Hide Idle (>14 d.) Chans
← 2022-07-14 | 2022-07-16 → ↓
01:52 signpost http://wot.deedbot.org/ << I made this index page a little more useful by showing folks in my l3 with >=2 positive ratings from someone.
↖
01:54 signpost there are plenty of names here I'd be happy to work with again as the world burns.
01:55 signpost graphviz organized the thing pretty sensibly once I deleted deedbot and switched to "neato". perhaps I can persuade y'all to update your ratings as you currently see them.
~ 8 hours 7 minutes ~
10:03 asciilifeform signpost: nifty. see what it'd look like if you take out the corpse and extend 1 level.
~ 52 minutes ~
10:55 shinohai http://logs.bitdash.io/pest/2022-07-15#1009589 << pretty neat, I'm seeing the graph images over user's fingerprint information when you click individual pages tho.
10:55 bitbot Logged on 2022-07-15 01:52:21 signpost: http://wot.deedbot.org/ << I made this index page a little more useful by showing folks in my l3 with >=2 positive ratings from someone.
11:06 PeterL signpost: on individual pages, e.g. http://wot.deedbot.org/F595258CD9F1FDBDEFAE5093B4E2E0DCACB63D25.html , it looks like the graph is sometimes covering up the text above and below it?
~ 39 minutes ~
11:46 signpost asciilifeform: dun wanna remove human ratings for folks; I only removed deedbot's. but I'll give another level a try.
11:46 signpost ideally the page indicates who interesting folks outside one's own l2 may be.
11:46 signpost shinohai: PeterL: ty, will try to unfuck the CSS. might need to ask for halp from billymg.
↖
11:47 signpost was having a bitch of a time getting the SVG to scale properly.
11:47 * signpost bbl
11:49 asciilifeform signpost: notion wasn't even 'bury mp cuz he's dead, haha' but that afaik nobody gives half a rat's arse re the over9000 whores etc he rated, who will fill the screen if you show +n levels
11:50 * asciilifeform would find, e.g. pankakke's ratings interesting, even tho just as ded
11:50 signpost I can get behind a general rule of "n00bs don't make the front page" by requiring rating of >=2
11:51 asciilifeform also worx
11:51 signpost that should remove all of the tits
11:51 asciilifeform a++
11:51 signpost well, most!
11:51 billymg signpost: nice, definitely a more useful visualization now
11:52 billymg the latest version seems to be pushing the header (search box, index) off the screen for me
11:53 asciilifeform still loox like needs +1 level (all the visible folx are immediately recognizable #t veterans)
11:53 asciilifeform well, nearly all
11:56 asciilifeform signpost: theoretically a 'rated somebody in past n yrs' filter might be useful
11:58 billymg http://logs.bitdash.io/pest/2022-07-15#1009598 << what i found easiest for the two line charts (also svg) on http://bitdash.io/ was to not touch the svgs themselves at all and instead scale their containers
11:58 bitbot Logged on 2022-07-15 11:46:58 signpost: shinohai: PeterL: ty, will try to unfuck the CSS. might need to ask for halp from billymg.
12:01 billymg i also remember it requiring some fiddling with the output scale from matplotlib, so that the proportions looked right when shrunk or stretched in the browser
12:04 billymg i actually ended up having to serve two different svgs for each chart, one for mobile/small screen proportions, and one for large screen. and then each of those scales freely in the viewport
12:06 billymg it wasn't possible to just pick a middle ground and shrink/zoom in the browser because then the axis labels would either look comically huge or be too small to read
~ 36 minutes ~
12:42 whaack !e view-height
12:42 trbexplorer block_height: 745102
12:42 trbexplorer mins_since_last_block: 7
12:43 whaack nice, back near tip
12:44 whaack i'm interested in creating a clean webpage for trbexplorer soon, i have trbexplorer.com claimed
12:48 asciilifeform whaack: nifty
12:51 whaack not sure yet whether i plan to support searching for txs in the mempool, or how to support heathen addresses (bc1.. 3...)
12:52 whaack one solution is just generating links to other block explorers if someone searches for a heathen address
12:53 whaack i.e. 'segshit not suppported, see results from heathen browser: blockstream.info/segwitaddressyoujustsearched/'
~ 2 hours 22 minutes ~
15:16 PeterL !!rate thimbronion 2 aka awt; authored blatta as prototype pestron
15:16 deedbot Get your OTP: http://paste.deedbot.org/?id=89bS
15:17 PeterL !!v 040EDD69958EECF9D4D9881539B59A9CA19BAFF9B1F3EEB01ED1AEB41C4EFE0C
15:17 deedbot PeterL rated thimbronion 2 << aka awt; authored blatta as prototype pestron
15:18 PeterL !!rate signpost 2 aka trinque; runs deedbot
15:18 deedbot Get your OTP: http://paste.deedbot.org/?id=oVtR
15:19 PeterL !!v 384EEAA7F8E619C0820FE860567F864B6ABEAD5D41174C14F076A4CE30E51173
15:19 deedbot PeterL updated rating of signpost from 1 to 2 << aka trinque; runs deedbot
15:22 PeterL !!rate asciilifeform 3 designed pest, ffa, other stuff. solid dude.
15:22 deedbot Get your OTP: http://paste.deedbot.org/?id=hADq
↖
15:23 PeterL !!v 1656CEF889B4B9412CD3EE07ECD6EE4E0478008F0204D6AEA8B1958809089A50
15:23 deedbot PeterL updated rating of asciilifeform from 1 to 3 << designed pest, ffa, other stuff. solid dude.
15:24 PeterL !!rate ben_vulpes 1 formerly board for The Real Bitcoin Foundation
15:24 deedbot Get your OTP: http://paste.deedbot.org/?id=SrAE
15:25 PeterL !!v 442E9ED243D035343FD7D813A1FB13AAA9C91A012AA8F42D66C4AE98A3E7CA9C
15:25 deedbot PeterL updated rating of ben_vulpes from 1 to 1 << formerly board for The Real Bitcoin Foundation
15:25 PeterL !!rate mod6 1 formerly board for The Real Bitcoin Foundation
15:25 deedbot Get your OTP: http://paste.deedbot.org/?id=rYhX
15:26 PeterL !!v 268FEE0E4E0A15C995D832C7010EABA56549F16EE08041F5FFFB04BF05C060F6
15:26 deedbot PeterL updated rating of mod6 from 1 to 1 << formerly board for The Real Bitcoin Foundation
15:27 PeterL !!rate shinohai 1 hangs around, is sometimes helpful
15:27 deedbot Get your OTP: http://paste.deedbot.org/?id=i8R2
15:28 PeterL !!v 985CD500782BDBEB062D7B51C2F4316D157B531DA21224C0A30660B8534A7394
15:28 deedbot PeterL rated shinohai 1 << hangs around, is sometimes helpful
15:29 PeterL !!v unrate cazalla
15:29 PeterL !!help
15:29 deedbot http://deedbot.org/help.html
15:30 PeterL !!unrate cazalla
15:30 deedbot Get your OTP: http://paste.deedbot.org/?id=lUSs
15:30 PeterL !!v E3C285F219C9CADBF1A296BE7108758CEF5394D4F522C90E329D9565778CE8B0
15:31 deedbot PeterL unrated cazalla.
15:32 PeterL !!rate billymg 1 did a lot of work fixing up MP-WP
15:32 deedbot Get your OTP: http://paste.deedbot.org/?id=6N1v
15:33 PeterL !!v F898D9227CCEBB4A57B059DD2E2DFB1F4703C03D5324A5BA6E06D72E495104F2
15:33 deedbot PeterL rated billymg 1 << did a lot of work fixing up MP-WP
~ 27 minutes ~
16:00 shinohai I guess that's better than "is never helpful".
~ 3 hours 17 minutes ~
19:17 phf trinque: is deedbot wot database available in a dump somewhere? or it's proprietary information :)
19:18 phf *signpost
19:19 signpost very serious trade secret pls no stela
19:19 signpost *steal
19:19 signpost !!wot
19:19 signpost ^ let's see if this still works
19:23 phf i was thinking, since forever, i think maybe even tried to bring it up to mp's "why do it that way what are you stupid or worse an engineer!", but maybe misremembering, that maybe wot ought to, or could, just be a public log of signed "rate <x> <n> <comment>"
19:24 phf where x is a symbol that might or might not map to a nick, but the nick to pgp hash correspondence and uniqueness is not actually property of wot
19:25 signpost the argument for not compiling a database of signed ratings was that it creates proof of a relationship which The Man may use against you.
19:25 signpost whereas current model creates a cache which suggests you go ask the human about their rating if you like.
19:26 phf this way can also accomodate the old tmsr condemnation rate `rate president-macron -100 le frog`. at the end get a kind of rated tag space of opinions and perspectives
19:26 signpost personally I like the world where one swears to their sayings, but can see the above argument also.
19:26 signpost wot command might be borked, used to shit an sexp of wot.
19:26 * signpost will see about fixing.
19:27 phf hmm, i possibly remember that conversation
19:28 signpost upstack, regarding the wot database it should eventually be an entirely decentralized mechanism of some kind. me dying shouldn't mean shit.
19:29 phf signpost: being the master of logs at some point, i obv see everything in terms of one long neverending scroll of communication, that one can mine for information, so e.g. having a process walk the log ought to produce a complete and current wot (subsequent rates override previous ones, etc.)
↖
19:32 signpost I think this is right. ideologically speaking the log should contain all history.
19:32 phf probably should reflect on what pest means for that kind of thing. with trust implications of pest, is it enough for station to say `i rate x at y with comment` and have an observe be able to make some claims about the implications of such statement. e.g. l1 treats that as `fact` l2 as useful hearsay, etc.
19:32 signpost like, All. History.
19:34 signpost there's a fundamental republican cultural object in there somewhere.
19:34 signpost vpatches also need this, a place to be recorded.
19:36 phf pest blockchain!!1
19:39 signpost re: observers and how truth propagates it's interesting. one could imagine swearing to having observed, and so on.
19:41 signpost one can start getting allergic about "PKI" but the problem there's the centralization and bureaucracy, not chains of trust.
19:42 * signpost also finds himself still pulled by Voice of Authorization to contemplate WoT, has various items on desk in that direction.
19:45 phf i'm so far designing my pest to keep the packets indefinitely. considering that the whole of tmsr, including that utterly pointless b-a ancient history, was about 450mb in memory. from that perspective if things like rating become simply some form of structured pronouncements, one essentially keeps “own wot”
↖
19:50 phf a bounces 0 packet from asciilifeform is as good as truth to his peers, but everything else requires additional investigation
19:56 signpost pest seems in the direction of the right model tbh, and not just a good-enough hack
19:56 bitbot (asciilifeform) 2022-06-06 asciilifeform: recognizes the 'inelegance' of the h(s) scheme , 'yer trying to get the win from rsa sigs w/out paying price of rsa, and will never be as good as Real Thing' but imho over9000x beats 'nuffin'
19:56 signpost non-opposable with hard assurance at l1, attenuates thereafter.
19:57 signpost everybody has as much assurance in their own copy of history as their relationships grant them.
↖
19:57 signpost and even at the edges one can find it pretty likely, even if not mathematically provable.
~ 2 hours 36 minutes ~
22:34 awt !!help
22:34 deedbot http://deedbot.org/help.html
22:45 asciilifeform signpost et al: per asciilifeform , the 'ask x about y' oughta be integrated pestronically. but iirc already described this
22:47 asciilifeform http://logs.bitdash.io/pest/2022-07-15#1009694 << best kind of signature for sumthing like ratings imho.
22:47 bitbot Logged on 2022-07-15 19:57:09 signpost: everybody has as much assurance in their own copy of history as their relationships grant them.
22:48 asciilifeform http://logs.bitdash.io/pest/2022-07-15#1009679 << indeed, given that errybody gets complete log (eventually), can even sift'em outta log.
22:48 bitbot Logged on 2022-07-15 19:29:49 phf[signpost|billymg]: signpost: being the master of logs at some point, i obv see everything in terms of one long neverending scroll of communication, that one can mine for information, so e.g. having a process walk the log ought to produce a complete and current wot (subsequent rates override previous ones, etc.)
22:49 asciilifeform (gets tricky in re 'bob problem' however)
22:49 asciilifeform will need sumthing like a solution.
23:03 phf what's bob problem again
23:04 asciilifeform e.g.
23:04 bitbot (asciilifeform) 2022-06-06 asciilifeform: phf: and when 2+ nets merge, and this produces a handle collison ? (aka the 'bob' problem)
23:05 phf oh right
23:05 phf i still refuse to recognize it as a problem :p
23:05 asciilifeform 1 possib. 'solution' lol
23:06 phf namespacing problems always end up with some hairbrained schemes of solutions that never the less fail irl
23:07 phf i have a solution to bob problem: when merging nets, if there's a collision, colliding individuals fight to surrender in a thunderdome. the loser is forever banished from civilized protocols.
23:08 * asciilifeform not has any reason to concretely believe in the solvability of 'namespace problem' in the general case; other than 'die trying' approach
23:08 bitbot (trilema) 2017-05-31 asciilifeform: and if someone wants to mention godel etc -- ethical engineer MAY NOT cite godel, EVER, just as a police detective MAY NOT cite the supernatural and admit a hypothesis of miraculous theft from a safe
23:08 phf of course one can surrender ones designator for a different one, and avoid himself the trouble
23:08 asciilifeform phf: a++ solution. nao, to implement...
23:09 phf if either parity refuses to surrender the identifier OR to participate in fight to surrender in a thurderdome, both parties are booted from the resultant network
23:09 phf there, a COMPLETE protocol
23:09 * asciilifeform in favour!111
23:10 asciilifeform even to have'em play 'rock/paper/scissors' would be great. nao if could distinguish'em..
23:11 phf a representative of the net responsible for the merger speaks on behalf of member
23:12 phf there are some issues to the protocol though. the members of the “losing” network have to cary the cognitive load of readjustment. perhaps bitcoin compensation distributed evenly is in order :D
23:12 asciilifeform naively might suppose that the 1st merge peer would know on what side he expects 'bob'. but in actuality not knows, because the 'wrong' side may in fact be already peered, via some # of hops, to 'right' bob
23:13 * asciilifeform fully expects, in all srsns, that there'll be pestnets where pay to play, somewhere
23:14 phf i agree that pest perhaps should be more explicitly about name management, which right now exists both out of bound (where you keep “operator” in the station's definition) and in bound
23:15 * asciilifeform can't find thread atm, but at one pt suggested nixing the 'speaker' field entirely, in favour of a 'i'm x' msg type, which folx dredge outta log
↖
23:15 asciilifeform then can decide what to call their 'bob' via sequencing of same
23:16 phf yeah, on related note, what happens when bounces is 0, but speaker is not station operator?
23:16 asciilifeform per current spec, only immediate or direct msg may have bounce 0
23:16 asciilifeform (if otherwise -- discarded)
23:17 asciilifeform but see, however, 4.3.2
23:18 asciilifeform ^ if immediate msg but speaker != peer's known handle, is specially marked in console.
23:18 * asciilifeform must bbl
23:19 phf ah, ty
23:34 awt Can we assume, given Dunbar's number, which is in the range from 100 to 250, that any two pestnets that fuse will likely be of size s and D - s?
↖
23:34 awt If so, how likely are namespace collisions to be and how difficult would they be to resolve "manually?"
23:39 phf hello world
23:39 phf hehe, i did a dope extension to slime
23:40 phf it's like we're half way to genera
↖
23:50 signpost awt: the weird thing about the internet is that our dunbars aren't one group.
23:51 signpost would be nice to build such that eat-the-world is possible, even while avoiding the hubris of expecting to.
↖
23:51 signpost phf: piiiix
23:54 phf http://logs.bitdash.io/pest/2022-07-15#1009739 << http://glyf.org/screenshots/pest-slime-say.png
23:54 bitbot Logged on 2022-07-15 23:40:01 phf[signpost|billymg]: it's like we're half way to genera
23:54 signpost re: bobs problem, was thinking that a proper design of pesttronic wot might help solve.
23:55 signpost phf: god damn that's nice
23:56 signpost having it in the REPL might be mighty handy.
23:57 signpost upstack on wot, one's identity is what others call them. (NOT what one "identifies as")
23:58 signpost suppose a rating pointed to a given message to indicate identity. "I'm rating whomever produced this chain of messages"
↖
23:59 signpost if you have two people claiming to both be bob, crawl up their self-chains and figure out where they fall in your wot. let UI indicate this somehow, "bob before biff, ben, and beelzebub"
23:59 * signpost might misunderstand the problem, invites correction if so.
← 2022-07-14 | 2022-07-16 → ↑

Random(pest) | Download hourly DB snapshot | Get Source Code

Theme: Light | Dark | System Default