Show Idle (>14 d.) Chans


← 2021-09-12 | 2021-09-14 →
00:07 verisimilitude Another aspect is treating others as nonhuman until proven otherwise.
00:08 verisimilitude I've a website, with real work, and original work at that; ergo, I'm human, more human than one who builds nothing.
00:20 verisimilitude Storing encrypted messeges in the deduplication buffer is an obvious optimization; I've been thinking about per-peer deduplication buffers, but there doesn't seem to be justification for them.
~ 16 minutes ~
00:36 thimbronion asciilifeform: encryption question: in updating to the new packet structure, I'm finding that a requirement of the encryption lib is that the packet must be divisible by 16, as 16 bytes is the block size. According to the spec, red packets are 424 bytes, which is not divisible by 16, causing the serpent lib to barf. Is the algo broken or am I missing a step maybe?
00:37 verisimilitude Padding would be the answer, right?
00:37 thimbronion I'm seeing one other anomaly while packing the red packet, but need to investigate further before I can ask anything.
00:38 thimbronion verisimilitude: yes, but then the packet wouldn't be the size specified in the spec
~ 1 hours 31 minutes ~
02:09 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057805 << it's no optimization of any kind, if yer storing unauthenticated and undecrypted (i.e. can't see the timestamp yet) enemy can fill you to the ears with replayed material
02:09 dulapbot Logged on 2021-09-12 20:20:15 verisimilitude: Storing encrypted messeges in the deduplication buffer is an obvious optimization; I've been thinking about per-peer deduplication buffers, but there doesn't seem to be justification for them.
02:09 dulapbot Logged on 2021-09-10 14:26:11 asciilifeform: punkman: if you're storing items you have not authenticated -- enemy can fill you
02:10 asciilifeform must. not. store. unauthed. packets. not hard concept imho.
02:10 dulapbot Logged on 2021-09-10 14:28:38 asciilifeform: but not if you can be induced to ~store~ garbage.
02:11 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057806 << cipherblock chaining. i suppose i oughta specify it explicitly. (is why the nonce is there.)
02:11 dulapbot Logged on 2021-09-12 20:36:51 thimbronion: asciilifeform: encryption question: in updating to the new packet structure, I'm finding that a requirement of the encryption lib is that the packet must be divisible by 16, as 16 bytes is the block size. According to the spec, red packets are 424 bytes, which is not divisible by 16, causing the serpent lib to barf. Is the algo broken or am I missing a step maybe?
02:11 asciilifeform ( and you pad it. )
02:12 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057809 << black packet's ciphertext is 444 byte, recall
02:12 dulapbot Logged on 2021-09-12 20:38:00 thimbronion: verisimilitude: yes, but then the packet wouldn't be the size specified in the spec
02:14 asciilifeform ( the ~message~ is 424 bytes long, but recall that there is also nonce/bounces/version/reserved/command. )
02:15 * asciilifeform was planning to put worked examples of encrypt/decrypt in 5.2.
02:19 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057793 << soo yer not only asking to double (triple? ntuple? in general case) the req'd amt of bandwidth taken up, but also asking receiver to burn a buncha cpu cycles finding the correct alignment?! all to avoid using checksums?
02:19 dulapbot Logged on 2021-09-12 19:50:15 verisimilitude: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-10#1057182 The idea is it uses multiple copies in a way which prevents tampering from reasonably happening, without a hash checksum or whatnot.
02:24 asciilifeform it's arguably the 1 application where hashes can be absolutely relied on.
02:25 asciilifeform (it's a fucking otp soup, there's 0 possibility of constructive manipulation of either body or hash forfuckssake)
02:26 asciilifeform ^ this in re [http://verisimilitudes.net/2021-09-09][[verisimilitude's article], nought to do w/ pest
02:26 asciilifeform grr
02:26 asciilifeform ^ this in re verisimilitude's article, nought to do w/ pest
~ 1 hours 14 minutes ~
03:40 vex yessir oughtta work
03:40 dulapbot Logged on 2021-09-12 19:32:10 asciilifeform: so far the only solution asciilifeform can think of, is 'pest is not for casualtards who don't have 24/7 fiber. if you don't have at home, get a box in a dc.'
03:45 vex If you can't afford shoelaces, you can't afford shoes
~ 18 minutes ~
04:04 vex comsci (crypto) lectures are the best. thanks professor alf
04:18 vex https://www.youtube.com/watch?v=FJC9mrX4y-Q
~ 5 hours 2 minutes ~
09:21 punkman http://logs.nosuchlabs.com/log/asciilifeform/2021-09-11#1057565 << there is anothe way to look at this. if you run a business, or a cult, you need a "funnel" where prospects fall in. maybe it has a wide "mouth", or not so wide, but you'll have to interact with prospects and reject/train/sell them
09:21 dulapbot Logged on 2021-09-11 23:18:45 billymg: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-11#1057553 << which is why it's crucial to not compromise anything for the normie (it sounds like we're all in agreement on this tbh)
09:23 punkman if your funnel is too narrow, if you don't work the funnel, you end up a lonely business/cult-man
09:29 punkman http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057831 << suppose I don't have a toilet-box where to run experimental network software. now I need to find a provider, buy vps, set it up, manage the thing. The vps costs $5 month, whatever no big deal, but it also costs time and space in head.
09:29 dulapbot Logged on 2021-09-12 23:45:20 vex: If you can't afford shoelaces, you can't afford shoes
~ 4 hours 29 minutes ~
13:58 PeterL What is the downside to having an intermitent connection? Is there a reason everybody needs to have a complete history?
~ 42 minutes ~
14:40 asciilifeform PeterL: see thread. (summary: 'selfchain' feature for distinguishing 'true' nick from a squatter relies on being able to see an unbroken chain of 'i'm the one who previously said x')
14:40 dulapbot Logged on 2021-09-12 19:21:26 asciilifeform: thinking again about 'getdata' : this won't work as given. because if timestamp indicates 'stale' (>15min in the past) the answer to the 'getdata' will be tossed.
14:41 asciilifeform (this in re: hearsay -- messages directly received from a peer, with his handle in'em, are prima facie authentic)
14:44 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057837 << you don't strictly speaking ~need~ a proper box; point was that certain imho key features of the protocol won't work as intended, fork alarm will go off regularly on your station, and you'll end up silencing it, losing the mechanism effectively, and eventually suffering the problem it is
14:44 dulapbot Logged on 2021-09-13 05:29:27 punkman: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-12#1057831 << suppose I don't have a toilet-box where to run experimental network software. now I need to find a provider, buy vps, set it up, manage the thing. The vps costs $5 month, whatever no big deal, but it also costs time and space in head.
14:44 dulapbot Logged on 2021-09-06 00:30:43 asciilifeform: they're there to 1) solve the 'cocks problem' 2) support, in future, usenet-style threading.
14:44 asciilifeform meant to defend against.
14:45 asciilifeform on top of this, people will send you PMs which go to /dev/null with no indication or warning
14:45 asciilifeform and you, similarly, to others
14:47 PeterL I suppose your ack method guards against PM's silently going to /dev/null
14:48 asciilifeform PeterL: ACKs are going away.
14:48 asciilifeform they're a massive waste of bw and entirely redundant when we have 'getdata'
14:48 asciilifeform punkman: without reasonably-stable connection, you will not reliably track unbroken selfchains. and so will have exactly 0 ability to distinguish genuine from fraudulent hearsay msgs.
14:49 asciilifeform if anyone can think of a 'have the cake and eat it' solution (aside from the unavailable, at line rate, constant-time rsa) i'm all ears.
14:49 PeterL nothing stops people from saying "got it" after getting a message, but that does not need to go into the protocol
14:49 asciilifeform PeterL: indeed
14:50 asciilifeform you need either reliable delivery, or gap filling, for maintaining selfchain integrity tho
14:50 asciilifeform the latter imho is preferable (getdata).
14:52 PeterL if I know I have been offline for a while, then I can assume I will have missed some selfchain from anybody who talked, so should there be a mechanism for ignoring the forking at the inial connection?
14:53 asciilifeform PeterL: right, but you'll ignore it enuff of the time that the feature may as well not exist for you.
14:53 dulapbot Logged on 2021-09-13 10:44:24 asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057837 << you don't strictly speaking ~need~ a proper box; point was that certain imho key features of the protocol won't work as intended, fork alarm will go off regularly on your station, and you'll end up silencing it, losing the mechanism effectively, and eventually suffering the problem it is
14:54 PeterL and if I have been offline then I would not have been talking, so it should not affect other people's expected selfchain from me
14:55 asciilifeform asciilifeform's aim was to come up with some means to enforce nick exclusivity on a 100% p2p net. this may be 'fried ice' and impossible in the general case. but i certainly have no desire to live on a net where there is NO ability to distinguish an original from a squatter.
14:55 asciilifeform PeterL: you can talk without knowing that you're offline.
14:55 asciilifeform bang, you've broken yer selfchain ~for everyone~.
14:55 PeterL heh, I guess that is true
14:57 PeterL so then when I reconnect everybody would use "getdata" to see what they missed. Would they then re-broadcast the old messages?
14:57 asciilifeform PeterL: understand, we have what genrich altshuller referred to as an 'engineering contradiction'. where e.g. you need a stick to be both long and short somehow at the same time; a pan to be both hot and cold; etc. superficially can't be done -- but sometimes MUST.
14:57 dulapbot (trilema) 2019-05-31 asciilifeform: not read whole thing, but appears to be inspired by altshuller's 'triz' , which in fact not a crock of shit, and asciilifeform personally uses.
14:58 asciilifeform in this case we need rsa. but we can't use rsa.
14:58 asciilifeform the entire pest algo is an attempt.
14:58 asciilifeform (at 'resolving the contradiction'.)
14:58 * asciilifeform will be back shortly
~ 22 minutes ~
15:21 billymg asciilifeform: your www's select mechanism has a bug where it also strips off the post id upon deselect
15:22 billymg at least on this chromium version i'm currently using
~ 18 minutes ~
15:40 asciilifeform billymg: well-known bug
15:40 asciilifeform billymg: if you've a fix, plox to write in
~ 16 minutes ~
15:57 billymg asciilifeform: this oughta do it
~ 41 minutes ~
16:38 asciilifeform billymg: seems to work!! ty!
16:44 billymg asciilifeform: np. it felt like your site was gaslighting me, i kept wondering if perhaps at some point i had hit the back button and didn't remember
16:58 asciilifeform billymg: ancient bug, inherited from the original (from archive.is, via mp's rip)
17:11 asciilifeform continuing this puzzler-thread : 'impostor' and 'original' are not always even applicable :
17:11 dulapbot Logged on 2021-09-13 10:40:59 asciilifeform: PeterL: see thread. (summary: 'selfchain' feature for distinguishing 'true' nick from a squatter relies on being able to see an unbroken chain of 'i'm the one who previously said x')
17:12 asciilifeform ... given that two previously-disjoint nets may merge , there can be nick collisions resulting from said merge w/out necessarily any malfeasance on anyone's part
17:19 asciilifeform ( a spec-compliant noad will not permit two+ direct peers having colliding canonical (i.e. recorded in WOT) handles. but is powerless to prevent the use of whatever handles, incl. colliding, in hearsay msgs )
~ 18 minutes ~
17:37 * asciilifeform firmly stuck on this issue, cannot proceed w/ this work until satisfactory solution is found
17:41 asciilifeform punkman signpost billymg thimbronion et al ^
17:53 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057866 << forgot to answer this -- doesn't work in the >15min case, recall.
17:53 dulapbot Logged on 2021-09-13 10:57:33 PeterL: so then when I reconnect everybody would use "getdata" to see what they missed. Would they then re-broadcast the old messages?
17:53 dulapbot Logged on 2021-09-12 19:21:26 asciilifeform: thinking again about 'getdata' : this won't work as given. because if timestamp indicates 'stale' (>15min in the past) the answer to the 'getdata' will be tossed.
~ 20 minutes ~
18:14 punkman asciilifeform: sort of feels like inverse of "tin woman" probem, where we are so poor, we are trying to figure out optimal way to fuck can of beans
18:14 asciilifeform punkman: imho virtually all real-world engineering falls into this category
18:15 asciilifeform in practice the avail. materiel inevitably much closer to can of beans than the desired antigrav unobtainium.
18:15 asciilifeform for instance, satellite biz still relies on strapping the 100-200M$ hand-made machine to a gigantic bomb...
18:19 asciilifeform so yes indeed, asciilifeform is searching for 'optimal way to fuck a can of beans' and invites others to participate.
18:19 asciilifeform cuz the alternative is, what, vodka and death.
18:19 punkman my point being, if we are "fucking can of beans", maybe not to worry too much about selfchains and forkage at this point
18:20 asciilifeform punkman: i'm uninterested in baking another irc.
18:20 asciilifeform 100% uninterested.
18:20 asciilifeform 100% ddos and sybilproof or fuckyou.
18:20 punkman we can only verify msg from direct peer, there is no magic way to verify hearsay for free
18:20 asciilifeform at the same time gotta.
18:20 asciilifeform otherwise, punkman , we have the star topology again
18:21 asciilifeform i.e. zero useful communication except between direct peers.
18:22 asciilifeform such a net would have precious little redundancy/resilience, and, importantly, almost impossible to grow to any interesting size without exponential explosion of overhead
18:22 punkman but we've already established "if peer sends shit hearsay messages, unpeer/gag/whatever"
18:22 asciilifeform punkman: how do you know they're shit ?
18:23 punkman mostly own eyes
18:23 asciilifeform aha, and if not obvious to eyes ?
18:24 asciilifeform and, say, this case. what, 'it won't happen' ?? and if it does ?
18:24 dulapbot Logged on 2021-09-13 13:12:22 asciilifeform: ... given that two previously-disjoint nets may merge , there can be nick collisions resulting from said merge w/out necessarily any malfeasance on anyone's part
18:24 asciilifeform prayer is not a valid engineering component.
18:25 punkman how would nets join?
18:25 asciilifeform punkman: say asciilifeform has peers a,b,c. punkman has peers x,y,z. asciilifeform and punkman became peers.
18:26 asciilifeform c and z are both 'nebuchadnezzar'.
18:26 asciilifeform now we have a hot mess.
18:26 punkman why would we do this?
18:26 asciilifeform what do you mean 'why' ?
18:26 asciilifeform you peered with someone. he had existing peers.
18:26 punkman if I join separate net, I don't bring peers with me
18:26 asciilifeform they are now your l2.
18:27 asciilifeform sure as fuck you do.
18:27 asciilifeform via hearsay.
18:27 asciilifeform it aint separate, is the thing
18:27 asciilifeform you've added a peer to a net of yours. he had existing peers of his own.
18:27 asciilifeform ( you can of course join separate nets, operated via separate pestrons. this is explicitly not the item under discussion tho )
18:28 punkman wait, so you run dulapnet, I run whorenet with 50 whores, why would you want me to bring the whores to dulapnet
18:29 asciilifeform punkman: likely you wouldn't. but in many cases involving sumthing other than obvious disjunction as with whores -- you may want to
18:29 asciilifeform this is part of the [file:///home/stas/pest/FD/b/pest_spec/pest.html#122-nets-instead-of-channels][explicitly specified!!!] model of net growth.
18:30 asciilifeform and, separately, broadcast messages. are. part. of . the. spec. meaning, even if packet from asciilifeform to punkman is lost, he will still get the message via 11 whatever other people.
18:30 asciilifeform i.e. non-star topology.
18:30 asciilifeform that message is to be distinguishable from noise.
18:33 asciilifeform exactly HOW to do this, remains to be seen. but it aint a question of 'whether'. it's a question of 'who wants to climb down from the fucking trees -- must find how to do this.'
18:36 punkman if we forget about packet size limitation, I suppose message could contain N encrypted copies, one for each peer, and hearsay broadcast now has meaningful sig for recipient, even if not received directly
18:36 asciilifeform punkman: 'onions'. this falls immediately, we have exactly 508 bytes.
18:36 asciilifeform for this reason.
18:36 dulapbot Logged on 2021-09-11 12:42:33 asciilifeform: permitting reassembly is an allcomers hole -- given as you can't authenticate the frags.
18:36 asciilifeform it's a permanent limit for so long as we're on earthling arpa-derived net.
18:37 asciilifeform even if packets could be MBs in size, this'd be nonsense, it doesn't scale to even coupla dozen peers
18:38 asciilifeform because, again, ANY incoming packet has to be handled fast enuff to allow rejection of martians at line rate.
18:40 asciilifeform this problem will have to be SOLVED, punkman . until someone solves it, nice things cannot be had.
18:40 asciilifeform not hand-waved around, no. solved.
18:45 asciilifeform http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057928 << lol, pasted local draft, let's redo :
18:45 dulapbot Logged on 2021-09-13 14:29:32 asciilifeform: this is part of the [file:///home/stas/pest/FD/b/pest_spec/pest.html#122-nets-instead-of-channels][explicitly specified!!!] model of net growth.
18:45 asciilifeform ... explicitly specified!!! model of net growth.
18:46 asciilifeform 'A broadcast message is sent to every member of a station's net. Nets may easily and organically undergo schismatic splits, or, on the contrary, combine into larger nets, whenever the individual station operators so desire.'
18:46 asciilifeform this is not optional.
18:49 asciilifeform in so far as asciilifeform can see, the only way to maintain a reasonably hard identity without asymmetric crypto, is to have a public, unbroken chain of 'i'm the one who wrote X' going all the way back to the 1st message ever sent by $speaker, where he had a pgpgram.
18:50 asciilifeform this however requires having the unbroken chain.
18:52 asciilifeform ( btw in the original unpublished draft, there was a ~third~ field, 'xorchain', containing the xor of all the message hashes from the speaker from t==0. asciilifeform removed it, because while useful adjunct to selfchain, does not actually resolve the fundamental headaches of it )
18:55 * asciilifeform ftr NOT INTERESTED in 'make yet-another chat like all the others'. interested STRICTLY in resolving the fundamental problem once and fucking for all.
19:01 * asciilifeform will post various unpublished/rejected crapola from chalkboard, in hopes that it will stimulate other folx to think :
19:04 asciilifeform ... eh nm, perhaps later.
19:04 * asciilifeform does not want to get people stuck in same mathematical dead end as himself
19:04 punkman it seems to me that any solution that is not "find way to live with hearsay", will amount to pubkeycryptosystem
19:04 asciilifeform punkman: the stick gotta be both long and short. hot and cold.
19:05 asciilifeform asciilifeform already proposed one solution, which by all rights 'ought not be possible', it simply relies on reliable net connectivity for all parties.
19:05 asciilifeform there oughta be another.
19:05 asciilifeform you somehow gotta be able to leverage the fact that direct-peer msgs are ipso facto authentic.
19:08 punkman "unbroken chain of 'i'm the one who wrote X' going all the way back to the 1st message" << is this a solution? I send message on Friday, then go fishing for weekend. anyone else can start growing this "unbroken chain" with hearsay messages
19:09 asciilifeform punkman: and then the genuine punkman shows up
19:09 asciilifeform so, to work, gotta be able to detect fork of arbitrary length.
19:10 asciilifeform either that, or -- from asciilifeform's chalkboard -- erry message includes 2 additional 256bit fields . one is H256(R) where R is a rng turd. the other is the R from previous.
19:10 asciilifeform a message not carrying correct R is immediately known to be fraudulent.
19:11 asciilifeform this scheme is however vulnerable to mitm, anyone who is able to read a message's successor can generate a variant message, because knows the requisite R.
19:11 asciilifeform however it does solve the 'fishing problem' described by punkman .
19:12 asciilifeform i.e. no one can 'extend the chain' without knowing the not-yet-revealed R.
19:12 asciilifeform (from the pov, of course, of those who received the most recent genuine msg from punkman .)
19:15 asciilifeform may as well post the other chalkboardism :
19:16 punkman it's the same problem again, so you commit to secret for next broadcast msg, the commitment only good for direct peers
19:17 asciilifeform ... K gets an additional 256bit component, G; and msgs get an additional 256b field, SG. G is shared with all peers; SG is HMAC signature using G. all of your peers can now authenticate (or impersonate) your hearsay.
19:17 asciilifeform punkman: the commitment would be impervious for direct peers, and ~pretty good~ for indirect
19:18 asciilifeform but still imho ugly hack.
19:18 cgra asciilifeform: pardon if i missed something, but: where do you end up, if handles aren't "first names of various people" and hence eventually collide, but specify 'must be properly random string min x characters'? -- and perhaps deal with the nicknaming issue on each console's end instead
19:19 asciilifeform cgra: they'd be obv. less likely to collide by happenstance. but what about when deliberately ?
19:20 asciilifeform the way asciilifeform sees it, the situation of deliberate collision probably cannot be ~prevented~. but it MUST be ~detectable~.
19:21 asciilifeform the other thing, cgra , is that one MUST be able to receive meaningful messages from l2 (and optinally l3 etc.) this requires human-readable handles inside msgs.
19:21 dulapbot Logged on 2021-09-13 14:45:28 asciilifeform: ... explicitly specified!!! model of net growth.
19:22 asciilifeform *optionally
19:22 punkman SG tells me that one from group signed the hearsay, but I already know this, peer signed and peer is from group
19:22 asciilifeform punkman: it'd be for the case where e.g. asciilifeform is talking but for some reason his packets cannot reach punkman , but can reach e.g. signpost .
19:22 cgra asciilifeform: this 'handle must not collide' kinda got me thinking every Joe must be separately introduced to me before the name carries a specific enough meaning
19:23 asciilifeform i.e. the non-star topology
19:24 asciilifeform cgra: at the same time a net oughta be able to contain a 'joe'.
19:24 asciilifeform but ~no other station~ on the net can at any point be a 'joe'.
19:26 asciilifeform this case has to be handled correctly, always.
19:26 dulapbot Logged on 2021-09-13 14:25:56 asciilifeform: punkman: say asciilifeform has peers a,b,c. punkman has peers x,y,z. asciilifeform and punkman became peers.
19:26 asciilifeform even if 'c' and 'z' are both 'joe'.
19:26 asciilifeform for one thing, there's no way to even know about this apriori.
19:28 cgra asciilifeform: which part of spec should i look at again, to grasp how the deliberate handle collision would possibly happen, if otherwise issue mitigated by random handles? or, if you can work a simple example for me
19:30 asciilifeform cgra: suppose that for whatever reason, a packet or series of packets from asciilifeform is able to reach only signpost and punkman . at the same time asciilifeform is addressing his entire net ;
19:32 asciilifeform meanwhile packets from cgra are able to reach asciilifeform's entire net. cgra has a peer 'nebuchadnezzar', who is not a peer of asciilifeform or anyone else currently on dulapnet. nebuchadnezzar is peer with shalmaneser, who also not peer w/ anyone else. the latter decides to impersonate asciilifeform.
19:33 asciilifeform so when the genuine asciilifeform speaks , folx see ' asciilifeform(signpost) : .... ' and asciilifeform(punkman) : ....'
19:33 asciilifeform when the sham asciilifeform speaks, people will see ' asciilifeform(cgra) : surrender to reich with me! ' or similar.
19:34 asciilifeform without something like the unbroken-chain scheme, it is impossible for anyone, incl. asciilifeform's l1, to distinguish the sham from the genuine.
19:34 asciilifeform i.e. without authenticable (in any sense) hearsay, we have a star topology again. where the only packets that can have any meaning are such that are sent between direct peers.
19:44 asciilifeform ... in the above case, the genuine and the sham are at least in ~some~ way distinct. but suppose that both are connected to the net via same peer !
19:45 asciilifeform this is easy to picture, i will omit explicitly drawing it.
19:48 asciilifeform punkman: returning to your 'fishing' example -- one way to distinguish a genuine from a squatter broadcast is that the latter will nearly inevitably get relayed by ~only one~ peer
19:49 asciilifeform whereas the genuine -- sends to his l1 ~simultaneously
19:49 asciilifeform there oughta be a way to make use of this fact.
19:51 asciilifeform it is exceedingly unlikely that e.g. asciilifeform will have to carry out a broadcast by hearsaying through ONE thin path into the net.
19:58 cgra asciilifeform: "all my buddies say asciilifeform said so" vs "cgra said nebuchadnezzar said asciilifeform said otherwise", kinda fits the 'hearsay' term
20:01 asciilifeform right, but how to mechanize.
20:01 asciilifeform remember, per current spec, the messages are deduped
20:01 asciilifeform so you will never see , on your console, the same message from multiple paths.
20:01 punkman http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1058001 <> http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057985
20:01 dulapbot Logged on 2021-09-13 15:51:11 asciilifeform: it is exceedingly unlikely that e.g. asciilifeform will have to carry out a broadcast by hearsaying through ONE thin path into the net.
20:01 dulapbot Logged on 2021-09-13 15:26:08 asciilifeform: this case has to be handled correctly, always.
20:02 asciilifeform punkman: the potential legitimate prior existence of two asciilifeforms is a separate (tho related) problem.
20:03 * asciilifeform will post, for completeness, the last of the chalkboardisms :
20:03 punkman meant to link http://logs.nosuchlabs.com/log/asciilifeform/2021-09-13#1057913
20:03 dulapbot Logged on 2021-09-13 14:25:56 asciilifeform: punkman: say asciilifeform has peers a,b,c. punkman has peers x,y,z. asciilifeform and punkman became peers.
20:03 asciilifeform suppose that any message having a variant netchain from the receiver's value is rejected.
20:04 * asciilifeform must brb
~ 21 minutes ~
20:25 asciilifeform sooo punkman , cgra , et al: let's try this 'solution' on for size :
20:26 asciilifeform a hearsay message is received; and there may be a certain number of duplicates of it received, in a defined time window (refer as Td.) normally we simply throw'em away and forget about'em. but suppose they get ~counted~
20:27 asciilifeform lemma : a station can only receive as many duplicates of message M, such that have a bounce==1, as it has peers.
20:28 asciilifeform per this lemma, you now have a useful number associated with any hearsay message : the # of peers from whom a dupe of said message , having bounce <= 1, was NOT received.
20:29 asciilifeform if this number (call it B... bogosity?) != 0, it is displayed in parens following the handle(peer) string.
20:30 asciilifeform e.g. in this example, if you received gen1 dupes of the msg from all but 2 of your peers, you will see e.g. ' asciilifeform(signpost)(-2) : ... '
20:30 dulapbot Logged on 2021-09-13 15:33:11 asciilifeform: so when the genuine asciilifeform speaks , folx see ' asciilifeform(signpost) : .... ' and asciilifeform(punkman) : ....'
20:31 asciilifeform but if in fact one of, say, shinohai's camwhores is impersonating asciilifeform , you will see ' asciilifeform(shinohai)(-12) ' supposing you have 13 peers.
20:33 asciilifeform there may be a cleaner algo to make use of the simultaneous burst of genuine msg. but i suspect will be minor variation on the above theme.
20:33 dulapbot Logged on 2021-09-13 15:49:09 asciilifeform: whereas the genuine -- sends to his l1 ~simultaneously
20:34 asciilifeform imho oughta be used ~in conjunction~ with selfchain, rather than in place.
20:34 * asciilifeform invites comments
20:36 * asciilifeform proposes a value of 1s for 'Td' .
20:38 asciilifeform in fact already specified this in 4.1.2.2.2. In-WOT Hearsay. simply, now also count the # of peers from whom got bounce<=1 dupes of a given msg.
20:40 asciilifeform station operator could configure a minimal bogosity that triggers the (-xx) . or see any where != 0. entirely matter of taste.
20:40 asciilifeform could even reject in-wot hearsay with bogosity over $threshold entirely.
20:42 asciilifeform ... if you got ~more~ dupes than you have peers, you now know that you have a traitor in your wot, who sent a bogon with bounce==0.
20:43 asciilifeform i.e. an attempted impersonation by someone in the victim's l1.
20:44 asciilifeform anyways will attempt a more rigorous formulation of the algo laters.
20:44 * asciilifeform bbl
20:55 punkman one problem is that all messages from poorly connected station, will have high bogosity
20:55 asciilifeform his problem.
20:55 asciilifeform and imho entirely appropriate. want folx to see low bogometer reading? peer with moar people in that net.
20:56 asciilifeform meanwhile lurkers write in.
20:56 asciilifeform apparently subject interests moar folx than simply dulapnet denizens...
20:56 * asciilifeform genuinely bbl
~ 2 hours 9 minutes ~
23:06 asciilifeform for future ref, the proposed pill. (btw billymg seems that you broke multiline highlight in your css...)
23:06 * asciilifeform equally at fault, tsk, signed w/out fully testing
23:07 asciilifeform correct link w/ anchor grr
23:07 dulapbot Logged on 2021-09-13 16:25:46 asciilifeform: sooo punkman , cgra , et al: let's try this 'solution' on for size :
23:09 billymg asciilifeform: will fix, i didn't know multiline highlight was a feature even
23:12 asciilifeform billymg: it has been since almost the start; but is awkward to use (have to form the url by hand)
23:15 billymg asciilifeform: i have the fix, it's a oneliner
23:15 asciilifeform 1moar refinement to the scheme -- only count as bogowitnesses (peers who did NOT bring a dupe of the hearsay msg) such peers as, within last 15m, have sent in ~anything~ (let's define'em as 'online'.)
23:15 dulapbot Logged on 2021-09-13 16:28:23 asciilifeform: per this lemma, you now have a useful number associated with any hearsay message : the # of peers from whom a dupe of said message , having bounce <= 1, was NOT received.
23:15 asciilifeform billymg: hm?
23:16 billymg for the multiline select highlight
23:16 asciilifeform billymg: let's have it ( and vpatch at billymg's leisure )
23:18 billymg asciilifeform: http://paste.deedbot.org/?id=wZuk
23:19 asciilifeform billymg: worx! ty!
23:19 billymg asciilifeform: np
23:19 billymg specificity issue, apologies for the oversight
23:19 asciilifeform no worry
23:26 billymg asciilifeform: i've started working on a sort of marketing page for pest at http://pest.bitdash.io/ -- right now the homepage just links to a neatly formatted copy of the current draft spec, but next step is adding an "overview" button to the homepage that will lead to a powerpoint-esque summary with diagrams/visual aids
~ 16 minutes ~
23:42 asciilifeform nifty, billymg , ty for mirroring !
23:43 asciilifeform billymg: i've been hoping to bake diagrams of key mechanisms (svg?) but nfi when will get to.
23:43 asciilifeform billymg: if you like that kind of thing please feel welcome to try your hand at it
23:43 * asciilifeform promises to review
23:47 asciilifeform billymg: i like the pretty-printing ; if you want to share your script for generating this from the .md -- plox to vpatch, i'ma make use of it in subsequent updates
23:48 asciilifeform indeed it is easier on the eyes. (esp. vs the mutilation which appears on asciilifeform's www when pasting in the output of 'markdown2')
23:48 billymg asciilifeform: ty, all i did was view-source on your www and copy/pasted the html you had generated
23:48 billymg the rest is just different css applied
23:48 asciilifeform a hm
23:48 billymg (with very very minor html adjustments only when absolutely necessary)
23:48 asciilifeform billymg: see the makefile btw, it's straight output from 'markdown2'
23:49 billymg asciilifeform: will check it out, the view-source, copy/paste was just a quick and dirty method
23:49 asciilifeform ftr markdown is a rancid piece of shit and barely usable.
23:49 dulapbot Logged on 2021-07-11 19:26:49 asciilifeform: meanwhile, asciilifeform went 'can use a footnote >1nce in 'markdown' formatter ?' -- apparently not! and the fuckwads/'maintainers' are proud!'
23:50 asciilifeform and and.
23:50 dulapbot Logged on 2021-07-11 19:37:41 asciilifeform: or say i want to rename a section, or insert one between existing sections.
23:50 dulapbot Logged on 2021-07-11 19:37:49 asciilifeform: now have to manually change 9000+ links? WHY?
23:51 asciilifeform a civilized replacement (sexprs!!!) for 'markdown' would be an a++ noob project.
23:51 asciilifeform sane handling of section names, numbering, endnotes, backlinks, table of contents...
23:52 asciilifeform 'if wishes were horses'
23:54 billymg asciilifeform: someday...
← 2021-09-12 | 2021-09-14 →