log☇︎
27000+ entries in 0.015s
girlattorney: if you have no wot, and you are on your own, and you rely on the goodwill of the republican nodes, currently you wait 28 nodes
asciilifeform: ( often -- 0, people tend to stand'em up via eatblock )
asciilifeform: the other side of the medal, is that a trb node spends only small portion of its life in initial sync.
girlattorney: as already told: i appreciate that TRB exist even if i still not able to using it. However reading logs when syncing has been a pain and i thought that mempool during syncing could create just overhead
asciilifeform: the 'reject peers sending garbage' mechanism is also mine. trb did not always have it. i submitted it for inclusion to mod6's flagship after determining that it results in substantially improved performance across the board (i.e. peers sending bloomism are, statistically, an unlikely source of the-next-block) ☝︎
asciilifeform: girlattorney: there is not currently such a knob. no one has felt pressing desire for it; if you wanna submit patch -- folx will read.
girlattorney: could be written on trb a feature that enable / disable mempool?
asciilifeform: ( not to mention, even when relaying actual bitcoin blocks & tx, blows gigatonnes of bandwidth on nonsense inserted by wreckers as prelude to their bigblockism scheme of '15 , e.g. bloomism )
asciilifeform: and nonsensical tx.
a111: Logged on 2018-10-30 18:16 asciilifeform: more interestingly, there was even 1 of 10/30/18 17:05:41 ERROR: ProcessBlock() : CheckBlock FAILED from peer 213.148.193.153
asciilifeform: i personally removed this nonsense, as 1 of the opening shots of trb story.
girlattorney: asciilifeform that's nice to hear, so it has been kept simple for the sake of security, correct?
asciilifeform: prb had 'orphanage' mechanism where it accepted antecedent-less inputs 'on faith'. this opens node both to memory exhaustion and algorithmic complexity attack (i.e. crafted input can prompt machine into wasting arbitrary amt of memory, + arbitrary amt of cpu cycle walking it)
asciilifeform: this includes e.g. a block whose parent was not yet seen; a tx whose antecedent input(s) not yet seen .
asciilifeform: girlattorney: 'discard excess' in trb algo is ANYTHING received when its antecedent is absent.
girlattorney: asciilifeform i'm aware of the fast sync part. Just saying that when i sync a node i actually get data from other prb nodes. I'm not getting garbage or errors, and I think (not being technically able unfortunately) that you could modify prb to verify all the blocks, and eventually discard the excess if I reecived them in sprayed order
a111: Logged on 2019-07-16 12:52 asciilifeform: trb, unlike prb, does not accept blocks 'on credit' (i.e. ones for whom the antecedent block is not yet on the disk)
asciilifeform: girlattorney: see also earlier thrd, and past . ☝︎
BingoBoingo: <girlattorney> but i could argue with this: when i do bootstrap a core node, it actually get fed from other core nodes << During initial sync trb wants to receive blocks in order while core will intentionally spray blocks out of order
girlattorney: so i can effectively say that i'm receiving blocks from other nodes, they aren't just connecting and stay silent. They are sending me blocks.
asciilifeform: girlattorney: 'core' (aka prb) when bootstrapping, asks randomly selected peer for 'headers' and eats'em on faith. afterwards does same with blocks (none of which it bothers to actually verify, in the traditional sense)
BingoBoingo: asciilifeform: I notice when I drop the version number down enough I'll get a wider number of version strings in my peerslist
girlattorney: but i could argue with this: when i do bootstrap a core node, it actually get fed from other core nodes
asciilifeform: BingoBoingo: fwiw none of the nodes i've operated were ever set to masquerade as prb (other than in the 0.9999... aspect)
asciilifeform: answr to given puzzle is very simple -- most of the supposed btc net, consists not of nodes, but of equivalent to sand in engine's crankcase.
BingoBoingo: Playing with the version string on a TRB node is the fastest and simplest way to change the sorts of peers your node encounters in the wild
asciilifeform: girlattorney: simple logical inference (there's a number of publicly advertised trb nodes; most of'em agree with heathendom re the height) points to : no, not 'only with themselves' dunnit.
girlattorney: i was asking about where a TRB node fetch the blocks, if all of the TRB nodes are only interconnected with themselves
a111: Logged on 2019-07-16 11:19 mp_en_viaje: girlattorney, your notion of identity is not adequate for the situation you're dealing with.
girlattorney: mp_en_viaje i'm re-reading the messages, what do you mean by saying this? http://btcbase.org/log/2019-07-16#1922517 ☝︎
feedbot: http://trilema.com/2019/thelastpsychiatristcom-academics-hide-drug-company-payments-adnotated/ << Trilema -- thel....com - Academics Hide Drug Company Payments. Adnotated.
asciilifeform: girlattorney: that's half the story, yes. other half it that it's ~100MB of unauditable obfuscated-c liquishit.
girlattorney: asciilifeform i know that it's centralized. A couple of CA that runs the game
BingoBoingo: girlattorney: It's been a while since I looked into it, but I believe if the version string on a peer is greater than X, they insist on SSL'ing
asciilifeform: girlattorney: are you familiar with the actual purpose of sslism ?
girlattorney: BingoBoingo can you rephrase about tunneling? Do you mean they use ssl to connect?
asciilifeform: BingoBoingo: lol, hearn's old thing , the cocktail where tried to slip heartbleed in
BingoBoingo: Nevermind the things nodes pass to other nodes are not the seekrit things
BingoBoingo: Also recent NSAware nodes insist on encrypted tunneling to connect to other nodes
asciilifeform: girlattorney: will still be banned by any working trb node, on acct of sending nonclassical protocol cmds (bloom filter garbage etc)
girlattorney: you still are on a core node, but even if you aren't completely ignoring segwit shit, you aren't touching it directly
girlattorney: but let's say the following: you get a core (prb) node, you set a minrelaytxfee very high, so you don't propagate at all, just include what new blocks have because you have to (to stay up to date)
BingoBoingo: "To the network, PseudoNode behaves the same way as a full node by relaying transactions, blocks, addresses, etc. However, unlike a normal full node, PseudoNode does not verify data (txs & blocks) itself. Rather, PseudoNode relies on neighboring peers (with configurable confidence levels) to do the verification on PseudoNode's behalf. As a result, PseudoNode is very lightweight."
asciilifeform: ( other half of problem is -- miners mining w/out actual working node. but this is elaborated in the logs, will not repeat )
BingoBoingo: At present its too easy for Bezos customers to spin up things that superficially act almost like nodes.
asciilifeform: and 99,999...% of their cpu cycles spent rejecting rubbish from the 9000 'nodes' running nsaware.
asciilifeform: girlattorney: problem, as such, is that right now there are maybe two dozen btc nodes (actual, working ones, w/out segshitness etc) and most of'em belong to the folx here.
girlattorney: Also, what's the problem in the nodes not being rewarded from being just nodes? If I want to run a serious business I'll need a node, otherwise i can stick with a third party wallet such as primedice or deedbot
asciilifeform: girlattorney: older article re subj, with illustration of the business end of the punishment weapon.
mp_en_viaje: they do what they're told.
mp_en_viaje: girlattorney, nobody asked them anything ; not in 2013, not in 2016, not now and not for the foreseeavble future.
girlattorney: First of all, do you really expect millionaire businesses such as bitmain or bitfury to support a fork that instantly invalidates many millions of already-in-production ASICs?
BingoBoingo: girlattorney: You aren't missing much. The protocol isn't changing any time soon for reasons largely explained by handicaps inflicted by Mandarin language's tendency to try insulating against the future.
mp_en_viaje: in other lulz, anyone recall the 1918 narkomprod trying to "organise the peasants" into an orange revolution, only to discover that, contrary to marxism-leninism, poor peasants do not actually hate rich peasants ; not to mention have exactly the same interests.
girlattorney: but i see the article is from 2016
girlattorney: From the article i read "In any case, a word to the wise : if you are designing ASIC chips, and you are not including the possibility of feeding a bitfield like this in blocks, you are deliberately ensuring failure not just for yourself, but for your customers as well. This change WILL eventually come in, start planning accordingly, today. ["
mp_en_viaje: asciilifeform, can has link to your original lohatronarticle ?
girlattorney: and is there a consensus about how much it should be raised?
girlattorney: another question that i think belong here. Tried to read the logs but haven't found enough clarity. When Bitcoin should scale the blocksize?
feedbot: http://qntra.net/2019/07/iran-detains-tanker-for-smuggling-fuel-in-persian-gulf/ << Qntra -- Iran Detains Tanker For Smuggling Fuel In Persian Gulf
feedbot: http://trilema.com/2019/apocatastasis/ << Trilema -- Apocatastasis
asciilifeform: thread moved to #a .
girlattorney: i think that the large number of writes could be a problem in the long term
asciilifeform: threatens to be long and uninteresting to erryone else.
asciilifeform: mp_en_viaje: lemme know if you want this thread moved to #a
girlattorney: i think that half MB won't be enough
asciilifeform: i wanted to see when was last time you got block, and from where
girlattorney: just like the one that i posted repeated till half MB
asciilifeform: girlattorney: post plz your last half MB or so of trb log
girlattorney: i'm very near to setup my blog to post screenshot and other stuff about the setup
girlattorney: i disabled the swap by not having it
girlattorney: i determine the writes using iotop
asciilifeform: ditch the virtualhosting thing.
asciilifeform: girlattorney: if this were in fact the case, my ssd would live for a week and not 2yrs.
girlattorney: bdb sits the index on disk, that's ok, but i got the impression that to archive 1MB it needs to write like 1GB
asciilifeform: re writes, bdb sits the index on disk, so naturally writes
asciilifeform: by how many behind ? and from where do you know the current height ?
girlattorney: and the amount of writes that TRB does with or without swap is insane
girlattorney: yes, still don't be able to catch the latest blocks
girlattorney: well, mempool rejected transactions etc...
asciilifeform: girlattorney: post the garbage ?
girlattorney: and even if i was stuck a dozen of blocks behind the general block height, the connection with the other peers was still present, just producing garbage in debug.log
girlattorney: every pal listed on thebitcoin.foundation
girlattorney: It's almost like TRB telling me: you're too poor to spend your time trying running me, do something different in your time
girlattorney: so, tried with many attempts to restart TRB and hoping it could fully sync. No fucking way. And at every shutdown (with ctrl - c) always a painful writing process of at least 50 gb, taking 30 minutes or so
feedbot: http://bvt-trace.net/2019/07/unrolled-x86_64-assembly-multiplication-for-updated-to-ch-14-ffa/ << bvt's backtrace -- Unrolled x86_64 Assembly Multiplication for updated to Ch. 14 FFA ☟︎
feedbot: http://trilema.com/2019/thelastpsychiatristcom-abusive-teens-force-their-girlfriends-to-get-pregnant-dont-let-the-truth-get-in-the-way-of-a-good-story-adnotated/ << Trilema -- thel....com - Abusive Teens Force Their Girlfriends To Get Pregnant! (Don't Let The Truth Get In The Way Of A Good Story). Adnotated.
feedbot: http://trilema.com/2019/thelastpsychiatristcom-8-characteristics-of-family-annihilators-adnotated/ << Trilema -- thel....com - 8 Characteristics of Family Annihilators. Adnotated.
asciilifeform: at some pt somebody's gonna have to audit a snapshot of bin gnat. which wouldja rather ? ☟︎
asciilifeform: or the nightmare of endianized byte-addressing permitted in ~every~ inst that refs mem. ( again x86. )
asciilifeform: or what even means to audit a binturd in arch where jumping into middle of instruction is permitted.
asciilifeform: ftr i want to see x86 and arm die , properly, and cremated, they belong in same place as winblows. srsly wtf, picture 'alphabet' of 700 symbols, that's x86..
a111: Logged on 2019-07-18 04:09 mp_en_viaje: http://btcbase.org/log/2019-07-17#1923138 << yep, it's what's inhibiting my otherwise natural and arguably required "holy shit wtf is this weird, stop now" reaction.
asciilifeform: http://btcbase.org/log/2019-07-18#1923194 << if mp_en_viaje or anyone else can think of less ugly pill for described problem, i'm all ears, cuz indeed this one's barbaric ☝︎
mp_en_viaje: much like viktor vs anatoli situation, come to think of it.
a111: Logged on 2019-07-17 16:42 asciilifeform: e.g. old man naggum -- evidently moar 'alive' nao than when alive..
mp_en_viaje: http://btcbase.org/log/2019-07-17#1923146 << sad fact of the matter, ballas has significant advantage over naggum on purely procedural grounds -- had enough fucking sense to make himself blog. ☝︎
a111: Logged on 2019-07-17 16:38 asciilifeform: http://btcbase.org/log/2019-07-17#1923088 << the circular situation of gnat , where 'need gnat to build gnat', is the third reason for the mips approach -- arch with 100 fixed-length instructions, simplify the eventual rewriting of backend so can ditch gcc dep.
mp_en_viaje: http://btcbase.org/log/2019-07-17#1923138 << yep, it's what's inhibiting my otherwise natural and arguably required "holy shit wtf is this weird, stop now" reaction. ☝︎☟︎☟︎
mp_en_viaje: but i'll see your modern africa and trade you historical america