk into a diner, agree that coffee sucks there -- one's a colombian, local brew is swill ; the other's from alaska, can not stand hot beverages.
mircea_popescu: yet they "agreed". but what was meant ?
trinque: what was meant is that I want others to study their nodes while I study mine
trinque: and not simply go "alf says it isn't so, lets not look"
mircea_popescu: yes but you did not specific a study paradigm
mircea_popescu: at least not public
trinque: I plainly said what I was going to do about it
asciilifeform: one useful item would be a 'from whom this came' log prefixer patch
mircea_popescu: "study this room" rarely works in my experience (and yes, i experienced ; we have small toys for hiding etc)
mircea_popescu: trinque i gathered you moved from knob set to x to knob set to y ; that's not the same as a "do this experiment".
mircea_popescu: for instance, what specific variable was to be measured, and how.
asciilifeform: trinque: one way to turn this into an experimental setup would be to make ver knob adjustable ~without reset~
asciilifeform: then could genuinely say 'it wasn't the reset, it was version mask'
asciilifeform: after certain # of shots
mircea_popescu: reset does add some noise ; but it can be also filtered out otherwise.
trinque: asciilifeform: clearly.
mircea_popescu: statistics works on anything, is the thing.
asciilifeform: it isn't 'some noise', mircea_popescu , reset triggers the one and only sync-catchup routine
asciilifeform: ( see the massive log turd earlier )
mircea_popescu: i'm aware, i'm aware.
trinque: I was brief on details because I fully expected that others are seeing the same behavior with the trb nodes they maintain. public indexes of nodes report that just about every single one is far behind.
asciilifeform: trinque: the 1 common thread in asciilifeform's lab notes from birth of trb to today, is that once they fall behind, they stay behind, until reset. (often more than one reset.)
mats: i've had some luck syncing with 'btcwire' nodes, when all of the advertised nodes were ~30 blocks behind or unreachable on sunday
asciilifeform: and i'm satisfied that i found the reason for it. ( and no i do not have a ready pill, the ~algo~ is broken )
asciilifeform: what's btcwire ?
mats: and at that time i just fed those blocks back to the advertised nodes, it caught up
mats: some golang thing
mircea_popescu: asciilifeform half-admitted sybil.
mircea_popescu: was something-network before that
trinque: one which probably doesn't implement enough of the prb protocol to get banned
asciilifeform: mats: hand-feeding always works, yes. but it's not a practical means of keeping a node up in regular practice.
trinque: think for a sec, that with the high version number you're advertising support for protocol "features" you then immediately issue a ban for.
asciilifeform: trinque: this is not wrong , but it's a pov that concedes to the powerrangers the privilege of owning version#s
asciilifeform: why should they get to filter us out by doing e.g. 'nothing below 0.7 can connect'
mircea_popescu: asciilifeform in his case, because he wants to at least occasionally get txn out
trinque: this sort of thinking creates gigantic blind spots in which ways 0.5.4 is broken will hide.
asciilifeform: if it in fact has become impossible to get a tx out without emulating all of prb, we have a Problem
trinque: I reported this in the log
mircea_popescu: we have had A PROBLEM for longer than i can recall.
mircea_popescu: we have no isp. we can't as much as fucking hire a chinese girl.
mircea_popescu: somehow nike manages to get its shoes sewn ; but not us
mircea_popescu: we're afloat in problems.
mircea_popescu: so then.
mats: in regular practice i'm uncertain how to masquerade as prb without running a heathen node and plugging trb into that
trinque: in the case of the version number (which I will obviously test extensively, and said so) it's possible it's a bad move to crank to maxver, because it's a lie saying "we support all misc novelties below this int"
trinque: mats: my node *immediately* synced to top after weeks of restarts and praying to satan when version number was set.
asciilifeform: trinque: that knob is there, to be turned. i've no argument that 'version dun matter', it may well matter. but oughta be tested correctly, as indep. variable.
mats: trinque: who did you ask for blocks?
trinque: I have the republican list in addnode and that's it.
trinque: asciilifeform: where was that in argument
asciilifeform: iirc recent prb ( as analyzed by jurov, still digging for the thread ) will simply drop you if you connect and report <0.7 .
a111: Logged on 2017-12-21 05:34 trinque will be comparing multiple test subjects.
asciilifeform: you gotta compare without reset tho.
trinque: I agree with that as well.
trinque: should be trivial to add an rpc to change the var
asciilifeform: needs a patch.
asciilifeform: another patch experiment whose time imho has come, is 'prod' rpc command : trigger the boot sync behaviour at arbitrary time.
asciilifeform: i'ma produce it this weekend, if no one else has.
trinque: indeed, should be tested independently of this, can possibly preserve the high version.
asciilifeform: incidentally when i wrote the versionknob patch, i did not expect that everyone here will run with ~same~ , default ver
mircea_popescu: yes for srs. prod!
mircea_popescu: asciilifeform meaningless magic number got set to the "closest ot inexistence" available.
trinque: asciilifeform: I don't blame the patch, only "now I have an observation which suggests touch the knob"
asciilifeform: aha , ^ is how i came up with the value
trinque: suggestion may be wrong
asciilifeform: trinque: item very much deserves proper test.
asciilifeform: and as 'ship is made to sail', knob -- is made to turn.
asciilifeform: asciilifeform's point was not 'don't turn', but 'gotta turn them knobs independently of one another, or 9000 years will not suffice'
trinque suddenly realises that there are two intuitive items, tit.. and cock
trinque: the second, stick everywhere!
BingoBoingo: Where's that trilema about cock eating rock
shinohai: <<< Pero si es lleno de recursos humanos, no?
a111: Logged on 2017-12-21 14:37 BingoBoingo: Inglaterra es un pais pobre
trinque: BingoBoingo: how goes rackquest?
BingoBoingo: shinohai: Para hacer biopetro.
shinohai: Mah corn
BingoBoingo: trinque: More slowly than I would like. The whole US citizen bullshit is giving the accountant quite a workout when it comes to making banking work to pay for the racks.
mircea_popescu: in other sads : one trb node was dead since fucking 22nd of august, because -- it ran into the fabled "terminate called after throwing an instance of 'DbRunRecoveryException' what(): DbEnv::txn_checkpoint: DB_RUNRECOVERY: Fatal error, run database recovery" which then kicked in a script to clean it up, which it did, but couldn't boot back up because for yet-unknown reasons there was a spurious .lock leftover ; corner case u
mircea_popescu: ncovered, 1k man-hours worth of spuriously idiotic automation down teh drain.
mircea_popescu: we have, as the man says, A problem.
BingoBoingo: trinque: As far as the racks themselves. Ready within 72 hours of signing a contract and paying for them, but IP address block from LACNIC will take 2 weeks from application to allocation.
asciilifeform: BingoBoingo: what was yer bandwidth, again?
mircea_popescu: shaping up BingoBoingo !
BingoBoingo: asciilifeform: The quote is for 100 MBPS decicated symmetrical. If there's more traffic in one direction, asymetrical dedicated lines are also available.
BingoBoingo: mircea_popescu: gracias
mircea_popescu: asciilifeform he's getting his foot in tyhe door ; it'll scale.
asciilifeform: mircea_popescu: should hope so. because 100mb ain't enuff for even 1 trbtron
mircea_popescu: measure it soemtime.
asciilifeform: !#s incitatus
a111: 40 results for "incitatus",
asciilifeform: ^ subj.
trinque: BingoBoingo: would easily host my sites and botworx
asciilifeform: prolly would work for, e.g., qntra, aha
mircea_popescu: asciilifeform i don't see a number in there ; be moar useful and speak the value.
mircea_popescu: !~calc 1e6 * 3600 * 24 * 30
jhvh1: mircea_popescu: 1e6 * 3600 * 24 * 30 = 2592000000000
mircea_popescu: 2.5 petabytes is many times over trilema, so.
BingoBoingo can't recall running a trbtron I could touch on a connection faster than 3 mbps until in Montevideo
asciilifeform: hey, slow modem beats no modem
BingoBoingo: asciilifeform: Well, router rule
mircea_popescu: by the time your trb node eats 10 MBps even, you've more serious problems than "buy more pipe". and that's inst not sustained.
asciilifeform: BingoBoingo: well yes, i'm aware that there is not a literal 100mb/s modem, lol
BingoBoingo: Once trb gets enough network speed to work, extra doesn't offer much
asciilifeform: mircea_popescu: most painful spot is prolly the bringup of new node with its 200G
mircea_popescu: that 200G is over a little more than 20 minutes.
mircea_popescu: bb broadfly has it, there's a narrow sliver of bw trb/prb can even use ; more than that it generally wastes.
asciilifeform does the arithmetic, yea mircea_popescu is prolly right
mircea_popescu: and the narrow sliver in question is in the kbps range
mircea_popescu: ie, unless you got a good reason running a centralizing node, capping the bw for the damned thing doesn't do anything observable.
asciilifeform: the bottom limit is what it takes to move in a block every 10m . i.e. ~2kB/s.
mircea_popescu: in practice more like 15x that, because it gotta chatter like a boring housewife.
asciilifeform: aha, mempoolism
BingoBoingo: 1-1.5 mbps cap tends to allow resonable block reception time and support a number of connections, and keeps pipe faily clear outside of block time
BingoBoingo: with an under 512 kbps cap, not many other nodes will want to be friends
asciilifeform: and speaking of all of this, asciilifeform has a gentootronic box ready to crate up and enthrone immediately when BingoBoingo proclaims his rack.
asciilifeform: ( and soon 2 moar )
BingoBoingo: asciilifeform: very nice
asciilifeform: mircea_popescu: << i have seen this item a grand total of twice, and both times -- dying hdd
a111: Logged on 2017-12-21 16:44 mircea_popescu: in other sads : one trb node was dead since fucking 22nd of august, because -- it ran into the fabled "terminate called after throwing an instance of 'DbRunRecoveryException' what(): DbEnv::txn_checkpoint: DB_RUNRECOVERY: Fatal error, run database recovery" which then kicked in a script to clean it up, which it did, but couldn't boot back up because for yet-unknown reasons there was a spurious .lock leftover ; corner case u
mircea_popescu: i see it on occasion ; and yes decent predictor for ssd death.
asciilifeform: afaik it's entirely impossible without rotting iron
mircea_popescu: this is bitcoin, errything's possible.
asciilifeform: all i can say is, to date only observed db death on boxes with actual bit rot.
mircea_popescu: the db is actually fine.
asciilifeform: i.e. copied it elsewhere, and it mounted ?
mircea_popescu: here, also.
BingoBoingo: <asciilifeform> all i can say is, to date only observed db death on boxes with actual bit rot. << Aha, saw when trying to start trb from blockchain loaded onto box from corrupted thumb drive
mircea_popescu: lmao watching BingoBoingo spanish is amusing. los republica ? how many of it are there!
mircea_popescu: and why is it masculine allovasudden
mircea_popescu: teh republic's a bitch.
BingoBoingo: Because the venezualana is too busy to deliver suffient humilation on her own.
BingoBoingo: Anyways, as many republicitas as los senors.
BingoBoingo: But consider "vos sos [adjective]" mastered.
BingoBoingo: Anyways, accomodating the truth that most people want a friend more than I do, and letting them shoulder their share of the burden helps make BingoBoingo spanish functionally amusing in person.
BingoBoingo: Blog post baking.
BingoBoingo: !!up freetlas
deedbot: freetlas voiced for 30 minutes.
freetlas: Indeed, it is. Btw, Venezuelan* or venezolana/o
freetlas: In case you're talking about people from Venezuela
BingoBoingo: Una persona, una odontologa
freetlas: Conoces una odontologa venezolana?
BingoBoingo: freetlas: Para vos, Yo soy usted!
mircea_popescu: he got it, and he ain't affraid to use it!
freetlas: BingoBoingo: It ain't fucking usted in Argentine Spanish bastard lol
freetlas: , bastard*
BingoBoingo: Uruguay no es Argentina.
BingoBoingo: Y aca los senores son usted.
freetlas: You're right. Uruguay is part of Argentina. It's just another state. :)
freetlas: Los senadores?
BingoBoingo: Los caballeros.
freetlas: Are you a Lord? :o
freetlas: I apologize, Lord BingoBoingo. Usted conoce a una odontologa venezolana?
BingoBoingo: Gracias, si la conozco.
freetlas: Gracias. Si, la conozco.*
BingoBoingo: !!gettrust deedbot BingoBoingo
deedbot: L1: 1, L2: 8 by 8 connections.
mircea_popescu: anyway, continuing the trinque discussion, it seems entirely unavoidable that trb will become 3 things : a wallet node, optimized for pumping out local signed tx ; a block node, optimized for keeping the blockchain, getting blocks, no mempool nonsense ; and a spy node, optimized to keeping track of the lies and nonsense flowing through the relay network (mempool, timing nodes, what have you).
mircea_popescu: there's absolutely no conceivable reason to have all these 3 items in the same place.
mircea_popescu: key to the split is a rewrite of the db so it's less fucktarded.
asciilifeform actually has the db item, based on 'horsecocks' routine posted earlier ; but not ready for primetime and on hold pending ffa/p release
asciilifeform: i.e. giant mmap.
mircea_popescu: not a bad idea, at that.
asciilifeform: O(1) fetches.
asciilifeform: ( and writes, for that matter. )
asciilifeform: btw 'block node' per mircea_popescu's cut , doesn't have to be a 'daemonic' process at all, even. can run as 1shot cmdline util, that can be asked to eat a candidate-block ( and report success/failure ), or disgorge a previously-accepted block (by header , or by height) ; then terminates
asciilifeform: it dun need to persistently grind.
mircea_popescu: possibly.
mircea_popescu: could even be a bash one-liner
asciilifeform: bash-driven, at any rate. ( still pretty massive cpp ball behind the scenes, considering that all of the remaining 'grandfather's pistol' logic lives ultimately in the is-this-an-actual-block evaluator )
mircea_popescu: and evidently the power rangers made the wrong cut. it is FINE to have a "light node" consisting of block and tx ; no spy.
mircea_popescu: but they cut out the essential part and left the nonsense.
asciilifeform: without is-this-a-block-based-on-all-known-blocks-down-to-genesis, it ain't a node, it's used cumrag
mircea_popescu: so in this sense, it can be said "it makes no sense to have A without B", which is true. though it DOES make perfect senser to have B without A, which is why they are to be cut.
mircea_popescu: because exactly, B is the basis of all things ; upon which basis A or C or both or neither may live.
asciilifeform: will even observe, that it is possible to have a bitcoin net with only sneakernet connectivity, with cmdline eatblock-ing nodes.
asciilifeform: and hand-carried payloads.
mircea_popescu: cut like this, yes.
mircea_popescu: and as it should be.
mircea_popescu: then B can simply send PushGetBlocks until it falls over and entirely ignore mempool idiocy
mircea_popescu: thereby having nodes forever on the brink of the chain no matter what.
asciilifeform: B in the 'cmdline' model wouldn't ever on its own power send anything, or act autonomously
asciilifeform: it'd be 'supreme court' for the whatever As connected.
asciilifeform: ( but contain no networking code on its own )
mircea_popescu: could be a peer too, why not.
asciilifeform: idea is that it should not be gummable by anything that happens on the net.
asciilifeform: could be a peer, but 'lan peer'
mircea_popescu: i guess. better have it capable and let operator design than have it incapable and no choice.
mircea_popescu: lan-only knob eminently useful.
asciilifeform: ideally reference-trb is capable-of-everything, incl. cpumining ( see the dozen or so threads at this point )
mircea_popescu: no more "reference" bs. functionality.
a111: Logged on 2017-12-21 04:03 Cerber248: ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ christel sold freenode to Private Internet Access Andrew Lee WHO ALSO OWNS SNOONET AND IS MOVING FREENODE TO THAT SERVER (NEXT MONTH) AND CLOSING DOWN OPEN SOURCE ROOMS PLEASE COMPLAIN IN CHAN FREENODErlpxbbo: gribble lobbesbot de
asciilifeform: yes functionality but gotta be careful that the cut parts sum to at least the customary functionality.
mircea_popescu: you know FOR A FACT that it's nsa-bs, because if it weren't, they'd have sold it to me not to some obscure mortodifame.
asciilifeform: how else.
mircea_popescu: but it wasn't a "sale", it was a "mommy calls children home".
asciilifeform: they have a whole fleet of these shills, 1000s
mircea_popescu: NOW, the first, and naive, thought, would be "o noes, all those poor but intelligent engineer minds who will lose their freenode ; we should try to help them"
asciilifeform: starting with 1940s crypto-ag and going on to today.
mircea_popescu: no. we should not. fuck them with a barbed stick. how DARE they be anywhere but here ? and why help such scum ?
mircea_popescu: they'll get what they deserve.
mircea_popescu: reference ? FUCK YOU.
mircea_popescu: "reference" is altspeak for "proof proofiness investigation"
asciilifeform: it isn't as if the heathens read, or ever read, it
asciilifeform: !!up mircea_popescu
deedbot: mircea_popescu voiced for 30 minutes.
BingoBoingo: !!up mircea_popescu
deedbot: mircea_popescu voiced for 30 minutes.
mircea_popescu: there's 0 reason to think in these terms. let the self-important impotent judges-of-all-things-makers-of-nothing wash their heads with their reference rights.
mircea_popescu is in car, will v later on.
asciilifeform: i gotta ask, in what terms instead
asciilifeform: as i understood trb historically was 'and here is bitcoin minus the post-2011 barnacles, deviate at your own risk, derps, it will still be there an' working'
mircea_popescu: <mircea_popescu> no more "reference" bs. functionality. <
mircea_popescu: trinque needs so and so for so and so, let him have it.
asciilifeform doesn't disagree.
mircea_popescu: what we need is all that matters ; anyone else's "rights" are indistinguishable from hallucinations.
mircea_popescu: show a need and get at the least a go ahead, ideally some help.
asciilifeform: sorta where i was going with experiments such as 'wires' etc
asciilifeform: 'fuck the scum'
asciilifeform: ( btw did everyone grasp why 'wires' dun work ? )
mircea_popescu: i'm absolutely NOT going to sit here and say to trinque yes you're here and doing things but your item should be even an iota less usefull to you than it could be "because public"
mircea_popescu: the public is henceforth a byword for they in the dungeons.
asciilifeform: the 1 caveat is that if the tx dun cut through the kilometre of prb and get to china, it dun get mined
mircea_popescu: it's always cheaper to deal with china without idiot white cucks than with.
asciilifeform: hence 'cut through'
mircea_popescu: true in 1817 as today
mircea_popescu: anytway, this car has been idliong in parking lot long enough. i will bbl fo reals.
asciilifeform: i'll admit ftr that this item was one of the underlying seekrit crackpotteries behind asciilifeform's interest in shortwave
asciilifeform: theoretically a miner who gives a damn re max rake, oughta tune in when if e.g. mircea_popescu were to pump out a fat tx on sw.
asciilifeform: rather than waiting for 10,000km of and usg.fiber relays to churn.
asciilifeform: ( as well as dozen hops of prb nonsense )
asciilifeform: whatever else ionosphere does, it doesn't 'bitbetate'.
BingoBoingo: !~ticker --market all
jhvh1: BingoBoingo: Bitstamp BTCUSD last: 15450.0, vol: 22162.77034256 | Bitfinex BTCUSD last: 15429.0, vol: 68944.99232496 | Kraken BTCUSD last: 14891.3, vol: 4117.38791642 | Volume-weighted last average: 15410.6382406
asciilifeform: waterlolfall
shinohai: `Long Island Iced Tea shares soared 500% after the company rebranded as Long Blockchain. The decision came at 3 AM, during a long "product testing" meeting, after one executive mentioned that his son had made a lot of money on "Betcoin [sic].`
shinohai: toppest of keks
asciilifeform: i gotta wonder how mircea_popescu squares pictures like this with 'net worth is proxy for intelligence', 'executives are skilled professionals', and similar
trinque: << I have already split the wallet from trb in a patch I'm testing.
a111: Logged on 2017-12-21 17:38 mircea_popescu: anyway, continuing the trinque discussion, it seems entirely unavoidable that trb will become 3 things : a wallet node, optimized for pumping out local signed tx ; a block node, optimized for keeping the blockchain, getting blocks, no mempool nonsense ; and a spy node, optimized to keeping track of the lies and nonsense flowing through the relay network (mempool, timing nodes, what have you).
BingoBoingo: asciilifeform: Likely by recognizing the damaging effect of wreckers and the despicable kind of theif. Pic unrelated, features the honorable kind of thief
trinque: the cut turned out to be very clean, delete wallet.h/cpp, rip all references out, and item syncs just fine. it of course breaks getmempool and GenerateBlocks, whatever else.
asciilifeform: trinque: this is neato and i'd like to try it
asciilifeform: ( i assume it also sheds various mempoolisms ? )
trinque: nah, mempool cut can be a separate patch
trinque: kept this strictly to wallet
asciilifeform: a 'pistol only' cut , as discussed earlier, would be interesting
trinque: I will post an experimental patch soon, when fit for other hands
shinohai: Neato trinque .... had discussed with mod6 before how divorcing wallet rpc functions from trb would be so much better.
trinque: what's needed is, say, a shivatronic script compiler for txn.
a111: Logged on 2017-12-21 01:52 phf: i think correct method would really be to get the transaction out as a binary array into shiva, and then have a transaction parser in shiva itself that'll break it down into a sexp or whatever
trinque: that, but both ways.
trinque: asciilifeform: beauty of v is that these paths can diverge cleanly, operator can choose which lineage to press. patches appropriate for multiple lines, ground onto multiple lines.
phf: “Romanian Hackers Infiltrated 65% of DC Outdoor Surveillance Cameras (”
asciilifeform: lol, for spam
asciilifeform: where were these folx on '9/11' !111
phf: cheering on?
asciilifeform: ( famously, all cams near pentagon 'oops, stopped working' that day )
asciilifeform: << plaintxtization of the lulgem from phf's link
asciilifeform: !!up freetlas
deedbot: freetlas voiced for 30 minutes.
freetlas: I'm very disapointted with myself for not even being a Beginner of The Republic.
asciilifeform: so go and fix
asciilifeform: ^ reg with deedbot
asciilifeform: then follow the intro linked in chan subj, on mircea_popescu's www .
phf: << i've picked up a book in new orleans, “Symbolic computation : applications to scientific computing”, it's all macsyma and various sussman style scheme hacks for exploring dynamic systems, but an interesting trick that almost all the code uses, which i guess was sop in lisp world at some point, is to do visualization and control on a lisp machine, but do the heavy numeric simulations in fortran 77. the in
a111: Logged on 2017-12-20 18:35 asciilifeform: this is possible. when's the last time you saw a cpu-bound proggy in cl tho
phf: teractive environment is basically there to parametrize, and then it spits out a fortran code that crunches the numbers as it is
asciilifeform: oyea phf pretty sure i saw this
asciilifeform: 'compile to fortran'
asciilifeform: this was even used quite recently, in certain schemetrons
asciilifeform: 'compile to c'
asciilifeform: really the endgame of this is 'compile to fpga'
phf: like there's a paper on a multibody simulation system "implemented in zeta-lisp on symbolics 3600" with symbolic part done in macsyma, but the "numeric simulator" is implemented in fortran-77 running on a lisp machine
asciilifeform: for those annoting little cpubound problems like hbomb implosion contours etc
phf: there's a separate differential geometry package for non-linear dynamic systems in macsyma, likewise does analysis in macsyma, and then spits out fortran 77 code
asciilifeform: at one time asciilifeform was a meat version of this, for a living
phf: right, i guess you were doing industrial robotics and such?
asciilifeform: that was at same place, but later
asciilifeform: earlier was , them crunchings. inner loops in asm etc.
phf: ah
asciilifeform: btw, ftr asciilifeform considered f77 as close runner-up for 'in what to write ffa and all from nao on'
asciilifeform: it loses to ada in re: clarity and bound-guarantees
asciilifeform: but otherwise strong alternative to the bell labs idiocies.
asciilifeform: and quite well-standardized.
asciilifeform: and compiles pretty cleanly.
phf: i have to admit, my math is so weak, it's reading these papers i realized that Proper Engineering actually uses something called state-space representation for discretizing dynamic systems, and that behaviors of, say, robotic manipulators, have time independent, closed form solutions. also of explains that scene from anathem, at the workshop
asciilifeform extends imaginary cylinder hat.
asciilifeform: state-space representation << recall lagrange's mechanics.
phf: well, i have an intuitive understanding of what's going on now, i just need to work through calca. i've also simultaneously understood why lapack is so valuable. and yes, i see how that ties to lagrange's mechanics.
deedbot: << Bingo Blog - He Tenido Una Cita En La Ciudad Vieja: A Lesson In Applied Trilema
a111: Logged on 2017-07-18 18:42 mircea_popescu: rooster sees hen and runs towards her. hen reflexively starts running away. after a short space the hen thinks "if i stop he'll think me a slut, if i keep running he'll think me stupid... how about i stumble."
shinohai is thoroughly enjoying these new pieces on the BingoBoingo blog ....
BingoBoingo: !~tyvm
jhvh1: BingoBoingo: Error: "tyvm" is not a valid command.
jhvh1: You are very welcome Daddy
deedbot: << Ossasepia - EuCrypt: Correcting MPI Implementation
asciilifeform: o hey hey hey , lbj!!!
diana_coman: it seems I also found a difference in press behaviour between asciilifeform's v99 and mod6's v: this new vpatch of mine si correctly identified as leaf and otherwise descendant of ch1_mpi.vpatch by both v ; however, when pressing ch2 vpatch, mod6's v presses this other leaf too from what I see, while asciilifeform's v99 does not press it; to me v99's behaviour seems correct but I don't know if this is something that was discussed before
diana_coman: ch2 vpatch is *another* leaf, sibling of this vpatch
mod6: ugh.
mod6: i'll have to take a look a bit later here diana_coman
mod6: i simply curse the day that i ever wrote the thing
diana_coman: mod6, no rush at all; I just reported it because unexpected to me; let me know if there is any other info I can provide to help
diana_coman: fwiw mod6 I'm rather happy you wrote it though, as it helped me a lot in the beginning when I was trying to get my head around v itself; having at least 2 implementations to compare is not a bad thing
mod6: yeah, it's not good though. literally everytime someone uses it, there seems to be some sort of ... 'it doesn't work'
mod6: i'll take a look and see what your thing is doing later. will report back.
asciilifeform: diana_coman: the v99 behaviour is the correct one : pressing 'siblings' automagically, is a mistake; if you want the contents of both 'siblings', you oughta have a 'unifier' patch that pulls both in.
asciilifeform: v as i originally devised it, specifically 'grasps' by ~1~ leaf, when calculating a press
diana_coman: asciilifeform, yes, I plan precisely a unifier; that makes total sense to me, see above where I already said I consider v99 behaviour as correct
danielpbarron: !!deposit 1.5
phf: asciilifeform: you will find this lulzy, little factoid that i didn't know till now
phf: "Считаю необходимым довести до сведения следственных органов ряд новых фактов характеризующих мое морально бытовое разложение. Речь идет о моем давнем пороке — педерастии." "24 апреля 1939 г. Н. Ежов."
asciilifeform: pretty sure i've read this
asciilifeform: and pictured that it was written while sitting on a stake
phf: obviously, i didn't realize that the stake took this particular shape in this case
BingoBoingo: !~ticker --market all
jhvh1: BingoBoingo: Bitstamp BTCUSD last: 14690.0, vol: 21271.31875099 | Bitfinex BTCUSD last: 14687.0, vol: 68537.0174885 | Kraken BTCUSD last: 14472.0, vol: 4419.01464118 | Volume-weighted last average: 14677.5943
BingoBoingo: stakes have many shapes
esthlos: i'm trying to implement V, without reading too much source (as asciilifeform suggested), so how is this so far:
mod6: Initial findings after digging into it here quickly, my V seems to have the leaves in a separate flow order than from Stan's (v99). Which, by itself, (as we've discussed before) isn't a problem per-se. It seems though that in the case here where the leaves are ordered differently, and you press (with my v (99994)) up through 'ch2_truerandom.vpatch', it also includes the 'eucrypt_mpi_fix_copy_incr.vpatch' a
mod6: s it comes earlier in the flow.
mod6: So in this case, it is a problem, because you may not want to press 'eucrypt_mpi_fix_copy_incr.vpatch' leaf when pressing to only ch2.
esthlos: there is a collection X of diffs, a collection Y of detatched signatures of those diffs, and a collection Z of pubkeys corresponding to those patches
mod6: I don't have a fix ready or anything of the kind, in fact, this could be the nail in the coffin for my v.
mod6: I'll stop working on all other things, and dig into this. Will advise as I have more to say.
esthlos: by selecting which keys are acceptable, we get a subsequence of acceptable diffs, which are then used to patch the source
BingoBoingo: <esthlos> i'm trying to implement V, without reading too much source (as asciilifeform suggested), so how is this so far: << Well, stop reading code and take V on a walk down the Rambla.
asciilifeform: actually i am reading my orig V nao and shuddering
asciilifeform: the press algo i wrote was extremely lame , and mod6's vtron went astray by copying it, rather than the item in my head, lol
asciilifeform: the correct press algo is
asciilifeform: ... 1) to take the 'out' hashes of the ~one~, specific, patch being 'pressed to'
esthlos: that's where I'm at, trying to figure out the order to apply the diffs
asciilifeform: 2) recurse backwards from these.
asciilifeform: pointedly the 'flow' should not be used as the input for any operation , other than to display to user
asciilifeform: ( the linear flow display was also a mistake, it misleads people, does not actually help anybody. really ought to have exclusively tree-flow display, as mod6 had )
mod6: aha, yeah, this is where mine eats shit ^
esthlos: BingoBoingo: yes, I think it's time for that
BingoBoingo: esthlos: Just remeber, V is more interest in you than you in V
mod6: speaking of tree-display... once I spit out the tree for eucrypt, then it makes sense. no reason that you should have to press 'eucrypt_mpi_fix_copy_incr.vpatch' to have pressed 'ch2_truerandom.vpatch'
asciilifeform: so to expand : we take each 'out' , and follow it back as far as we can. ( when we meet a nil, we stop recursing, in that branch. ) when we meet a non-nil, recurse. when the recursion unwinds, the actual unixpatch (or phfian patch, for that matter) util invocations, happen.
phf: that's what btcbase does™
asciilifeform: phf did it correctly !
asciilifeform: by not copying asciilifeform's orig. hasty, wartime vtron.
asciilifeform: but instead implementing the Troo Algo.
phf: he get a lot of mileage out of a simple graph walk algorithm
asciilifeform: there's definitionally 1 correct graph traversal for any given vtree.
asciilifeform: per leaf.
asciilifeform: ( per node, really )
asciilifeform: you can press to any node.
mod6: Hmm. Ok, well, there *might be* a quick fix to this on my end -- not withstanding important original design flaws pointed out by asciilifeform above ^, seems that I have a possible idiotic thing in my toposort. Once removed, *appears* to do The Right Thing.
mod6: Will be looking at this closely/testing, etc. If that indeed becomes the fix, doesn't negate some of the refactors Stan is discussing. And those may be warranted in a later release.
mod6: Anyway, more as I have it. Thanks[
danielpbarron: !!up l0de
deedbot: l0de voiced for 30 minutes.
danielpbarron: !!up KutmasteR
deedbot: KutmasteR voiced for 30 minutes.
phf: "whatever happened to strong, silent types"
danielpbarron: hi KutmasteR, who are you?
KutmasteR: who are you?
danielpbarron: i'll give you one more chance before the down
KutmasteR: heard some guy talking about this chan
danielpbarron: yeah that was me
KutmasteR: well hello
danielpbarron: tell your boy l0de to get his key regged if he wants that free money
KutmasteR: i honestly don't know him that well
a111: Logged on 2017-12-18 01:18 phf: "The l0de radio hour is a long-running live call-in show that irregularly broadcasts over the internet, and to pirate shortwave and FM stations. It is hosted by Lode Ray Dio, a gypsy hustler with a long history of fraud, grifting, and treason. The show is notable for its frequent cancellations, obscene and bizarre content, and a high rate of listener suicide."
danielpbarron: yes, that's how i found it. it's live again tonight and i just called in
l0de: ya
danielpbarron: guy's apparently been doing a show for the last 20 years, and in the beginning did it from a pirate radio station
l0de: phones are acting up tonight
l0de: I'm wokring on it
danielpbarron: and for the record, (he claims) that he wasn't behind the spam. and has kline to back him up on that
phf: i thought the spam was funny, because i want to his youtube channel and there was a bunch of very upset and confused people there. i thought it was an intentional, as they say, good show
danielpbarron: !!up l0de
deedbot: l0de voiced for 30 minutes.
trinque: when do they start the treason part?
danielpbarron: idk, l0de picked my brain on how we're taking down the USG and i guess he didn't like my answer, but i told him i'd send him some bitcoin with the bot if he gets his key regged. he said he's never heard of GPG..
danielpbarron: !!up nisstyre
deedbot: nisstyre voiced for 30 minutes.
phf: if only there was some way to capture a conversation so that it gets preserved for future readers
danielpbarron: !!up MTW
deedbot: MTW voiced for 30 minutes.
MTW: hell yeah
mod6: werd
trinque: what's an MTW ?
danielpbarron: he's with l0de
trinque: lol, this guy is bitching about money
trinque: how much coin did you offer him, danielpbarron ?
danielpbarron: 800 USD worth
danielpbarron: !!up rainrawr
deedbot: rainrawr voiced for 30 minutes.
deedbot: 5048EC06FA366D1350F5DBDD55A423FD89074FC4 registered as MTW.
danielpbarron: !!up weevlos
deedbot: weevlos voiced for 30 minutes.
danielpbarron: !!rate MTW 1 l0de radio guy
weevlos: !!help
rainrawr: ok wtf do i do
a111: Logged on 2015-01-31 01:28 assbot: Successfully updated the rating for danielpbarron from 1 to 2 with note: He's sort of like an apostle.
deedbot: B9185B95152960A319100120740941791782B3CD registered as weevlos.
danielpbarron: alright MTW you can voice yourself with !!up
trinque: rainrawr: announce yourself. who are you? what does your daddy do?
MTW: hell yeah
danielpbarron: lemme send you some coin here
weevlos: hook me up with btc dawg
danielpbarron: !!pay MTW 0.01
rainrawr: hi im rainrawr im just here to chat
MTW: !!balance
danielpbarron: give it a sec
MTW: kk
danielpbarron: !!rate weevlos the weev, friend of Emily Youcis
danielpbarron: did we crash the fucking bot?
MTW: looks like it
MTW: !!help
MTW: no it's slightly aliv
MTW: e
trinque: danielpbarron: you didn't put a number in your command...
danielpbarron: oh right
danielpbarron: but my verify for the pay command didn't return yet
danielpbarron: !!rate weevlos 1 the weev, friend of Emily Youcis
trinque: and one sec on the wallet ops.
weevlos: !!pay weevlos 0.1
deedbot: You must be registered.
MTW: !!balance
trinque: danielpbarron: should be fine now
MTW: danielpbarron that bot work or nah
trinque: MTW: I have to make the wallets.
MTW: ahhhh
danielpbarron: !!v 6DF767B9CA157362E7B69C67342313A0AA02329F917870BA3965BBAFC1027200
trinque: weevlos: anyhow you're the stormer guy eh? there was a thread
a111: Logged on 2017-09-01 06:14 mircea_popescu: AlfredAlfer the correct solution here is to pay 1-200 per month for a dedicated server in a sane jurisdiction (ie, outside of nato). make sure you get a chunk of ips with it, then host all your shit.
weevlos: yeah we have no problem with server hosting
weevlos: our servers stay up just fine
weevlos: the issue is the registrars
weevlos: everybody thinks we don't have servers
rainrawr: danielpbarron: how do i register
trinque: weevlos: thread was more that DNS is and will always be in the hands of the enemy.
weevlos: yeah
← 2017-12-20 | 2017-12-22 →