Forum logs for 13 Feb 2018

Monday, 16 March, Year 12 d.Tr. | Author:
hanbot: speaking of that wednesday, it's rather incredible how little discussion there is on the matter, especially considering http://trilema.com/forum-logs-for-10-feb-2018#2408406 [01:48]
a111: Logged on 2018-02-10 14:58 mircea_popescu: nevertheless, reality can only be elided at significant personal sacrifice and over short intervals. so, to repeat it : there is no future for the republic without an isp. [01:48]
hanbot: i suppose the amount of time it's taken to get to this point while the same death knell was present makes it easy to imagine things could just continue as they are, but i tell you that seems pretty dangerous. [01:50]
hanbot: and i'd think salvaging bb's mission closer to the goal --any point closer than zero--, would be just about top priority. hence figure it out, and well before wednesday morning comes around. [01:51]
diana_coman: I admit I don't really get what the sticking point is there in asciilifeform's calculations: the potential tax on shipped equipment? the cost of BingoBoingo ? the cost of racks? not enough sure revenue (how the fuck)? what? [02:56]
ave1: Hanbot and Diana_coman, I thought so too. The past couple of days has cost me a few nights sleep. My thoughts here: http://ave1.org/2018/from-the-proverbial-peanut-gallery/ [05:58]
shinohai: Gah, the logs I wake to of late. [06:53]
BingoBoingo: <mircea_popescu> BingoBoingo soo, you're buying your ticket wednesday or what was it ? << I was planning on picking it up Thursday if there has been no interest through Wednesday in channel on a takeover mission for this venture. I can extend the timeline a bit, if it is prudent to do so. [07:11]
BingoBoingo: <hanbot> and i'd think salvaging bb's mission closer to the goal --any point closer than zero--, would be just about top priority. hence figure it out, and well before wednesday morning comes around. << I can only hope other people are doing math and weighting costs versus warming up their other ISP plans they may have put on hold. [07:15]
diana_coman: quite some very good points there in my opinion ave1 [09:10]
diana_coman: BingoBoingo, have you read that? [09:12]
BingoBoingo: I am re-reading it a few times. Certainly it is helpful in a way that the gabriel_laddel-ism http://www.thedrinkingrecord.com/2018/02/09/a-long-descent-into-the-cargo-cult/#comment-114684 is not [09:23]
TomServo: I'd very much like to see this endeavor succeed and would gladly offer assistance. I don't want to get out line if this discussion is for L1. [09:26]
BingoBoingo: TomServo: At this point the discussion looks like it would be a way to get into L1 [09:28]
lobbes: http://btcbase.org/log/2018-02-13#1783471 << I can't speak for others, but talk (and opinions) are cheap! Only one willing to consider shouldering the post-mp burden/risk thus far has been asciilifeform. [09:32]
a111: Logged on 2018-02-13 06:48 hanbot: speaking of that wednesday, it's rather incredible how little discussion there is on the matter, especially considering http://trilema.com/forum-logs-for-10-feb-2018#2408406 [09:32]
lobbes: I'm a regular poorfag, but fuck, what is it FOOD for BingoBoingo that is standing in the way in alf's calcs? If that's what it takes I can throw a months worth of food in btc to bb as a one-time gift.  (Though I think issue was longterm coin-to-food conveyor) [09:33]
lobbes: Bbl, mines [09:33]
TomServo: For clarification, is the S.NSA gear verboten? Are phf's flight/courier arrangements still on the table? [09:37]
BingoBoingo: lobbes: If necessary BingoBoingo can do some ugly things to keep fed. http://btcbase.org/log/2018-02-11#1783247 Follows from BingoBoingo being a sort of religious fanatic ideologically aligned with the Republic, even if my inexperience and inability to effectively self direct are an impediment lords taking over this venture would need to compensate for. [09:39]
a111: Logged on 2018-02-11 14:00 mircea_popescu: BingoBoingo what your lack of experience isn't informing you is that you have a very significant advantage. your ~7k per annum cost compares very VERY favorable to what empire pays to hire a walmart drone (25k or so base + 20k or so in overload). the walmart drone costs 6x what you do, and i daresay they do not do 1/6th of what you do. just because you came out short of lordship in the field does not now mean the baseline of [09:39]
BingoBoingo: That being said for my health I would like a proposal that includes the possibility of a BingoBoingo exit date away from cargo cult Alcoholics Anonymous [09:44]
shinohai: I've done ugly things to stay fed .... worked with fatty camgirls, etc. But echoing http://trilema.com/forum-logs-for-10-feb-2018#2408406 none of my ventures have any future without isp. [09:44]
a111: Logged on 2018-02-10 14:58 mircea_popescu: nevertheless, reality can only be elided at significant personal sacrifice and over short intervals. so, to repeat it : there is no future for the republic without an isp. [09:44]
shinohai: Aside from Qntra I have no reliable income, so that brings all to screeching halt here. [09:45]
* BingoBoingo to shut up for a bit and compile as much of the ave1 requested material together into a blog and explore the possibility of making arrangements for a Dentist to follow the BingoBoingo camp to old country. [09:47]
mod6: I don't want to through good money after bad here. [10:14]
mod6: So, indeed, I think we all want to see the ISP idea work. But after this entire snafu, I'm still confused as to what went wrong here. The foundation's plan here was in motion to bring machines there directly. Because of the ugly possible VAT or dissappearing of boxes themselves. [10:15]
mod6: If part of the plan was to build machines on demand for customers, it sounds like we're in the wrong locale for that. [10:16]
mod6: I'm gonna look at some things, will report back later. [10:23]
deedbot: http://www.thedrinkingrecord.com/2018/02/13/twelve-runs-of-speedtest-cli-from-the-qntra-server/ << Bingo Blog - Twelve Runs Of speedtest-cli From The Qntra Server [10:32]
phf: http://btcbase.org/log/2018-02-13#1783489 << S.NSA gear is not verboten, but my absolute deadline to leave expires in less than an hour [10:40]
a111: Logged on 2018-02-13 14:37 TomServo: For clarification, is the S.NSA gear verboten? Are phf's flight/courier arrangements still on the table? [10:40]
mod6: On it. [10:40]
TomServo: Sucks, I think getting that gear there was a critical piece. [10:47]
mod6: Yeah. Stand-by. [10:48]
lobbes: << this is the key question needing resolving, imo. "Is it worth it to continue throwing good money into orc-hole?". After BB's deep descent into the cargo cult looking for chassis n' such, it is certainly suggestive of more "tree-forts everywhere" ala http://btcbase.org/log/2018-02-08#1782956 [10:52]
a111: Logged on 2018-02-08 23:20 mircea_popescu: http://btcbase.org/log/2018-02-08#1782932 << the only point to my eyes is that someone fucked in the head enough to run something they call "datacenter" without one ton of spare racks laying about is not distinguishable from three twelve year old white boys in a "tree fort" running a "porn business" through taking complicatedly filtered ipad shots of their own palms. [10:52]
lobbes: But for now I'll shut up and leave room for my betters to work out that question. Back to saeculum mines, and then archive-o-tronics [10:52]
ben_vulpes: http://btcbase.org/log/2018-02-13#1783475 << this is not that clear to me. what i've penciled out is that there are some ~1.8kusd monthly of rack rental buy side floating around, and mircea_popescu's plan as i understand it was to build and rent vpsen to get to breakeven, which still needs someone in teh forums hawking vps [10:56]
a111: Logged on 2018-02-13 07:56 diana_coman: I admit I don't really get what the sticking point is there in asciilifeform's calculations: the potential tax on shipped equipment? the cost of BingoBoingo ? the cost of racks? not enough sure revenue (how the fuck)? what? [10:56]
shinohai: Can't sell what doesn't exist, neh? [10:56]
ben_vulpes: shinohai: elaborate? [10:57]
mod6: BingoBoingo: how much revenue was going to be pulled down by non-vps customers per month? The foundation had said "we want 2U", so that accounts for us. [10:58]
shinohai: If there are currently no vps boxxen to sell, cannot exactly sell anything besides promisetronic [10:59]
mod6: (trying to establish how much VPS needs to be sold to make up the difference just to break even on bills) [10:59]
ben_vulpes: shinohai: obviousl. [10:59]
phf: TomServo: it's critical only in a sense that the immediate opportunity was squandered, but right now most important is to provide bb with enough fiat to stay afloat, next requirement is to provide bb with enough fiat to pay dc fees, because those two are the ones that are going to shut down the operation [11:00]
ben_vulpes: i'm guessing the initial subscription of just rack space comes to around 1.7k/mo waving hands at a vps setup "special get us to breakeven" deal slicing up one of the mega boxes asciilifeform described implies a breakeven at some 57 monthly rentals at 20 usd/mo (which buys 4.5GB of ram, notbad.jpg in the grand scheme of things, but the tiny straw of bandwidth weighs against this) [11:05]
ave1: BingoBoingo: read carefully these are not requests but suggestions, I even left some out. Also, at this point you still have a partner and I do not know your arrangements, so check before publishing what might be important / secret business operational data. [11:08]
BingoBoingo: ben_vulpes: Another higher margin activity which can be marketed is good old fashioned shared hosting. [11:08]
BingoBoingo: ave1: At this point the journey to here is documented in existing blog posts, there is no harm in collecting that data into a single larger post for the convenience not only of people who might be considering taking on this particular effort, but for those who may want to undertake their own as well. [11:15]
ben_vulpes: neglecting shared hosting, and again waving hands at a sales rate on vps slicen of 10/mo and zero churn etc, thing could hit breakeven on the rack alone sometime in august [11:15]
phf: ben_vulpes: it's $1423 for 20/200 pipe and $900 for 2.5KW rack (his pipe is effectively 18/183) [11:39]
ben_vulpes: phf: yeah i rolled IVA into that number, is that wrong for some reason? [11:40]
phf: ben_vulpes: didn't you lose a $1000 then, it works out to 2741/mo [11:46]
ben_vulpes: i pulled 2834.06/mo from http://www.thedrinkingrecord.com/2018/02/03/pricing-information-fun-with-numbers/ , estimated some 1.7k in rental subscriptions from l1/2 [11:48]
phf: oh oh [11:49]
phf: i missed the "estimated from l1/2" bit [11:49]
ben_vulpes: yeah, it was a bit further up, "~1.8k monthly of rack rental buy side floating around" [11:50]
asciilifeform: !!up alienatu [12:07]
deedbot: alienatu voiced for 30 minutes. [12:07]
diana_coman: phf I think what is actually shutting down the operation is a. lack of someone capable & willing to take it in hand i.e. provide very detailed instructions to BingoBoingo b. fuzziness as to what operation there really is atm + what operation can actually be had at the location [12:35]
asciilifeform: fwiw asciilifeform is able to supply (a) . problem is (b) [12:36]
diana_coman: asciilifeform, so ask BingoBoingo exactly for what is needed to un-fuzz that's what all the detailed instructions is all about [12:36]
asciilifeform: diana_coman: there's two problems , 'technical' and 'business'. technical is solvable. the part that bites, is the 'business' [12:38]
diana_coman: so a.1 is technical capable person a.2 business capable person [12:39]
diana_coman: and apparently you can supply a.1 missing a.2 [12:39]
asciilifeform: it is not clear to me that 'vps marketed to heathens, for btc' is a viable model. but then again i am not business wizard, possibly someone with mp-level brain could have done this [12:39]
asciilifeform: as far as i can see, there is a shortage of heathens who have any btc to spend to begin with. (much less who want to spend it on a vps) [12:40]
diana_coman: out of curiosity why marketed to heathens? [12:40]
asciilifeform: because l1 doesn't use , afaik, vps [12:40]
asciilifeform: ( am i mistaken ? ) [12:40]
mod6: i use a few for various things. [12:41]
diana_coman: iirc there were some vps ran by l1, what as sandboxes if nothing else [12:41]
asciilifeform: l1 appears good for a certain amount of raw verticalspace rental. but not the 0.5btc's worth (at current usd rate) required to keep op in the black [12:41]
diana_coman: asciilifeform, so maybe make the plan + offer and see what feedback you get on it? [12:41]
diana_coman: fwiw I am certainly interested in being a customer of TMSR ISP but I've been waiting and waiting to see concrete offers being a bit on the other side of the world doesn't help much re moving iron there either but presumably at some point I can even get that sorted in some way if there is some ISP to get sorted to, of course [12:44]
diana_coman: but I suppose asciilifeform's obv is correct in that a.2 is missing still [12:45]
asciilifeform: diana_coman: dare i ask if you have a theory of how a bisp could stay in the black without taking heathen coin ? [12:45]
diana_coman: I think it can sell enough to l1(+l2 perhaps) but I haven't run the numbers etc (because I'm already juggling lots as it is) [12:47]
ben_vulpes: diana_coman: i've only tallied up ~1.7k/mo in l1 space rental [12:51]
diana_coman: in other words this can be put as: tmsr's l1 atm is not big enough yet to support an isp of its own [12:54]
asciilifeform: diana_coman: could be enuff. depending on what people are willing to pay. [12:55]
ben_vulpes: aaand that assumes that the 5 slots mircea_popescu wants don't overlap with s.nsa [12:55]
diana_coman: asciilifeform, the point is to sell more not pricier, neh? so: depending on how much people are willing to USE the isp [12:56]
asciilifeform: diana_coman: i meant, in sum [12:56]
asciilifeform: ( if we have 5 people who each want 3 boxes, is just as good as if we have 3 who each want 5 ) [12:56]
diana_coman: yes, how big is not a head count strictly of course put it as how powerful if that is clearer [12:58]
asciilifeform: right [13:05]
asciilifeform: the rough measure appears to be, if l1 is good for 0.5 btc / mo. ( at the current rate) -- bisp runs. if rate loses a 0 -- it turns into 5. if grows a 0 -- becomes 0.05 . [13:10]
asciilifeform: ( if someone believes this equation to contain a mistake -- i would like to hear about it asap ) [13:12]
ben_vulpes: i'll buy it [13:13]
asciilifeform: ben_vulpes: which it [13:14]
ben_vulpes: the equation, it pencils out over here as well. [13:14]
asciilifeform: aa. [13:14]
mod6: yeah, this seems to make sense to me. [13:15]
mod6: BingoBoingo: Are you there? [13:22]
BingoBoingo: mod6: I am baking blog post summarizing where this thing is. [13:23]
mod6: We have some questions for you. [13:23]
BingoBoingo: Well, present the questions? [13:24]
lobbes: http://btcbase.org/log/2018-02-13#1783544 << fwiw I will certainly purchase/use a bbisp vps as well. [13:24]
a111: Logged on 2018-02-13 17:41 mod6: i use a few for various things. [13:24]
deedbot: http://www.thedrinkingrecord.com/2018/02/13/a-summary-of-bbisp-activity/ << Bingo Blog - A Summary Of BBisp Activity [13:27]
mod6: !~later tell jurov Please to ping me as soon as you get a chance. [14:27]
mod6: damnit [14:28]
mod6: !~ later tell jurov Please to ping me as soon as you get a chance. [14:28]
mod6: ok maybe bot is down. [14:29]
mod6: jurov: please ping me when you're around. [14:29]
lobbes: !Qlater tell jurov plox to ping mod6 once ya get the chance [14:31]
lobbesbot: lobbes: The operation succeeded. [14:31]
lobbes: Redundancy ftw! [14:31]
mod6: oh, thanks, didn't even know. good work lobbes. [14:31]
deedbot: http://trilema.com/2018/experimental-results/ << Trilema - Experimental results [16:58]
diana_coman: in other issues: rfc on oaep for tmsr, see http://trilema.com/2017/tmsr-rsa-spec-extremely-early-draft/#comment-124596 [17:06]
asciilifeform: diana_coman: that was the openproblem that produced the thread which ultimately turned into mp's-hash etc [17:06]
shinohai: http://btcbase.org/log/2018-02-13#1783582 <<< lobbes this may be something you and I might discuss you overtaking, since I likely am not going to be able to bring back jhvh1 with current isp woes, etc. [17:16]
a111: Logged on 2018-02-13 19:31 lobbes: Redundancy ftw! [17:16]
ben_vulpes: somehow lobbes trinque phf asciilifeform can run bots without the republican isp but all of a sudden shinohai cannot? [17:18]
asciilifeform: ben_vulpes: not to mention eulora, trilema, somehow seem to run.. [17:22]
ben_vulpes: aha [17:22]
asciilifeform: ( since mircea_popescu seems to be sleeping i'ma have to run the emulator : 'it's because they are part of my Adult World, not this stinking kindergarten' ) [17:22]
shinohai: eulora and trilema have funding that shinohai does not. Might move it to home box, remains to be seen. [17:22]
mod6: re rota: http://btcbase.org/log/2013-01-21#-156893 [17:26]
a111: Logged on 2013-01-21 21:12 mod6: wat [17:26]
mod6: Alright, the Bitcoin Foundation has been discussing all day about using funds to support the bbisp. We are ready to make that commitment. [18:19]
mod6: Arrangements are being made currently. [18:20]
BingoBoingo: Is this a takeover? [18:20]
BingoBoingo: Please say yes [18:20]
mod6: The aim is to supply you with funds to continue operations. And to relieve you of debt to Mr. Popescu. [18:21]
mod6: We propose to pay him outright, relieving him of obligations to the venture. [18:22]
mod6: We hope, to retain Mr. Popescu as a customer, as well as ``Minigame'', S.MG, so known. [18:22]
asciilifeform: somebody plox wake up mircea_popescu [18:23]
asciilifeform: ( let's to hope he has a bell that rings when called... asciilifeform at one time did ) [18:23]
mod6: Someone also wake up jurov. [18:23]
mod6: Furthermore, we will bring management and retain BingoBoingo for his services as 'Remote-Hands'. [18:31]
mod6: Happy Valentines Day, TMSR~. <3 [19:00]
deedbot: http://www.loper-os.org/?p=2238 << Loper OS - Finite Field Arithmetic. Chapter 10: Introducing Karatsubas Multiplication. [21:05]
asciilifeform: phf: plox to snarf patches, ty [21:06]
mod6: hey hey hey, lbj [21:09]
asciilifeform: as always, q's, barfola, corrections of obvious flipped signs, etc. welcome. [21:15]
mod6: Thanks for posting. [21:15]
asciilifeform: fwiw asciilifeform posts these 'from cause, not purpose' (tm)(r) [21:17]
asciilifeform: who knows, 20y from nao , a different asciilifeform will meet a different mircea_popescu and they will look for a sane numeric lib and ~find~ one, instead of sitting for 4y without... [21:17]
asciilifeform: ( ... or not. 'not my paygrade' (tm) ) [21:18]
asciilifeform: >> http://btcbase.org/patches/ffa_ch10_karatsuba << ty phf ! [21:46]
mod6: I can confirm that the seal verifies and that pressing up through chatper 10 worked. [21:56]
asciilifeform: mod6: now try [21:58]
mod6: did you change something? [21:58]
asciilifeform: time echo "??\`\`*[print 0x]##[ == 0x]#[ * 0x]#" | ./bin/ffa_calc 1048576 32 /dev/urandom | tr -d '\n' | python [21:58]
mod6: oh. ok sure. [21:58]
asciilifeform: ...then substitute /dev/zero for /dev/urandom, and same [21:59]
asciilifeform: ( to do this very rigorously, oughta buffer the 'random' reading from /dev/zero is unsurprisingly slightly faster ) [22:00]
asciilifeform: but i think errybody gets the idea. [22:00]
mod6: http://p.bvulpes.com/pastes/K2jR0/?raw=true << for the command with /dev/urandom [22:01]
mod6: will try with /dev/zero [22:01]
asciilifeform: ahahahahawaaat [22:02]
asciilifeform: dafuq [22:02]
asciilifeform: do you not have tr ? [22:02]
asciilifeform: !!up pehbot [22:02]
deedbot: pehbot voiced for 30 minutes. [22:02]
mod6: hmmmod6@localhost ~/ada/alf/ffa/ch10/ffa/ffacalc $ command -v tr [22:02]
mod6: /usr/bin/tr [22:02]
asciilifeform: !A ??``*[print 0x]##[ == 0x]#[ * 0x]# [22:03]
pehbot: asciilifeform: print 0x3E5D0A1494714608C690023711EFA9F54CF89F9B8B6868E1D67349192693CFCE40FCAF5754AA88528ADEB8ED715E6E8AE88B42D309FAB89CA70AB0F7C302C4C9 == 0x5B76C0C5A47092525D9522B8B6BEF9B4DE1EA7F4C7F89A3275C15B4B779BE2E3 * 0xAE8CDAA17D3E6F67085DE25998BA005E2ADF86FB1A57BFED68BE051B47A18D63 [22:03]
asciilifeform: ^ 256bit version [22:03]
asciilifeform: !A ??``*[print 0x]##[ == 0x]#[ * 0x]# [22:03]
pehbot: asciilifeform: print 0x698CE911132F8468C4E194B8DE368D399B8AE8C9DCF00FC13622C69C77D352F353E6B7BF5B55BF3DA41638E1205BF3BB08656A1A8A8E6E44C6BFC7305DE792F0 == 0x97E739A43C6FDC54D30868703F7F3352A0E6B8145D73DE069DD0D5DAE15D40B0 * 0xB1E1E52D89EE143D129759B6ECDA2B57FB3A51F52A7903E840BDF7DDFC9361ED [22:03]
asciilifeform: etc. [22:03]
asciilifeform: when you strip the newlines, this kompyootes [22:03]
asciilifeform: aaaahahahahah you have fucking python3 dontcha mod6 . [22:03]
asciilifeform: they CHANGED the syntax for 'print'. [22:03]
mod6: with /dev/zero was 0x00000... * 0x0000... huge number. : http://p.bvulpes.com/pastes/MuO8r/?raw=true [22:03]
mod6: lemme check [22:04]
mod6: ah, aha: Python 3.4.5 [22:04]
asciilifeform: bahahahha [22:04]
mod6: kill it with fire? [22:04]
asciilifeform: or put in the parens it wants, lol [22:04]
asciilifeform: but certainly won't work as shown. [22:04]
mod6: ah. alrighty. [22:04]
asciilifeform: !A ??``*[print (0x]##[ == 0x]#[ * 0x]#[)] [22:05]
pehbot: asciilifeform: print (0x8EEE68CAA9C397919CECDBD05C4302B99F58EAB4A817CEA09AAA5CDC4BA3B6E0F8F9E79110C128FE5855DFAED823D747291502C911E8822424FF68EA61E163F0 == 0x9523A93AE0C7CA410E9D96BA917CC81F6AE261D3E66761697EEA50D42887C714 * 0xF5580622DB331381B3CCDFFB01647F4CBEE85A173D5CC98939237F4A02E7F0CC) [22:05]
asciilifeform: ^ maybe like-so [22:05]
mod6: ooh ok [22:05]
* asciilifeform tried locally, seems to satisfy p3 [22:05]
mod6: Like so? http://p.bvulpes.com/pastes/RLrQZ/?raw=true [22:06]
asciilifeform: exactly like-so. [22:06]
mod6: werd. nice :] [22:06]
asciilifeform: aaand it looks like we have a 'winner' : [22:06]
asciilifeform: ... or not. gotta compare /dev/zero and file of 0xFF . [22:07]
asciilifeform: but know that 'iron mul' has been re-introduced (will be made selectable soon enuff) and you can use ch10 ffa to diagnose 'evil' cpu [22:07]
asciilifeform: that have variable-timed MUL. [22:07]
mod6: ahh. [22:08]
asciilifeform: mod6: your timing result is not yet proof that you have such a box gotta compare a static 'random' consisting of 00000.... and one of FFFF ... [22:08]
asciilifeform: ( you can substitute a file for the /dev/whateverthefuck ) [22:08]
asciilifeform: for the given example, gotta make sure that it has at least 2MB of mass. [22:09]
asciilifeform: (megabit, not megabyte) [22:09]
asciilifeform: this is a deterministic test. ffacalc is made deliberately to permit one. [22:09]
mod6: oh, i see. lemme see what i can drum up here... [22:09]
asciilifeform: a pro seekrit : you can pipe ffacalc output to base64 -d [22:11]
asciilifeform: or grrr nm [22:12]
asciilifeform: not base64, but [22:12]
asciilifeform: xxd -r -p [22:14]
asciilifeform: so lessay you want a Mbit of 0xFF..... [22:14]
asciilifeform: echo ".0~#" | ./bin/ffa_calc 1048576 32 | xxd -r -p > lotsaones.bin [22:15]
asciilifeform: is your ticket. [22:15]
asciilifeform: y'know, [22:15]
asciilifeform: !A ".0~#" [22:15]
pehbot: asciilifeform: EGGOG: Pos: 0: Stack Underflow! [22:15]
asciilifeform: lolahaaha [22:15]
asciilifeform: !A .0~# [22:15]
pehbot: asciilifeform: FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF [22:15]
asciilifeform: etc. [22:15]
asciilifeform: so you get however many you want, you got it by simply using ffacalc to compute the desired integer. [22:16]
asciilifeform: after that you can e.g. time echo "??\`\`*[print 0x]##[ == 0x]#[ * 0x]#" | ./bin/ffa_calc 1048576 32 lotsaones.bin | tr -d '\n' | python [22:18]
mod6: ooh, ok lemme give that a shot. [22:18]
asciilifeform: ( note that you do not need the 0 in .0 , i put it in for readability ) [22:19]
asciilifeform: !A .~# [22:19]
pehbot: asciilifeform: FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF [22:19]
mod6: aha, sure. [22:19]
mod6: http://p.bvulpes.com/pastes/IKJAK/?raw=true << ok with feeding it in with '<' it seems to do somthing here... had some trouble trying to read it directly said 'Could not read from RNG!' [22:31]
asciilifeform: mod6: you need 2Mb, recall [22:34]
asciilifeform: i oughta have reminded. [22:35]
mod6: yah, i thought they might be a bit small. lemme see what I can do. [22:35]
asciilifeform: cuz you have two invokes, recall, "??" [22:35]
mod6: yup, gotcha [22:35]
asciilifeform: not 'a bit' but fully 2x toosmall [22:35]
asciilifeform: aaha [22:35]
mod6: how about this? http://p.bvulpes.com/pastes/1dfHJ/?raw=true [22:40]
asciilifeform: right, but ideally you want a few thou of these, and a correlation calc [22:41]
asciilifeform: (or at least basic stats) [22:41]
asciilifeform: and replace the python with > /dev/null [22:41]
asciilifeform: cuz python is (mega-unsurprise) variably-mul-ing [22:41]
mod6: ah, sure. [22:42]
asciilifeform: i left this as exercise but if mod6's nerves are as tight as asciilifeform's , is forgivable , lol [22:42]
mod6: heheh. so you're saying do this: http://p.bvulpes.com/pastes/1VeEL/?raw=true : a few thosand times and create a graph of the differences in the timings [22:43]
asciilifeform: can skip the tr lol [22:43]
asciilifeform: but yes [22:44]
mod6: oh my bad [22:44]
asciilifeform: can even skip the #'s, you don't need the output foranything [22:44]
asciilifeform: but overall yes. [22:44]
mod6: so for those following along, this: http://p.bvulpes.com/pastes/OHXDh/?raw=true [22:45]
asciilifeform: also if you want to experiment with even bigger numberz, turn the 32 ( that i habitually use as stack height ) into the minimal necessary 4 for this experiment [22:46]
asciilifeform: that way you will not exhaust memory (gcc's default stack) when you use genuinely massive ffawidth [22:46]
asciilifeform: ( if this does happen, it happens during warmup , and is harmless ) [22:47]
asciilifeform: e.g. 4194304 ( 4 megabit ) takes ~3s on my box here. [22:48]
asciilifeform: ( go and try same with ch9 ! ) [22:49]
asciilifeform: this of course is not a surprise to folx who actually read. but still fun. [22:49]
* asciilifeform bbl,food [22:50]
asciilifeform: in other oddities: http://www.riotsys.com [23:29]
asciilifeform: ( before anyone asks -- i have nfi who this is ) [23:29]
asciilifeform: 'RS-16 has a single-tasking "operating system" whose source fits on a few pages of paper and is easily understandable by anyone who speaks C. The operating system only implements the functions that you need for editing text, crypto and transferring files. It lives in ROM and cannot run programs from disk, eliminating a range of attack vectors.' [23:29]
asciilifeform: but for some reason also 'Optional: Bluetooth Module RN42-XV' ... [23:30]
asciilifeform: ( author also of e.g. http://interim-os.com and a few other interesting nuggets ) [23:31]
asciilifeform: in lulz*2, the last link in 1st para of 'rs-16' goes to a ... 'This video is no longer available because the YouTube account associated with this video has been terminated.' [23:32]
mod6: asciilifeform: http://p.bvulpes.com/pastes/kkTSP/?raw=true [23:48]
asciilifeform: oh hey hey hey [23:48]
asciilifeform: mod6: leave it to run overnight for coupla million shots [23:50]
asciilifeform: (each) [23:50]
mod6: sure. will do. [23:50]
asciilifeform: then post the magic figures, and include what type of iron was used. [23:50]
asciilifeform: ^ and for anybody else who wants to try. [23:50]
asciilifeform: i'ma set some up later this wk meself. [23:50]
mod6: cool :] [23:50]
asciilifeform: yay. [23:51]
asciilifeform: will be lulzy incidentally if the diff goes the other way when you move the files somewhere else on disk and rename'em etc [23:52]
asciilifeform: ( i.e. if it's due to where the inodes are, or the like ) [23:52]
asciilifeform: anyffinspossible. which is why experiment. [23:52]
Category: Logs
Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.
Add your cents! »
    If this is your first comment, it will wait to be approved. This usually takes a few hours. Subsequent comments are not delayed.