Hide Idle (>14 d.) Chans
← 2018-05-25 | 2018-05-27 → ↓
lobbesbot: jurov: Sent 6 days, 14 hours, and 59 minutes ago: <BingoBoingo> http://p.bvulpes.com/pastes/QodBe/?raw=true
spyked: to clear up the confusion in http://btcbase.org/log/2018-05-25#1818920 (and maybe for some bonus lulz): somehow last night my eyes completely overlooked the last part in http://btcbase.org/log/2018-05-25#1818894 (yes, after reading *twice*), leading to a very wtf contextless spyked upon ingesting http://btcbase.org/log/2018-05-25#1818901 . /me still has much practice to do to follow discussions in real-time, most of my reading happens
a111: Logged on 2018-05-25 21:31 mircea_popescu: what metaphor! it's an metasyntactic variable, you never encountered before ?!
a111: Logged on 2018-05-25 21:11 mircea_popescu: well, did you protest that anywhere ? preferably as "hey $x, im trying to use ircbot to make rssbot, but you've not got voice going ?!"
a111: Logged on 2018-05-25 21:16 mircea_popescu: nevermind that part for now. do you have a value for $x in "hey $x".
spyked: in the logs.
spyked: also, re. http://btcbase.org/log/2018-05-25#1819054 my intention is to implement the http://btcbase.org/log/2018-05-20#1815920 approach, where: a. user obtains OTPs manually by prodding deedbot with !!ups, decrypts them and feeds them to trilemabot; b. trilemabot maintains a list of OTPs and uses the next one in the list when it needs to self-voice (i.e. after it /joins #trilema). c. user must ensure that the OTP list is not empty,
a111: Logged on 2018-05-25 22:46 spyked: mircea_popescu, no, not yet. it's next on the todo list.
a111: Logged on 2018-05-20 14:17 trinque: spyked: one could make many voicing OTPs, and put them in a hopper for the bot.
spyked: otherwise bot won't self-voice.
spyked: I'll leave this open for discussion for a coupla days or so. meanwhile any comments from other lisp bot operators ( ben_vulpes, lobbes, phf et al.) are muchly appreciated.
mircea_popescu: that's what second pass is for, huh.
spyked: mircea_popescu, word.
mircea_popescu: re the voicing approach : the above is fine. the alternative, where you keep a dedicated key on the bot, is also fine. the damage it can do is very limited (what, spend money it doesn't have ? rate people ?) and the upside is that you get to find out about possible attacks. neither method is wrong or anything.
a111: Logged on 2018-05-25 22:47 trinque: I suspect logbot is terribly named. it's more aptly an IRC-pg bridge.
mircea_popescu: are we specifiing a gns standard for these ?
mircea_popescu: and in other lulz, anil dash of http://trilema.com/2014/the-nigger-a-practical-exercise/ still exists!
mircea_popescu: "Anil Dash (/ɑːˈniːl ˈdæʃ/; born September 5, 1975) is an American blogger, entrepreneur, and technologist. Previously an independent technology consultant and new media developer for the Village Voice"
mircea_popescu: the things imbecile journahos "become" in the empire! TECHNOLOGISTS!!
mircea_popescu: He was the director of Expert Labs, a "Government 2.0 initiative that aims to connect United States government projects with citizens who want to become more involved in the political discussion"
mircea_popescu: "Anil Dash‏Verified account @anildash 30m30 minutes ago
mircea_popescu: But just a word to journalists aghast at the idea of naming a source who wanted to be off the record: telling the truth and exposing the President’s lies is actually more important than j-school decorum. You’re being played."
mircea_popescu: "I cofounded Activate, the strategy consultancy that guides the leaders of the biggest media, technology and entertainment companies in developing their strategies for growth" << consider the politruk-icity. so, "the biggest" have absolutely nothing to say. they are so inept at creating their own fucking strategies, they'll just hire the political commissar to give them a sheet of paper that's certified by the office of polit
mircea_popescu: ruks and therefore won't get them into political trouble.
mircea_popescu: i'll be very keen to hear someone explain the difference between usg and kpss.
deedbot: http://trilema.com/2018/the-v-questionarium-answerarium-2018-edition/ << Trilema - The V questionarium & answerarium, 2018 edition
mircea_popescu: meanwhile in other alt-science and cvasi-technology, https://www.slaveplanet.net/
diana_coman: has anyone installed php successfully on a rockchip? it fails to build on mine with a lot of "error: dereferencing pointer to incomplete type" and a search for this sort of thing seems to suggest it might have to do with libxml version; anyone ran into this and sorted it out already?
diana_coman: for completeness: it's about php 5.6.34 with gcc 4.9.4 ; libxml2 is 2.9.7
mircea_popescu: meanwhile in finance and banking, http://bankofbaroda.co.bw/
BingoBoingo: I like how the Botswana Bank's international outreach is to India
BingoBoingo: And nowhere else
lobbes: diana_coman: I've been on a quest for the last few days trying to get a LAMP running on my rockchip. I was getting barfs on emerging php-5.6 but managed to at least get php 7.2.4 installed
lobbes: did your php 5.6 barfs look like mine too? >> http://p.bvulpes.com/pastes/KNEJ4/?raw=true
lobbes: the "error: dereferencing pointer to incomplete type" sounds like we have the same issue. Either way, I've no solution yet myself. I figure if I can get mp-wp going with php 7.2 I guess I'll be satisfied for now
mircea_popescu: BingoBoingo, it's really owned by the indian govt.
esthlos: hey mircea_popescu, you used to study physics, right? any opinion on mach's principle?
mircea_popescu: i dunno ; grav theory is particularily opaque.
esthlos: certainly
mircea_popescu: anyway, celestial reference is practically useful, and the question "but why should it be" seems ill posed.
mircea_popescu: in any case i find the web 2.0 ish approach intellectually repugnant ; if i'm to explain why one "feels his arms pulled" when spinning under the moon, perhaps i'll be next held to explain why the same one doesn't "feel" the normal pressure in his bladder, or the heliocentric nature of the solar system in his horse's clavicle. let einstein, ziggler, marx, godin and friends answer this nonsense ; anil dash stands ready to angri
mircea_popescu: ly tweet their productions.
mircea_popescu: in short, it's not only ill posed, but rather seems ill posed a certain, familiar way.
esthlos: most admirable physicists I've encountered feel ~ the same. still, idea secudes me from time to time
esthlos: trinque: any time to look at the new vtron? the currently outstanding issues I'm aware of are 1. need to reintroduce a defpackage; 2. weirdness with ccl and building the gpg keychain
mircea_popescu: i suppose the proper term is "new age". funny how that principal designation of pantsuit idiocy during the 80s disappeared once the soviets gave in.
esthlos: funny, they conjure literal beings of pure energy
esthlos: also mircea_popescu , I'm curious what your formula for learning new languages is; or to be more precise, what your advice is to an esltard trying to not be so tarded
mircea_popescu: get a girlfriend.
esthlos: i see
mircea_popescu: not just theory, either. afaik it's how BingoBoingo learned spanish.
esthlos: now to find a deutsche Frau in pantsuit central
trinque: esthlos: yep, I just got to the keyboard. iirc there was something wrong with ccl in regards to keychain path?
trinque fires it up
esthlos: well, try this out: load in the file from a ccl repl, run make-wot, back out of the failure, and run wot again
mircea_popescu: no, you go there.
esthlos: ah nevermind, I thought I remembered this working
esthlos: see the different error messages? I find it odd
trinque: esthlos: I wager neither of us knows how to use that ffi macro correctly. doing some reading.
esthlos: I found the piece of code causing the trouble
esthlos: CCL::%ERRNO-DISP
esthlos: on the first invocation of make-temp-dir, (ccl::%GET-ERRNO) returns -38, which according to https://www.thegeekstuff.com/2010/10/linux-error-codes/ is "function not implemented"
esthlos: sorry, don't mean CCL::%ERRNO-DISP is causing the trouble. CCL::%ERRNO-DISP is setting errno to -38, even though it seems to successfully create the directory
esthlos: err, #_mkdtemp sets errno to -38
trinque: http://p.bvulpes.com/pastes/kHzNl/?raw=true << this is why I think it's being called incorrectly.
trinque: appears to want some C type there as the argument, which makes sense. I wouldn't know how to provide that, just yet.
trinque: "There's no supported way to directly pass lisp data to foreign functions: scalar lisp data must be coerced to an equivalent foreign representation, and lisp arrays (notably strings) must be copied to non-GCed memory." << right there in the https://ccl.clozure.com/manual/chapter13.2.html
esthlos: example: (ccl:with-cstrs ((x "x")) (#_puts x))
trinque: yep, was over here trying to jam it in with make-cstring
trinque: appears to work, i.e. (#_mkdtemp (ccl::make-cstring "/tmp/fooXXXXXX"))
esthlos: can you check the errno with (ccl::%GET-ERRNO) ?
esthlos: as in (progn (#_mkdtemp (ccl:with-cstrs ((x "/tmp/fooXXXXXX")) (#_mkdtemp x))) (format t "~a" (ccl::%GET-ERRNO)))
trinque: getting a -22 out of that
esthlos: yep, and that's causing the error I believe
esthlos: but, verrry strangely, doing it with make-cstring works for me
trinque: well, expand the with-cstrs macro and see if it does what you assume
trinque: also the code you pasted had a #_mkdtemp wrapped in a #_mkdtemp ?
esthlos: oh, yeah
esthlos: I might need some sleep
esthlos: now it's returning errno 0
trinque: so what are we doing here?
esthlos: I'm looking into the expansion of ccl:with-cstrs to see why it's #_mkdtemp might raise an error
esthlos: make-temp-dir is still setting errno to -38. only my snippet was ill-formed
trinque: I don't think %GET-ERRNO does what it seems. mkdtemp returns a null pointer on error, proper pointer otherwise, and I'm seeing that change depending on whether e.g. there's enough X in the pattern.
trinque: phf, if you're around, would you be more familiar with ^ ?
trinque: esthlos: http://p.bvulpes.com/pastes/QjAE4/?raw=true << here's a fixed make-temp-dir
esthlos: excellent, I was trying to make heads or tails of http://p.bvulpes.com/pastes/eFqSm/?raw=true
esthlos: I am curious though, if http://btcbase.org/log/2018-05-23#1817664 was working for phf, because it wasn't for either of us it seems
a111: Logged on 2018-05-23 15:04 phf: http://btcbase.org/log/2018-05-23#1817293 << i started writing a lispy make-temp-directory but the implementation is not particularly elegant (C concerns are at odds with lisp concerns), you can pouch the ccl bits though. i'm not sure if there's a better way to do errno handling, without relying on private ccl symbols
trinque: esthlos: http://p.bvulpes.com/pastes/odjHx/?raw=true << fixed run-subprocess for ccl. external-process-status return is a "values"
trinque: without the format, of course.
esthlos: ah, nice
trinque: with those changes I was able to press trb with a ccl-built binary
esthlos: have to change the make-pathname of make-temp-dir back btw, because sbcl barfs
trinque: ah k
trinque: well, seems like this is pretty close to done!
esthlos: so, what's the path from here?
esthlos: yep!
trinque: get the defpackage in there, test, then when you're satisfied, post a genesis patch on your blog.
esthlos: hey douchebag feel like some pentesting?
esthlos: break my shit
esthlos: sounds good trinque
trinque: eh, I wouldn't rely on him for that. he's yet to show any signs he's pressed a trb himself using the previous V
trinque: I'll be banging on it myself as well.
esthlos: heh, there was some attempted sarcasm there
trinque: oic
trinque: what's this contents thing, emacsism?
esthlos: very light literate programming
esthlos: only to stay organized
esthlos: though wrt the pentesting, I would be grateful for any flaws found
esthlos: anyway, think its bedtime. defpackage tomorrow
esthlos: night
esthlos: oh I should add, the vast majority of time of the proggy is spent calling gpg. would be nice to replace that turd
trinque: sure would, cya
← 2018-05-25 | 2018-05-27 → ↑
Random(trilema) | Download hourly DB snapshot | Get Source Code