Hide Idle (>14 d.) Chans
← 2022-01-19 | 2022-01-21 → ↓
dulapbot: Logged on 2022-01-19 17:38:56 asciilifeform: promises to fix shortly, unless billymg gets there 1st
dulapbot: Logged on 2022-01-19 22:58:08 jonsykkel: breaks/forks handling is trickyest stuff imo by far, all kinda difrent cases, multiple breaks fusing togheter, breaks turning into forks, forks within breaks etc etc
whaack: morning
whaack: verisimilitude: you gave me some good chuckles reading the log yesterday
whaack: i skimmed it so i knew what was coming, and then reading down the lines.... 'ascii i have found a bug i can exploit *goes silent for a while* Ladies and gents, Behold! *nothing happens* Sorry about that, behold! *nothing happens* okay this time *nothing happens* i give up ... nvm got it *spam ensues*'
billymg: !. version
bitbot: I am bot version 719639.
dulapbot: Logged on 2022-01-19 17:18:57 dulapbot: Logged on 2022-01-19 17:18:27 verisimilitude: http://logs.bitdash.io/asciilifeform/2022-01-19#1074821
bitbot: Logged on 2022-01-19 22:16:37 verisimilitude: I've the math figured out.
billymg: hrm
billymg: should've tested locally
billymg: ok that's strange, the same code seems to work locally, perhaps i messed up the config somewhere
asciilifeform: billymg: make sure the bots list in the deployed one is up to date
billymg: !. version
bitbot: I am bot version 719639.
billymg: !. version
bitbot: I am bot version 719639.
dulapbot: Logged on 2022-01-19 17:18:57 dulapbot: Logged on 2022-01-19 17:18:27 verisimilitude: http://logs.bitdash.io/asciilifeform/2022-01-19#1074821
billymg: ok, i indeed messed up my config, but in a pest related knob. i got tripped up when casting a False in the config to a boolean (is True because non-empty string)
whaack: billymg: what are your plans for using trbexplorer? i think we discussed this a long while back, but if you are planning on putting an interface on bitdash then i recommend you (temporarily) use a python script to get the data from explorer.ztkfg.com.
billymg: whaack: yeah, i'd like to put a fancy frontend on it at some point
billymg: if your plan is to genesis the "backend" then i would have to do the work of standing up my own and modifying it to spit out e.g. JSON for easy parsing on the flask side
whaack: billymg: i already have the script that takes what explorer.ztkfg.com returns and puts results into a python object
billymg: ah, that would be even better then
whaack: only reason i haven't published / sent it out is i'm still testing and making sure i don't want to make any changes to the api first
billymg: whaack: makes sense
whaack: right now i'm wrestling with the idea that i need to keep the block explorer up to the chain tip as fast possible, from a practical use scenario it's going to be quite annoying that trbexplorer is always 2-3 blocks behind the tip
billymg: yeah, definitely, that's a solid half hour in block time
billymg: no rush on my end, there are still features i'd like to add to the crawler, and some guides i'd like to publish
billymg: btw, verisimilitude, you can try to break the bot(s) again if you like
whaack: alright cool, i'm also going to make a nice interface myself, but it'll be good to have multiple
billymg: yup, the more the merrier
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-20#1075218 << it's by far the least fleshed out mechanism in asciilifeform's spec so far
dulapbot: Logged on 2022-01-20 08:48:49 thimbronion: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-19#1075215 << my experience as well
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-19#1075215 << re the matter of chain breaks -- imho the objective should be 'no long-term chain breaks from station's pov'.
dulapbot: Logged on 2022-01-19 22:58:08 jonsykkel: breaks/forks handling is trickyest stuff imo by far, all kinda difrent cases, multiple breaks fusing togheter, breaks turning into forks, forks within breaks etc etc
asciilifeform: if you went into a submarine and emerged 3wks later, yer station oughta be able to fully close whatever gaps in both netchain (i.e. broadcast msgs) and indiv. folxs' selfchains
asciilifeform: the simplest way to make this happen is that folx specify x MB for how much pestnet log they're willing to store, and then chances are that 1 of yer peers will fill you up when asked.
asciilifeform: longbuffer still oughta cap at 1h / N MB of msgs (whichever greater) tho, to cap cost of dedup
asciilifeform: ( if folx recall, the entire point of timestamp is that you dun have to walk a century of log to dedupe incomings )
thimbronion: whaack, etc: the source for the the ruby bot and indexer I made (that actually worked) is here: http://share.alethepedia.com/lekythion/
asciilifeform: ( also recall that ~anyone~ who is able to get hold of a valid packet can send it to you x9000 as a dupe flood )
whaack: thimbronion: ty
asciilifeform: ( see also -- that's approx. how long you have to dedupe an incoming. no more. )
dulapbot: Logged on 2021-09-23 15:16:44 asciilifeform: 1e9 / 8*551 ~= 226860 packets/s, if all received packets match the 446byte length (ones which do not, you throw out immediately, they do not cost very much except in bw terms)
dulapbot: Logged on 2021-09-23 15:21:07 asciilifeform: i expect you'd have at least 8cores on reasonable box, which gives you ~105840 ticks to process a packet, imho entirely doable.
asciilifeform: ^ in light of which 'x MB' is rather optimistic.
asciilifeform in fact thinking that for line-rate operation, the receiver prologue in fact oughta look at chain hashes ~before~ searching longbuffer, and deprioritize anyffin that aint pointing to the immediately prior msg
asciilifeform: 1e5 cycles aint enuff to query even a 128kB buffer, likely
asciilifeform: need a rejection pass b/w 'stale' and 'dupe', which could consist of e.g. 'packet that doesn't chain to the immed. previous one received is interned and processed on as-cpu avail. basis'
asciilifeform not sure whether this monologue makes sense to anyone, but hopefully the elaboration in 0xfa will
asciilifeform: also this ('SelfChain hashes are computed strictly for Broadcast Text and Direct Text Messages. For all other Message types, the values of these fields are undefined.') may be a catastrophic mistake.
asciilifeform: under the current algo, a flood of captured packets turned into a dupe flood which includes e.g. a prod will end up querying the longbuffer errytime and exceed the avail. cycles to maintain linerate rejection.
asciilifeform: jonsykkel, thimbronion , et al ^ invited to comment
asciilifeform: in other gnarl, this is probably the Wrong Thing -- 1st oughta ask the originator, if you can, or Rm set, if you can't
asciilifeform: grr, broken highliter,
asciilifeform: points to 'If the "offending" message was a Broadcast Text, at least one GetData request is issued to each peer of the station.'
thimbronion: whaack: here is a dump of the lisp indexer, which
thimbronion: lol, which rendered links using headless chrome and saved the resulting text in a postgres db: http://share.alethepedia.com/lekythion/log-indexer.tar.gz
whaack: thimbronion: nice, i'll probably be able to grok that quicker than the ruby code
whaack has only done 3 months in the ruby saltmine
asciilifeform: $ticker btc usd
busybot: Current BTC price in USD: $42921.54
asciilifeform: !w poll
watchglass: Polling 14 nodes...
watchglass: 71.191.220.241:8333 : (pool-71-191-220-241.washdc.fios.verizon.net) Alive: (0.038s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666 (Operator: asciilifeform)
watchglass: 205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.081s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719666
watchglass: 205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.082s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 205.134.172.27:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666 (Operator: asciilifeform)
watchglass: 205.134.172.26:8333 : Alive: (0.141s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719666
watchglass: 54.39.156.171:8333 : (ns562940.ip-54-39-156.net) Alive: (0.112s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 208.94.240.42:8333 : Alive: (0.143s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 205.134.172.28:8333 : Alive: (0.083s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719666 (Operator: whaack)
watchglass: 54.38.94.63:8333 : (ns3140226.ip-54-38-94.eu) Alive: (0.258s) V=88888 (/therealbitcoin.org:0.8.88.88/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 82.79.58.192:8333 : (static-82-79-58-192.rdsnet.ro) Alive: (0.277s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 94.176.238.102:8333 : (2ppf.s.time4vps.cloud) Alive: (0.308s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=718917
watchglass: 103.36.92.112:8333 : (terebe.ns01.net) Alive: (0.642s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 75.106.222.93:8333 : Alive: (0.469s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719666
watchglass: 143.202.160.10:8333 : Busy? (No answer in 100 sec.)
jonsykkel: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-20#1075268 << according to new spec longbuffer is indexed by hash, shud be like <10e3 cycles regardless of baffer size?
dulapbot: Logged on 2022-01-20 11:55:29 asciilifeform: 1e5 cycles aint enuff to query even a 128kB buffer, likely
asciilifeform: jonsykkel: recall that the interval also includes verifying sig against erry k(s) in wot
jonsykkel: sure, talking about just querying the buffer step
asciilifeform: indeed oughta be queryable in o(n log n) or around
asciilifeform: absolutely must rule out disk assess tho
asciilifeform: if an arbitrary packet can force you to access disk, yer sunk
jonsykkel: inded
whaack: asciilifeform: (or anyone else running trb) do you perhaps know what the lag is between when heathen block explorers see new blox and when trb sees new blox (in either direction, perhaps trb is getting them first)
whaack: restarting trb explorer...
whaack: !e help
whaack: !e help
trbexplorer: whaack: my valid commands are: src, uptime, version, help, view-address, help, view-height, scan, view-raw-tx, verify-block, view-raw-block, view-tx, reset, view-balance, view-block, view-merkle-root, remove-top-blocks, verify-all-blocks, view-utxos, push
whaack: !e help
trbexplorer: whaack: my valid commands are: src, uptime, version, help, view-height, view-address, view-utxos, view-balance, view-block, view-raw-block, view-tx, view-raw-tx, push
whaack: !e view-height
trbexplorer: block_height: 719671
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-20#1075307 << asciilifeform not noticed a substantial lag. (and often enuff my noades see'em before certain heathen exploders do)
dulapbot: Logged on 2022-01-20 15:29:40 whaack: asciilifeform: (or anyone else running trb) do you perhaps know what the lag is between when heathen block explorers see new blox and when trb sees new blox (in either direction, perhaps trb is getting them first)
asciilifeform: iirc was pocket change, too
asciilifeform: 'illegal bitcoin'(tm)(r)(c)
bonechewer: speaking of "illegal bitcoin", ru seems to now want to stamp it out in favor of a digital ruble
← 2022-01-19 | 2022-01-21 → ↑
Random(asciilifeform) | Download hourly DB snapshot | Get Source Code