Hide Idle (>14 d.) Chans


← 2018-08-11 | 2018-08-13 →
asciilifeform: in other noose, 1st mats parcel -- eaten. will grind today.
asciilifeform: !Q later tell mod6 do we have spare ssd remaining ? i'ma need one asap, running out of temp storage .
lobbesbot: asciilifeform: The operation succeeded.
asciilifeform: ... in fact, loox like there won't be any moar phuctor output until i get a 1TB disk hooked up ( will have to be on a usb snake, luckily i left BingoBoingo with one -- as all 4 drive bays on dulap are full )
asciilifeform: mircea_popescu: we are gonna have to activate the s.nsa spare box, as a storage cluster ( and get disks to fill same )
asciilifeform: mechanical disks would actually suffice, for this purpose.
asciilifeform: ( 4x, as large as can be had. )
asciilifeform: meanwhile... i'ma do a speshul singleshot run with ~only~ mats parcel1 (i.e. minus errything prior to june) as input , just to get a sense of what lies ahead. output will appear in #asciilifeform later today.
BingoBoingo: Oh, how things grow!
mod6: http://btcbase.org/log/2018-08-12#1841025 << Well, I'm under the impression that there might be a spare for the bitcoin foundation.
a111: Logged on 2018-08-12 15:05 asciilifeform: !Q later tell mod6 do we have spare ssd remaining ? i'ma need one asap, running out of temp storage .
lobbesbot: mod6: Sent 1 hour and 53 minutes ago: <asciilifeform> do we have spare ssd remaining ? i'ma need one asap, running out of temp storage .
mod6: BingoBoingo ben_vulpes : is this true? ^
BingoBoingo: mod6: I'll dig out the cabinet inventory
mod6: Thanks, Sir.
BingoBoingo: There's a TBF marked spare 1 TB samsung in the inventory sheet. If thoroughput isn't absolutely necessary there are also locally available external spinninf rust HDD's: https://www.zonatecno.com.uy/ProductDetail/90355/Disco%20duro%20externo%20Seagate%201TB%20STEA1000400%20USB%203-0 and https://www.zonatecno.com.uy/ProductDetail/90469/Disco%20duro%20externo%20Seagate%202TB%20STEA2000400%20USB%203-0
mod6 looks
mod6: ah, cool. can you get by with an external drive like those then? that might be the best idea, especially for $100-$120
asciilifeform: mod6: problem is that afaik we dun have a 12v snake
asciilifeform: only the typical 5v one ( i.e. needs lappy-sized drive )
BingoBoingo: asciilifeform: Those are the bastard lappy drives in the plasic shells with snake already
BingoBoingo: 1 TB U$S 99, 2 TB U$S 129
mod6: Sincen we only have ~one~ spare remaining on hand until we get our other drive shipments, I kinda wanna hang on to the foundation
mod6: 's spare for emergency.
asciilifeform: lessee later today
asciilifeform: i'ma not ask for anyffing to be bought until mircea_popescu wakes up.
mod6: ok cool.
mod6: will stay tuned.
mircea_popescu: ftr, "crashing" in a "wooded area" sounds like someone didn't look for sidewinder debris at the site.
mircea_popescu: http://phuctor.nosuchlabs.com/search?q=amd+turd << ok this is beautiful. nice work stan!
asciilifeform: i've a irc bot also, but currently just short of working, connection drops
mircea_popescu: http://btcbase.org/log/2018-08-12#1841012 << this, seems good as it is tbh. let the republican style development work, if anyone for some reason needs, the anyone asks by explaining the reason and we see.
a111: Logged on 2018-08-12 00:11 trinque: if all it does is cough up the link to search, wouldn't really need an API
mircea_popescu: asciilifeform connection's been shit, i suspect they're "fighting spam"
asciilifeform: it's a regged luser & all
mircea_popescu: even so.
mircea_popescu: http://phuctor.nosuchlabs.com/search?q=bitcoin << very probably some "platform" or other, keybase ?
asciilifeform: mircea_popescu: they're in order of submission, so the top MB or so is ssl certolade, the rest -- sks
mircea_popescu: http://btcbase.org/log/2018-08-12#1841029 << wouldn't it be better to get mechanicals for this actually ?
a111: Logged on 2018-08-12 15:20 asciilifeform: mechanical disks would actually suffice, for this purpose.
asciilifeform: it would
asciilifeform: 4 fat ones.
mircea_popescu: what am i agreeing with here ? yes i agree pizarro should have on hand a dozen of each type of disk it uses, at the least. and yes as per ancient http://btcbase.org/log/2018-06-13#1825188 it seems a good move for it.
a111: Logged on 2018-06-13 21:44 mircea_popescu: "get the rigging match the wind".
mircea_popescu: or else yes i agree s.nsa can afford more drives if it wants them
asciilifeform: mircea_popescu: this box dun belong to pizarro, recall
asciilifeform: it's ours
asciilifeform: so we're the ones provisioning disks
mircea_popescu: aleright, so then tell me this, why is our box provisioned in such a way you ran out of space ?
asciilifeform: mircea_popescu: erry box is doomed to one day run out of space
asciilifeform: when i provisioned the boxes, eater operated at ~10 units / sec, instead of 150 .
mircea_popescu: never saw this wonder. but anyway, you want basically to keep a 4ssd raid cluster for speed and add a mechanical on usb ?
asciilifeform: ( and there was nobody to collect all of ipv4's ssl etc )
mircea_popescu: wouldn't it be better to have 3ssd raid and the mechanical inside ?
asciilifeform: mircea_popescu: primary dulap has 4 holes, all fulla ssd ; spare dulap has 4 empty holes
asciilifeform: usb suggested as stopgap
asciilifeform: and there is not an 'inside' in either
mircea_popescu: alright, so basically you want to make the spare into a dumb disk box ? isn't that a waste of board ?
asciilifeform: not necessarily waste of board ; could rent out space to l1 folx .
mircea_popescu: i'm not strictly against the configuration, put four large disks on it w/e. see what pizarro says too, but i'm amenable.
asciilifeform: mod6 et al ^
mircea_popescu: do we actually own that spare too ? iirc we did sell them something.
asciilifeform: we own 1 spare .
asciilifeform: ( sold s.mg spare . )
mod6: yeah, this ^
mircea_popescu: ah ok. anyway, if you get spinners, don't get 1tb bs. get 2 or 4 or even 8
mircea_popescu: http://btcbase.org/log/2018-08-12#1841047 << ok, but i mean, can't have phuctor grind to a stop because of it neh ?
a111: Logged on 2018-08-12 17:09 mod6: Sincen we only have ~one~ spare remaining on hand until we get our other drive shipments, I kinda wanna hang on to the foundation
mircea_popescu: anyway, ima go take hussy uphill for a spell, bb in a few hours.
asciilifeform: 10tb mech disks are ~400bux where i sit
asciilifeform: mircea_popescu: depending on how long it takes to get disks, we can do what i suggested a while back ( but for some reason can't dig up, in l0gz) where we trade space for time, and run on randomized subsets of the db continuously
mod6: <+mircea_popescu> http://btcbase.org/log/2018-08-12#1841047 << ok, but i mean, can't have phuctor grind to a stop because of it neh ? << If it's critical, it's critical, and will sell disk to S.NSA.
a111: Logged on 2018-08-12 17:09 mod6: Sincen we only have ~one~ spare remaining on hand until we get our other drive shipments, I kinda wanna hang on to the foundation
asciilifeform: i had this knob on the implementation conveyor regardless, because of a certain peculiarity of bernstein's algo which i noticed, where if there is an even multiple of any given factor present in the whole, it will not get found
asciilifeform: ( hence why the debian collection gets run in two steps, 'p's' then 'q's' )
mod6: asciilifeform: anyway, think it over. let me know.
mod6: we are trying to get more drives down there, as you know. but could be a while until we have any spares at all.
mod6: (spare drives, that is -- which could put is in a pretty bad position if it comes to it)
asciilifeform: mod6: drive shortage will have to be cured ; if i gotta fly with crate of disks, even.
mod6: agree.
mod6: but, VAT.
asciilifeform: mod6: what's the latest on ben_vulpes's disks misadventure ?
mod6: i dunno, I just put in a checkin with him in #pizarro to see about the refund. best case, he got the refund and can ship more drives, perhaps today even.
mod6: BingoBoingo should be ordering drives within a day or so from the btc I sent him on friday.
mod6: But could be a month out, at best.
mod6: Maybe sooner, depending. BingoBoingo is sending his expidited.
asciilifeform: !Q later tell ben_vulpes plox to give mircea_popescu & asciilifeform a quote for getting 5 10TB mech disks of reasonable quality to BingoBoingostan
lobbesbot: asciilifeform: The operation succeeded.
mod6: So what you're thinking is that instaed of the 5 EVO 1tb ssd's that we were supposed to get for S.MG, he'd send 5 10Tb disks for S.NSA instead?
asciilifeform: !Q later tell BingoBoingo plz lemme know re the 1 mech disk in the bilge ( whose is it ? and what size ? and can fit in the snake ? )
lobbesbot: asciilifeform: The operation succeeded.
asciilifeform: mod6: that's what i was thinking
mod6: Well, it could work. But then the drives that BingoBoingo is ordering must go to S.MG's UY3.
mod6: And then we still have no spares until BingoBoingo can place another order. Which I'm fine with, but still leaves us kinda in a sticky spot.
mod6: I'm reluctant to order more drives back to back until we can confirm we can actually receive a shipment.
mod6: (otherwise would just throw him more funds to place a second order immediately)
mod6: Let's think on it anyway. I guess worst case, BingoBoingo could take a day trip to BSAS and buy some drives.
mod6: Or maybe we can see if mthreat wants to ferry over a few.
asciilifeform: iirc mthread went awol aeons ago
mod6: Yeah, just throwin it out there.
mod6: Anyway, your idea sounds like a plan. But can S.NSA live without the mechanical 10Tb spinners for a month?
mod6: (guessing it would take somewhere around that amount of time for the shipment to complete from ben_vulpes)
asciilifeform: mod6 can, either via http://btcbase.org/log/2018-08-12#1841098 , or disk on snake ( if BingoBoingo turns one up )
a111: Logged on 2018-08-12 18:29 asciilifeform: mircea_popescu: depending on how long it takes to get disks, we can do what i suggested a while back ( but for some reason can't dig up, in l0gz) where we trade space for time, and run on randomized subsets of the db continuously
mod6: Ok then.
mod6: !Qcalc 1500/6200
lobbesbot: mod6: 0.241935483871
mod6: !Qcalc 1411/6200
lobbesbot: mod6: 0.227580645161
mod6: We're right under the threshold here.
mod6: !Qcalc 1500/5800
lobbesbot: mod6: 0.258620689655
mod6: !Qcalc (0.23437500+.1)*5800
lobbesbot: mod6: 1939.375
mod6: that'll do it then.
mod6: !!pay BingoBoingo .1
mod6: !!v 37B3B34DF2780C02556CD2B45DA8B9D33786C7E3D6CF33E7A9CA71D0D1162A75
deedbot: mod6 paid BingoBoingo .1
mod6: !!ledger
mod6: So for the record, the total BTC amount for the BingoBoingo SSD order is: 0.33437500
mod6: Whatever we don't spend will be returned to Pizarro, or I suppose can be held in fiat if need be.
mod6: !Qcalc 0.23437500+.1
lobbesbot: mod6: 0.334375
mod6: ok got a few chores to do here quick, bbs.
asciilifeform bbl,meat
a111: Logged on 2018-08-12 18:30 asciilifeform: i had this knob on the implementation conveyor regardless, because of a certain peculiarity of bernstein's algo which i noticed, where if there is an even multiple of any given factor present in the whole, it will not get found
asciilifeform: not multiple , sorry
mircea_popescu: asciilifeform really, i wouldn't put that much finnicking into it.
mircea_popescu: so it waits a finite interval for disks, not end of worl.d
mircea_popescu: in other fucking lulz, check out usg.freenode : http://logs.minigame.biz/2018-08-12.log.html#t20:01:43
lobbesbot: Logged on 2018-08-12 20:01:43: *** Joins: adamg (~adamg@171.240.3.192)
mircea_popescu: they actually patched their ircd to sed 's/spam//'
asciilifeform: mircea_popescu: the even powers thing is a real effect tho, it's why we still occasionally get 'mod phuctored! factors : 3' etc
BingoBoingo: <mod6> Let's think on it anyway. I guess worst case, BingoBoingo could take a day trip to BSAS and buy some drives. << They aren't cheaper over there
mod6: ok noted.
lobbes: http://blog.lobbesblog.com/2018/07/auctionbot-eta-and-status-report/#selection-71.0-71.11 << update: I have completed a successful test run of an external .py script communicating on IRC via logbot and the listen/notify mechanisms of postgres. Best part is that I've got a solid foundation for bot command handling as a result of the testing. Blog post soon on the details of playing 'ping-pong' with logbot and postgres as the paddle
asciilifeform: buncha phuctor output, btw, in #asciilifeform , and prolly will through the night
asciilifeform: or hm, possibly this is it.
asciilifeform: ( from parcel1 )
asciilifeform: pretty modest haul.
asciilifeform: ( for 26707136 mods )
a111: Logged on 2017-02-07 18:49 asciilifeform: hanbot: conceivably, a sabotaged generator that yields 2**32 possible keys, will result in ~0 natural collisions, and remain invisible to phuctor unless i first find it via some other means.
a111: Logged on 2016-09-02 18:26 asciilifeform: well other working hypothesis is that it is ~not~ tiny fraction, but a tip of the 'birthday theorem' iceberg.
asciilifeform: the huawei crapola, for instance, i'm quite convinced at this point would be 100% breakable if anybody got hold of a unit and forced it to generate keys 24/7 for a coupla months
asciilifeform: ( on 'clean' boot naturally , each time )
asciilifeform: ditto mikrotik etc
BingoBoingo: A lol in the spew: "deedbot> http://phuctor.nosuchlabs.com/gpgkey/23B2173C2FF1A9C43007D526720EA2B9EC1CB4AC21503429ACFBA1DA022517B3 << Recent Phuctorings. - Phuctored: 3 divides RSA Moduli belonging to 'James Bottomley <jejb@kernel.org>; James Bottomley <JBottomley@Odin.com>; James Bottomley <JBottomley@Parallels.com>; James Bottomley <James.Bottomley@HansenPartnership.com>; '"
asciilifeform: BingoBoingo: old
asciilifeform: see #a log
BingoBoingo: I'll have to start reading the phuctor chunks from the bottom up
asciilifeform: was optimizing rss crapper so that it doesn't idiotically do whole cpu/hour of work each time cache expires, evidently introduced a sad where a new find of $factor triggers old mods as report
asciilifeform: ( old mods containing same )
asciilifeform: the actual count of new factors found, was 28 .
asciilifeform: next we will see what , if anything, happens when we run on a ~random 3%~ subset of whole collection.
asciilifeform: ok, output is happening. up side : there's output, that there was not before. down side : so far all of it is small change crapolade. i'ma really have to do a mod-by-mod walk and gcd(mod, 8192bit primorial from oldthread) . and henceforth on submission. cuz this is lame.
asciilifeform: 'this is divisible by 3' should not be news.
asciilifeform: and nobody fucking cares that 'there are 2 moduli divisible by 1547665'. aint prime.
asciilifeform: what i'ma end up doing, is to factor all of the small composites, remove'em from db, enter the factors instead, and henceforth never enter composites into factor table.
asciilifeform: ( and omit 'new small factor for known rubbish mod' from rss, nobody particularly needs to 'see the rubble bounce' )
← 2018-08-11 | 2018-08-13 →