←2014-11-14 2014-11-15 2014-11-16→ ↑2014 ↑all
00:00:15 <oerjan> yeah
00:02:34 -!- NATT_SiM has quit (Remote host closed the connection).
00:04:13 <int-e> ok, so conversion to string doesn't enter the picture; if anything that part will slow the Haskell version down.
00:07:10 -!- adu has joined.
00:10:03 -!- AndoDaan has joined.
00:13:29 <int-e> what fun. So apparently the last lambdabot downtime was caused by a Windows Server 2012 bug, leading to a switch being flooded with IPv6 packets, overloading the switch's CPU, and thus affecting traffic of completely unrelated hosts.
00:17:22 -!- Bicyclidine has joined.
00:17:30 <oerjan> just bill microsoft hth
00:21:50 -!- AndoDaan has quit (Ping timeout: 272 seconds).
00:22:33 -!- olls_ has quit (Quit: Lost terminal).
00:26:25 -!- AndoDaan has joined.
00:27:29 <Bicyclidine> this serial text thing's manual says to use control sequences starting with "the escape character" but doesn't say what that character is, so i'm like the shit is this fuck
00:27:37 <Bicyclidine> and it turns out it's the ascii ESC control character. i feel very young atm
00:28:09 * oerjan throws Bicyclidine a lollipop
00:28:17 <Bicyclidine> what flavor
00:28:21 <oerjan> strawberry
00:28:28 <Bicyclidine> excellent *licks*
00:28:32 <AndoDaan> !blsq ""Pp,20rz{2j**2j**1.+lnPP_+' _+Pp}m[p\t]Q
00:28:33 <blsqbot> | Ain't nobody got time fo' dat!
00:28:54 <AndoDaan> why is string formatting so hard!
00:29:15 -!- adu has quit (Quit: adu).
00:29:21 <AndoDaan> Why cant computer intuitively know what I want by now!?
00:29:29 <AndoDaan> I blame turing.
00:29:47 <int-e> it's teaching you to be precise
00:30:28 <AndoDaan> pfft.
00:31:03 <int-e> if you manage, it'll reward you with the illusion that you're in control
00:31:40 <int-e> oh, cheers, I missed this part. "We've also disabled new uses of Windows Server 2012 for now."
00:32:13 <Bicyclidine> i think for formatting you can blame the complications of language
00:41:58 -!- AndoDaan_ has joined.
00:43:02 -!- AndoDaan has quit (Ping timeout: 244 seconds).
00:45:38 <oerjan> <b_jonas> does someone have a roman numeral to number translator bot here? <-- hm i remember doing that once
00:45:52 <oerjan> `run ls bin/*oma*
00:45:52 <HackEgo> bin/fromroman \ bin/randomanonlog \ bin/toroman
00:46:04 <oerjan> oh right you made it afterward
00:46:27 <olsner> do we have a radixal to numbers translator?
00:46:28 <oerjan> maybe i just made a haskell function
00:59:51 -!- aloril has quit (Read error: Connection reset by peer).
01:03:06 -!- NATT_SiM has joined.
01:07:37 -!- Bicyclidine has quit (Ping timeout: 244 seconds).
01:07:39 -!- NATT_SiM has quit (Remote host closed the connection).
01:10:04 -!- NATT_SiM has joined.
01:29:32 -!- NATT_SiM has quit (Remote host closed the connection).
01:31:13 -!- NATT_SiM has joined.
01:35:53 -!- NATT_SiM has quit (Ping timeout: 272 seconds).
01:37:26 -!- aloril has joined.
01:38:06 -!- MoALTz_ has quit (Quit: Leaving).
01:43:28 <vanila> i not really convinced rule 30 or whatever is turing complete
01:43:34 <vanila> is there a simple explanation of why it is
01:45:17 <vanila> tromp, I wonder how I couldl write a program to find that term without prior knowledge of it?
01:45:27 <vanila> maybe that isn't really doable
01:45:37 <vanila> the short Y combinator
01:47:16 -!- shachaf has left.
01:47:25 -!- MoALTz has joined.
01:48:52 <tromp_> vanila: int-e has such a program; ask him
01:49:07 <elliott> vanila: I don't think there's any proof rule 30 is TC
01:49:10 <vanila> it seem s very hard to proof
01:49:16 <elliott> just wolfram believes it is because it looks complex
01:49:27 <AndoDaan_> it;s rule 110 isn't it?
01:49:48 -!- AndoDaan_ has changed nick to AndoDaan.
01:49:59 <tromp_> https://en.wikipedia.org/wiki/Rule_110
01:50:06 <elliott> it looks kinda like rule 110 and rule 110 is TC :p
01:53:59 <oerjan> except rule 30 is reversible
01:54:17 <Bike> this snail? turing complete
01:54:18 <oerjan> > showIntAtBase 2 intToDigit 30 ""
01:54:20 <lambdabot> "11110"
01:54:41 <oerjan> which 110 isn't
01:55:16 <vanila> reversible turing machine wold be cool
01:55:38 <Bike> imo reversible thermodynamics
01:55:58 <oerjan> vanila: those exist. also, see http://esolangs.org/wiki/Reversible_Brainfuck
01:57:23 <vanila> i dont understand that
01:57:48 <oerjan> what part
01:57:52 <vanila> how can you implement normal BF using reversible BF
01:58:02 <vanila> is it storing state in between the cells
01:58:09 <oerjan> yes
01:58:16 <Bike> so does anyone know anything funny i can do with a 2×40 terminal
01:59:21 <oerjan> vanila: see the Tape Layout table (i made that)
01:59:30 <oerjan> *layout
01:59:41 <vanila> I see
02:00:36 <tromp_> Bike, you can turn play a very narrow tetris where no piece can be rotated:(
02:01:18 <vanila> I meant rule 110 though
02:01:40 <vanila> how is the polynomial overhead accepted?
02:02:12 <oerjan> vanila: rule 110 works by implementing a cyclic tag machine with colliding "gliders"
02:02:35 <oerjan> this part has polynomial overhead.
02:02:49 <oerjan> (although the setup is an infinite pattern)
02:03:09 <vanila> steup too
02:03:12 <vanila> setup too
02:03:21 <vanila> am I just being too strict?
02:03:23 <vanila> i feel like this wrong
02:03:52 <vanila> i'm not rejecting the proof just don't really understand how it relates to the claim that it's TC
02:04:12 <oerjan> vanila: it is dubious in a way, tc-ness only makes _complete_ sense if you are looking at finite setup and output.
02:04:20 <vanila> so i can simulate a cyclic tag machine if I encode it in this particular way into rule 110
02:04:37 <vanila> yeah
02:04:39 <oerjan> ais523's proof for the 2,3 turing machine has the same problem, only slightly worse
02:04:40 <vanila> that's kind of what I feel
02:04:54 <vanila> hthese are great results
02:04:55 <vanila> of course
02:04:58 <elliott> isn't the setup repeating though
02:05:06 <oerjan> elliott: for 110 yes
02:05:12 <elliott> that's not too bad
02:05:38 <vanila> also I don't doubt these ARE TC
02:05:54 <vanila> i.e. that it can be done with a finite setup
02:06:03 <vanila> or is there a reason to think that's not possible?
02:06:24 <Bike> ais is pretty smart and he couldn't figure it, how about that
02:06:54 <Bike> also there's finite information in the initial setup, right? so like whatever man.
02:07:01 <vanila> Bike: QED
02:07:40 <oerjan> vanila: well you really want a finite _output_ too, which is hard because these CA's create intrinsically growing patterns
02:08:06 <elliott> Bike: yeah but with 2,3 it requires actual computation, the initial pattern is non-repeating
02:08:14 <Bike> yeah i remember
02:08:21 <elliott> in the limit this is obviously cheating (if the required computation was TC)
02:08:37 <elliott> *fix that parenthical to be more pedantically worded
02:09:10 <Bike> well we're not in the limit
02:09:29 <oerjan> once you have infinite setup, it becomes really subtle whether your real computation takes place in the setup or afterward
02:09:49 <Bike> is that even a meaningful distinction? imo there are no boundaries anywhere, man
02:10:02 <vanila> yes
02:10:04 <vanila> that's a good point
02:10:09 <vanila> who even are we?
02:10:16 <oerjan> Bike: i'm not talking about CAs specifically
02:10:16 <Bike> maybe we could determine when computation is occuring through some rigorous physical measure, like watching waste heat
02:10:19 <Bike> this plan has no flaws
02:10:20 <Bike> oerjan: me neither
02:10:26 <elliott> Bike: well, presumably you don't want to consider the identity function turing complete
02:10:29 <oerjan> Bike: i think you may be high
02:10:33 <elliott> but it is if you prepare its initial state with a universal turing machine
02:11:40 <vanila> so a 1D CA has some initial state: an n-bit string (any natural number n)
02:11:43 <Bike> well i mean how formalizable is the idea of a computer, really? how do you compare lambda calculus and a turing machine? you can emulate one in the other, great, but i can emulate a turing machine with toilet paper and crayons
02:11:47 <Bike> bla bla bla
02:11:56 <vanila> and I guess there is already a question of termination
02:11:57 <Bike> anyway annoying thing about CAs, nobody told me they were used in fluid dynamics
02:12:05 <vanila> how do you get output?
02:12:47 <vanila> Bike: toilet paper and crayons is TPC, this is differnet than TC
02:12:57 <Bike> how do you get output out of a computer? you pick some time after your computation is running and say "okay, this is the output"
02:13:29 <Bike> we need more lattice gas automata
02:13:45 <elliott> toilet paper complete??
02:13:50 <elliott> oh
02:13:55 <oerjan> vanila: i think for rule 110 you can basically search each generation's state with a regexp to see if it has the "i am finished" tag
02:14:20 <elliott> vanila: isn't the initial state always infinite?
02:14:26 <elliott> just often it's all 0s outside of a finite area
02:14:32 <vanila> for my personal satisfaction, initial state should be finite
02:14:38 <vanila> oh
02:14:42 <Bike> "the initial state has finite support"
02:14:47 <vanila> fair
02:14:51 <Bike> go on. go full real analysis. i dare you
02:15:23 <Bike> homework: design a computing formalism that works when the initial state has arbitrary entropy everywhere but a finite area??
02:15:34 <oerjan> i recall once classifying the ways rule 110 could grow (leftwards, it never grows rightwards) into an infinite region of 0's. so i'm not so sure that you _cannot_ make it tc even with finite support.
02:15:38 <Bike> do you think i should eat a calzone for dinner tonight
02:15:49 <oerjan> (initially. it will grow infinitely nevertheless.)
02:16:10 <vanila> maybe we should just study 1D ca more and not try to shoehorn the kind of computation it does do into the 'TC' framework
02:16:50 <Bike> perhaps you could study.... lattice gases
02:17:02 <oerjan> you just need some way of doing glider guns to get the effect of the infinite setup while having only finite support at the start
02:17:16 <vanila> oerjan, ooh
02:17:21 -!- shikhin has quit (Ping timeout: 265 seconds).
02:17:29 <vanila> if someone worked that out it'd be worth a publication I think
02:17:39 <Bike> gasses
02:17:40 <oerjan> better i stay out of it then
02:17:42 <Bike> gassssses
02:18:31 <oerjan> my underload minimization attempts basically halted when ais523 said it might be publishable >:)
02:18:46 <oerjan> (well i already found the minimum)
02:19:22 -!- shikhin has joined.
02:21:09 <oerjan> <Bike> anyway annoying thing about CAs, nobody told me they were used in fluid dynamics <-- have you heard about terence tao's idea that one might theoretically prove navier-stokes (millennium problem) to blowup by embedding something like CA computation in ideal fluids
02:22:42 <Bike> yes
02:22:53 <Bike> unfortunately for me tao's blog is incomprehensible
02:22:54 <vanila> That's a cool idea
02:23:05 <vanila> tao is good at explaining things in a simple way...
02:23:10 <Bike> fluid dynamics is also incomprehensible, but i'm working on it
02:24:05 <oerjan> <Bike> homework: design a computing formalism that works when the initial state has arbitrary entropy everywhere but a finite area?? <-- i recall that there's an obvious prood that in a symmetric CA you cannot create a pattern that can expand into arbitrary chaos, namely just put a competing version of the pattern somewhere out there
02:24:13 <Bike> «The story is told of many giants of modern physics, but most plausibly of Heisenberg, that, on his death-bed, he remarked that the two great unsolved problems were reconciling quantum mechanics and general relativity, and turbulence. "Now, I'm optimistic about gravity..." »
02:24:46 <elliott> vanila: I guess the CA equivalent of TC would be, like, 110-complete?
02:24:53 <vanila> haha
02:24:58 <vanila> that's a good idea
02:25:08 <vanila> show that one CA is possible to simulate all the others
02:25:08 <Bike> oerjan: well that would figure, yeah.
02:25:15 <elliott> oh, yeah, that's a better definition.
02:25:25 <Bike> oerjan: i wonder if you can make that into the fluctuation theorem somehow!
02:25:30 <elliott> it still might involve infinite initial states and not-completely-trivial encodings though...
02:25:34 <elliott> but you don't have to worry about halting
02:25:40 <oerjan> *proof
02:25:47 <vanila> the meta interpreter for game of life is so cooooool
02:26:03 <elliott> the metacell thing? yeah!
02:28:16 <Taneb> Today I used Lua for the first time
02:28:21 <vanila> rule for CA simulation: initial state must be START CELL(C[0]) MID CELL(C[1]) MID ... STOP each of these is finite and CELL is a function that takes takes a 0 or 1
02:28:24 <Taneb> Well, in the past 24 hours definition of "today"
02:29:47 <Bike> good times: installing awesomewm, installing a lua browser, crashing lua browser which brings down awesome somehow
02:30:59 -!- NATT_SiM has joined.
02:31:02 <Bike> still waiting for guidance re: calzones
02:32:02 <oerjan> un calzone per favore, and then you'll have some more
02:32:46 <Taneb> I made a game
02:33:14 <oerjan> Bike: you don't need our guidance, it's all baked in hth
02:33:32 <elliott> Bike: is this like when I used to get you to tell me when to go to bed? :p
02:33:55 <Bike> no it's like whne i ask about fucking calzones!
02:33:59 <Bike> honestly i don't think this is hard
02:34:28 <Bike> did you ever go to bed btw
02:35:09 <elliott> nope. never.
02:35:21 <oerjan> i think i've eaten calzone a few times hth
02:35:28 <Taneb> I've had calzone once or twice
02:35:59 <oerjan> but i cannot help getting the feeling "why are you folding the pizza?"
02:35:59 <Bike> that sounds terrible
02:36:06 <Bike> both the not sleeping and the not calzoning
02:36:39 <vanila> are you degreasing?
02:36:50 <oerjan> maybe he's degaussing
02:37:09 <Taneb> I fancy some calzone now
02:37:23 <Taneb> Maybe I should go to a dodgy takeaway and get some
02:37:53 <oerjan> la calzone nostra
02:41:49 <oerjan> something tells me dmm isn't always careful when queueing these http://www.mezzacotta.net/postcard/
02:43:43 -!- centrinia has quit (Ping timeout: 244 seconds).
02:54:29 <Sgeo> What would you call a rocket intended to observe quantum mechanical effects in outer space?
02:54:44 <vanila> a rocket
02:55:05 * oerjan waits for the pun
02:55:38 <Sgeo> No pun, but I'm considering building something like that in Second Life
02:55:53 <AndoDaan> That still exists?
02:56:10 <Sgeo> I originally estimated that it would take 7 days before completing its mission, but considering that I'm predicting the latter half to be sped up beyond the normal SL speed limit...
02:56:34 <Sgeo> Either that or it dies half a week before I'm expecting it to
02:57:18 <vanila> The Rule 90 automaton (in its equivalent form on one of the two independent subsets of alternating cells) was investigated in the early 1970s, in an attempt to gain additional insight into Gilbreath's conjecture on the differences of consecutive prime numbers
02:58:05 <vanila> "every six steps of the Rule 22 automaton simulate a single step of the Rule 90 automaton"
02:59:04 <oerjan> > showIntAtBase 2 intToDigit 90 ""
02:59:05 <lambdabot> "1011010"
02:59:29 <oerjan> > showIntAtBase 2 intToDigit 22 ""
02:59:30 <lambdabot> "10110"
03:00:23 <Sgeo> std::mem::transmute is really convenient when playing with floats
03:02:42 <vanila> woah :D
03:02:52 <vanila> apparnelty Wang of Wang-tiles proves that 2-cyclic tag machines are TC
03:08:03 <Sgeo> I think if I could just get 70 million meters in the sky, that would allow for a dramatic test of my theories
03:08:12 <Sgeo> I always thought it would be much higher
03:10:28 <elliott> Sgeo: I was briefly terrified and confused before I realised you were talking about second life
03:12:23 <vanila> second life has quantum physics?
03:14:12 <Sgeo> Second Life has floating point numbers in its physics. Floating point numbers have quantums that increase as the numbers get larger. Therefore, Second Life has quantum physics.
03:22:06 <vanila> lol
03:23:21 <Bike> oh hey my assembler treats '\0' as meaning '0', sweet
03:23:50 <int-e> Sgeo: "nonuniform planck length explorer" ;)
03:24:07 <Sgeo> NPLE?
03:24:46 <int-e> Sgeo: there, now you have a suggestion for the name of your rocket, now you can shoot it down by thinking of something better.
03:25:00 <Bike> i was wondering why my professor would put '\0' instead of just zero, and i have learned the answer: no fucking reason
03:26:02 <int-e> \cargo{cult}
03:27:21 <Bike> i don't think he actually ran this code, unless he wanted it to display 0Hello world0Hello wor
03:28:29 <Bike> so, nerds, does anyone know why terminal control sequences start with [?
03:28:47 <int-e> ^[[ is easy to type?
03:29:17 <Bike> god, i can almost believe that's the reason
03:30:20 <Bike> hm, apparently ^[[ is equivalent to a one eight-bit control character in C1
03:31:06 <Bike> which nobody uses, so that's great
03:32:45 <vanila> can someone help me understand cyclic tag machine
03:33:20 <Bike> wikipedia seems pretty straightforward
03:33:31 <Bike> whatchu havin problem with
03:33:37 <vanila> Initial word: baa
03:33:37 <vanila> acca
03:33:40 <vanila> where does the a go
03:34:13 <Bike> Um. That seems wrong.
03:34:40 <Bike> Where are you going off of? Wikipedia and esolang both would seem to say that's not how it works.
03:34:47 <Bike> assuming you mean acca is the product from baa
03:36:46 <int-e> vanila: note the m here: "t(S) is the result of deleting the leftmost m symbols of S and appending the word P(x) on the right." ... in the example, m=2.
03:37:06 <vanila> ill try to implement this so i can undersatnd it
03:37:07 <Bike> oh.
03:38:10 <oerjan> vanila: that's a tag system, but not a _cyclic_ tag system. those are different.
03:38:27 <int-e> and it's a 2-tag system, whereas tag systems are 1-tag systems
03:38:46 <vanila> :( this is so confusing
03:38:47 <int-e> (hence the m)
03:38:52 <oerjan> int-e: *cyclic tag systems
03:39:15 <int-e> oerjan: no.
03:39:32 <vanila> A tag system is a deterministic computational model published by Emil Leon Post in 1943 as a simple form of Post canonical system
03:40:45 -!- centrinia has joined.
03:41:18 <oerjan> int-e: um yes.
03:41:40 <Bike> maybe
03:41:49 <Bike> (height of comedy)
03:42:48 <oerjan> int-e: or rather, ordinary 1-tag systems are uninteresting because not tc, or at least not proven so
03:42:48 <int-e> oerjan: The Wikipedia article introduces tag systems, m-tag-systems (where tag systems arise as a special case by taking m=1), and then cyclic tag systems.
03:42:57 <int-e> oerjan: I'm not sure what you're trying to correct.
03:43:09 <oerjan> hm... let me check that.
03:43:37 <int-e> it also introduces cyclic tag systems further down in the article
03:44:19 <int-e> mm
03:44:23 <vanila> so im confused about all this
03:44:30 <int-e> no, actually they say that m is implicit, with no default.
03:44:32 <Bike> yaeh it's all bullshit
03:44:48 <int-e> but still there's no cyclicity there at all.
03:45:12 -!- NATT_SiM has quit (Remote host closed the connection).
03:45:21 <int-e> oerjan: so there was something to correct, but your suggestion made it more wrong
03:45:48 <oerjan> OKAY
03:46:27 <vanila> http://link.springer.com/article/10.1007%2FBF01343730#page-2
03:46:43 <vanila> T_i : s_i --> E_i, i = 1, .., n
03:47:20 <vanila> if the first symbol of a string is s_i then the first beta symbols are removed and E_i is appended at the end
03:47:34 <oerjan> vanila: the confusing thing here is that cook/wolfram changed the definition of tag systems because it was easier to make a pattern of gliders deal with single bits one at a time
03:48:13 <vanila> if the alphabet contains m symbols then n <= m
03:48:16 <vanila> that's odd
03:48:19 <oerjan> and it was easy to change the gliders used from one step to the next since they're all part of the infinite setup.
03:48:54 <oerjan> (easy once you have found useable gliders in the first place)
03:49:27 <int-e> vanila: love the use of \beta without saying what it is first.
03:50:00 <vanila> not only is it not say first, but not said second either
03:50:11 <vanila> in fact I have no idea what beta is
03:50:12 <int-e> "monogenic"?!
03:50:30 <int-e> \beta seems to be the 'm' from wikipedia's article.
03:50:33 <Bike> q: is it sane to have a function called 'puts' that also works for terminal control sequences? i don't know if C puts does
03:51:00 <Bike> i guess maybe
03:51:14 <vanila> this is so hard
03:51:35 <vanila> so I guess beta is just a fixed constant
03:52:08 <int-e> yes. a positive integer
03:52:28 <oerjan> there are no \0s in terminal control sequences are there?
03:52:54 <oerjan> and isn't that all puts demands
03:53:59 <vanila> so basically this whole area is a confusing nightmare
03:54:05 <elliott> puts does append a newline though...
03:54:30 <Bike> oh i guess i'm not doing that
03:54:44 <vanila> http://arxiv.org/abs/cs/0612089
03:54:45 <oerjan> vanila: um no, just a bit unfortunate naming, a cyclic tag system is not a tag system, but analogous to one
03:54:49 <Bike> i guess this is more like fputs then
03:56:11 -!- shachaf has joined.
03:57:35 <vanila> why are they barring symbols? beforelemma 1
03:58:17 -!- AndoDaan has quit.
03:59:47 <vanila> I can't understand this
04:00:37 -!- coppro has quit (Ping timeout: 256 seconds).
04:01:20 -!- coppro has joined.
04:02:50 <vanila> http://dspace.mit.edu/handle/1721.1/6107#files-area
04:02:52 -!- ^v has quit (Remote host closed the connection).
04:03:51 <int-e> vanila: those are 7 different variants each of 0 and 1: barred, dotted, barred with underset 1, possibly dotted, and plain with underset 2. ... ouch.
04:04:50 <vanila> I see
04:05:05 <vanila> I found an explanatio of tag system which makes sense
04:07:22 <int-e> Ok, Lemma 1 is correct, but the Proof requires some doodling on paper to believe.
04:08:53 -!- copumpkin has joined.
04:09:07 <int-e> the key is that the x -> \epsilon rules delete pairs of characters, so you're left with \bar x_1 \dot\bar x_1 (and then x_2) if l is even or \dot \bar x_1 if l is odd.
04:09:26 <vanila> oh yeah!
04:11:05 <int-e> But it's confusing because 'w' does not match the encoding of words introduced in the preceding paragraph.
04:19:34 <vanila> http://lpaste.net/114334
04:19:36 <vanila> i made this
04:20:00 <vanila> I should create a format so people can make .tag files for it to run
04:21:42 <vanila> I have a programming challenge to #esoteric
04:22:12 <vanila> Write a program which finds a fixed point combinator for SK combintors (or some language like that)
04:22:20 <vanila> it's hard...
04:23:52 <int-e> ... http://sprunge.us/TRPL
04:24:30 <vanila> I declare int-e the winner
04:25:11 <int-e> you must have missed the earlier discussion on the topic
04:25:38 -!- oerjan has quit (Quit: leaving).
04:25:39 <vanila> yeah and i wrote a program to try this yesterday but it didnt wokr
04:26:28 <int-e> (about 5 1/2 hours ago)
04:28:07 <vanila> which of those output are fixed point combinators?
04:29:18 <int-e> the first one is a proper one; the last three work in a Böhm tree model (and Scott topology, I presume); they have the property that M f unfolds to f (f (f (f (f ...))))
04:29:35 -!- shikhin has quit (Ping timeout: 264 seconds).
04:30:18 <int-e> (but the program is unable to verify this.)
04:31:39 <vanila> I can see why you'd use tag systems for a CA with diagonal gliders
04:31:45 <int-e> the ones with [3] belong in the "unknown" category (where I think it's quite a safe bet that none of them are fixed point combinators, but there's no proof.)
04:32:00 <vanila> that's really interesting, I wonder what they are!
04:32:05 <vanila> maybe i should convert them to lambda terms
04:33:37 <elliott> int-e: is the difference between the proper and bohm tree model ones that M f isn't f (M f) but f (M' f) where M' f is f (M'' f) and so on?
04:34:06 <elliott> that seems like a fixed point combinator to me, since the Ms are essentially "implementation details"?
04:34:07 <int-e> elliott: yeah, the shape of the combinator changes.
04:34:15 <elliott> like, there's no way to observe the difference as f, right?
04:34:44 <int-e> elliott: but the resulting limiting infinite tree (that's the Böhm tree) is the same
04:34:57 <int-e> elliott: right.
04:34:57 * elliott nods
04:37:02 <vanila> i dont know what tool i can use to investigate a lambda term
04:39:21 <vanila> interesting that you can encode turing machines into tag systems
04:39:31 <vanila> i wonder if there's an easier way to show them TC
04:41:13 <vanila> http://codegolf.stackexchange.com/questions/40300/simulate-a-cyclic-tag-system
04:43:52 <elliott> those are so short...
04:46:03 <int-e> elliott: the first one is (\y f. y f y) (\f y. f (y f y))
04:46:27 * elliott nods. (was that meant to be for vanila?)
04:46:45 <int-e> probably.
04:47:00 * int-e is a random nick selector.
04:47:17 <vanila> looks like a y combinator to me!
04:47:22 <vanila> a very nice one too
04:47:49 <vanila> http://labs.orezdnu.org/lambda/ i foun d this which lets you manually reduce lambda terms
04:55:33 <Bike> hot damn is terminfo organized alpahbetically
04:56:28 <vanila> can the bot turn unlambda into lambda?
04:59:34 <elliott> vanila: that would violate the second law of thermodynamics :p
05:00:01 <vanila> haha
05:07:04 <vanila> http://lpaste.net/114335
05:07:10 <vanila> here are the terms then
05:13:35 <int-e> those are the ones that are conjecturally not fixed point combinators
05:14:17 <vanila> yeah mi interested in those
05:14:25 <vanila> if theyre not fixed poits they must be something weird and cool
05:17:25 <elliott> vanila: they could just be f (f (f (f (f (f _|_))))) or the like
05:17:39 <elliott> or, less nesting probably
05:18:20 <vanila> its hard to reduce it
05:18:30 <vanila> i tried in the javascript program but if you make a mistake it gets loads bigger
05:25:56 -!- AndoDaan has joined.
05:34:22 -!- AndoDaan_ has joined.
05:34:45 -!- AndoDaan has quit (Ping timeout: 255 seconds).
05:35:22 -!- nys has quit (Quit: quit).
05:36:57 -!- vanila has quit (Quit: Leaving).
05:45:36 -!- AndoDaan_ has quit.
05:47:45 -!- adu has joined.
05:57:16 <Bike> update: i ate a calzone anyway. it was kind of a lot so i feel a bit sick but i'm sure i'll pull through, no thanks to you jerks
05:59:49 <elliott> our uncertainty was meant to represent the idea of you eating half a calzone!
06:52:15 -!- adu has quit (Quit: adu).
07:04:42 -!- centrinia has quit (Ping timeout: 245 seconds).
07:28:49 -!- shikhin has joined.
07:49:48 <MDude> I've made something that works very poorly http://orteil.dashnet.org/randomgen/?gen=http://mdude1350.webs.com/generators/random-code/IBNIZ-simple-generator.txt
07:50:43 <MDude> And no, it's not Windows.
07:56:19 -!- Patashu_ has joined.
07:56:19 -!- Patashu has quit (Disconnected by services).
08:14:42 -!- shikhout has joined.
08:17:30 -!- shikhin has quit (Ping timeout: 244 seconds).
08:28:22 -!- AndoDaan has joined.
08:33:01 -!- MDude has changed nick to MDream.
08:48:23 -!- Patashu_ has quit (Quit: Soundcloud (Famitracker Chiptunes): http://www.soundcloud.com/patashu MSN: Patashu@hotmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
08:57:23 -!- bb010g has quit (Quit: Connection closed for inactivity).
09:01:11 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
09:06:26 -!- Patashu has joined.
09:12:57 -!- j-bot has quit (Remote host closed the connection).
09:14:10 <J_Arcane> MDream: IBNIZ! IBNIZ is rad.
09:21:19 <MDream> It seems to ahve some problems generating, though.
09:21:55 <MDream> I think however the generator code works might be expecting something special from question marks.
09:26:50 <nyuszika7h> anyone interested in my slightly unusual approach to 99 bottles in Haskell? http://lpaste.net/2310281802079535104
09:29:52 -!- shikhout has changed nick to shikhin.
09:30:44 <elliott> that is such an abuse of show :)
09:31:18 <elliott> (also, technically you want toTitle there, but...)
09:33:57 <J_Arcane> nyuszika7h: I actually did laugh out loud. :D
10:10:08 -!- Phantom_Hoover has joined.
10:42:43 -!- S1 has joined.
10:49:40 -!- AndoDaan has quit (Ping timeout: 240 seconds).
10:50:10 -!- S1 has quit (Quit: S1).
10:50:26 -!- copumpkin has quit (Ping timeout: 244 seconds).
11:10:04 -!- fungot has quit (Remote host closed the connection).
11:11:29 -!- fungot has joined.
11:14:08 <fizzie> fungot: Still feeling all right?
11:14:09 <fungot> fizzie: so the source is encoded in some meta language. to do what you want the suspend button to suspend or hibernate? granted, it's easy to implement all fingerprints in cfunge.
11:50:06 <mroman> Working on weekends is so relaxing.
12:05:30 -!- boily has joined.
12:19:10 <mroman> ^8ball Does blsqbot need sha1?
12:19:10 <fungot> No.
12:19:33 <mroman> fungot: Why not?
12:19:33 <fungot> mroman: it's defined as a function? can't you use alien? beings)
12:19:45 <mroman> I'm not sure if aliens know what sha1 is.
12:24:03 <fizzie> Yes, they're probably stuck with MD5 still.
12:29:55 -!- boily has quit (Quit: HASH CHICKEN).
12:32:26 <shachaf> fungot means that it should be using sha-2 or sha-3, obviously.
12:32:26 <fungot> shachaf: there you go :d, say how you want to
12:32:36 <shachaf> fungot: you tell 'em
12:32:36 <fungot> shachaf: im even fnord. browse like that into the, syntax is discussed half as much mem
12:35:46 <mroman> Lousy MD5 aliens.
12:36:27 <mroman> `learn MD5 is a hash algorithm mainly used by underdeveloped aliens.
12:36:29 <HackEgo> Learned 'md5': MD5 is a hash algorithm mainly used by underdeveloped aliens.
12:37:18 <myname> :D
12:42:47 -!- Patashu has quit (Ping timeout: 272 seconds).
12:43:43 <mroman> Damnit I have no headphones
12:43:51 <mroman> but looks like nobody's here anyway.
12:43:53 <mroman> *music*
12:48:02 <mroman> I see there are new "assisted suicides" debates flaming on.
12:49:16 <mroman> I don't really see the point in debating that though
12:49:24 <mroman> It's pretty much obvious anyway
12:49:41 <mroman> At the end of the spectrum you have the religious people who are against it for religious reasons
12:50:06 <mroman> then in the middle of the spectrum are regular people who have never been in the situation of thinking about suicide and think suicide is bad no matter what circumstances
12:50:23 <mroman> and at the other end of the spectrum are those who say "who want's to do it shall be able to"
12:50:55 <mroman> what they should be debating about is whether the state/government has the right to keep a person alive against the person's will.
12:52:03 <mroman> (and then there are people who think all people thinking about suicide can be treated and healed)
12:53:18 -!- qlkzy has quit (Excess Flood).
12:54:28 <mroman> I hate it when people argue for themselves rather than for the state/government
12:54:44 -!- qlkzy has joined.
12:54:45 -!- qlkzy has quit (Excess Flood).
12:56:04 -!- qlkzy has joined.
12:56:05 -!- qlkzy has quit (Excess Flood).
12:56:44 -!- qlkzy has joined.
12:56:45 -!- qlkzy has quit (Excess Flood).
12:57:07 -!- drdanmaku has joined.
12:58:14 -!- qlkzy has joined.
12:58:15 -!- qlkzy has quit (Excess Flood).
13:02:00 -!- qlkzy has joined.
13:02:00 -!- qlkzy has quit (Excess Flood).
13:03:21 -!- qlkzy has joined.
13:03:21 -!- qlkzy has quit (Excess Flood).
13:04:01 -!- qlkzy has joined.
13:04:01 -!- qlkzy has quit (Excess Flood).
13:04:41 -!- qlkzy has joined.
13:04:41 -!- qlkzy has quit (Excess Flood).
13:04:51 -!- qlkzy has joined.
13:04:51 -!- qlkzy has quit (Excess Flood).
13:05:21 -!- qlkzy has joined.
13:05:21 -!- qlkzy has quit (Excess Flood).
13:06:00 -!- qlkzy has joined.
13:13:14 <mroman> [java.lang.RuntimeException](java.sql.SQLException: [SQLITE_MISUSE] Library used incorrectly (not an error))
13:13:18 <mroman> lol
13:18:21 -!- oerjan has joined.
13:34:35 <mroman> git can't handle multiple * in gitignore patterns?
13:35:39 <mroman> wth
13:46:38 <fizzie> In my (admittedly small) tests, those worked just fine.
13:46:51 <fizzie> (Didn't test any interaction with ** though.)
13:47:22 <mroman> */sandboxes/sandbox_*/gen/* doesn't seem to work
13:48:05 <fizzie> I tried http://sprunge.us/VPKe and it properly ignored a/b/c/foo and foo.baz.bar.quux but not the other two files.
13:48:42 <mroman> so it added bar and foo?
13:49:06 <mroman> or hm
13:49:30 <mroman> let me test something
13:49:37 <fizzie> It added a/b/c/bar and foo.bar.baz.quux, neither of which match the ignore patterns.
13:50:44 <fizzie> (I suppose you remember that * doesn't match a /.)
13:53:20 <mroman> ow
13:53:21 <mroman> ok
13:53:49 <mroman> I see
13:55:08 <fizzie> (That's what ** is for.)
13:59:15 <Melvar> I wish to know whether gcc is inlining certain function calls. How should I go about this?
14:01:02 <mroman> Melvar: Have you tried -Winline
14:01:35 <mroman> Using -Winline will warn when a function marked inline could not be substituted, and will give the reason for the failure.
14:01:52 <mroman> I guess this should tell you what it DIDN'T inline
14:02:13 <Jafet> objdump -dlx a.out
14:02:18 <Melvar> Well, if I want to use inline, I’d have to transfer the functions to the header file I think.
14:02:27 <mroman> Melvar: yes
14:02:35 <mroman> it doesn't work with linking
14:02:51 <Melvar> I’m specifying -flto both times.
14:02:51 <mroman> GCC doesn't cross-module optimizations afaik
14:03:00 <mroman> (unlike ghc which does exactly that)
14:03:48 <mroman> hm. didn't know gcc hat lto
14:03:53 <mroman> *had
14:04:34 <mroman> I just know that ghc favors optimization over binary compatibility :)
14:09:57 <Melvar> Hm, looks like it inlined two of three and the code for the remaining one looks larger than expected.
14:10:25 <mroman> I think GCC can even do partial inlining?
14:13:28 <Melvar> Hm, it also gained a numerical suffix.
14:13:51 <mroman> GCC even copies the same functions to different locations
14:14:10 <mroman> which is an awesome thing to do actually
14:14:44 <Melvar> I wonder what the .2536 suffix on this function means.
14:14:44 -!- shikhout has joined.
14:15:08 <mroman> because it guarantees that the function is in the instruction cache
14:16:10 <mroman> or maybe that was some other compiler
14:17:19 -!- shikhin has quit (Ping timeout: 244 seconds).
14:18:33 <Melvar> Huh, no, looks like I was wrong, it didn’t inline them after all.
14:19:53 <Melvar> I’ll have to move them to the header and specify inline, I guess.
14:24:17 <mroman> There's no guarantee gcc will inline your functions btw
14:24:53 <mroman> (unless you use __attribute__((always_inline)) )
14:26:27 <mroman> but I don't know much about gcc optimizations :(
14:26:47 <mroman> I only know that __builtin_expect makes a hell of a difference if used correctly
14:28:06 <mroman> With __builtin_expect my emulator suddently executed 10 MIPS more than without
14:28:16 <Melvar> Yeah, two of these are functions to index a C array, so they consist of a single small C statement. Since these are bindings, they each get called from one other function which gets called a bunch of times.
14:32:30 <mroman> I wonder how gcc implements atomic stuff on CPUs that don't have LOCK prefixes and the like
14:32:53 <mroman> Although I don't think gcc even targets more esoteric CPUs :D
14:34:40 <Jafet> In software, as usual
14:39:36 <mroman> which is usually just CLI; do stuff; STI
14:39:46 <mroman> not a lot of other choices there I think
14:43:01 <mroman> which won't work for exceptions actually
14:45:42 <mroman> @tell AndoDaan you could write a MNNBFSL interpreter in Burlesque .
14:45:42 <lambdabot> Consider it noted.
14:46:57 <mroman> Interpreters in Burlesque for Esolangs: Brainfuck (doesn't support output), Deadfish
14:48:27 -!- nys has joined.
14:50:32 <mroman> (and trivially Burlesque since there's eval)
14:50:53 <mroman> !blsq "\'5 5.+\'pe"pe
14:50:53 <blsqbot> | 10
14:51:36 <mroman> !blsq "\'\\5'\\'\'pe"pe
14:51:36 <blsqbot> | '"
14:51:36 <blsqbot> | ERROR: Unknown command: (\5)!
14:51:45 <mroman> !blsq "\'\5'\\'\'pe"pe
14:51:45 <blsqbot> | '"
14:51:45 <blsqbot> | ERROR: Unknown command: (\5)!
14:51:48 <mroman> hm
14:52:00 <mroman> !blsq "\'\'pe"
14:52:00 <blsqbot> | "\"\"pe"
14:52:03 <mroman> !blsq "\'\'pe"Q
14:52:03 <blsqbot> | ""pe
14:52:10 <mroman> !blsq "\'\'\'pe\'pe"Q
14:52:10 <blsqbot> | """pe"pe
14:52:19 <mroman> !blsq "\'\'5\'pe\'pe"Q
14:52:19 <blsqbot> | ""5"pe"pe
14:52:24 <mroman> !blsq "\'\'5\'pe\'pe"
14:52:25 <blsqbot> | "\"\"5\"pe\"pe"
14:52:29 <mroman> !blsq "\'\'5\'pe\'pe"pe
14:52:29 <blsqbot> | ERROR: Burlesque: (e!) Invalid arguments!
14:52:29 <blsqbot> | ERROR: Burlesque: (ps) Invalid arguments!
14:52:29 <blsqbot> | 5
14:52:38 <mroman> !blsq "\'\'5\'pe\'pe"psQ
14:52:38 <blsqbot> | ["", 5, "pe", pe]
14:52:46 <mroman> that looks wrong
14:52:57 <mroman> !blsq "\'\\'5\\'pe\'pe"psQ
14:52:57 <blsqbot> | ["\"5\"pe", pe]
14:53:02 <mroman> !blsq "\'\\'5\\'pe\'pe"pe
14:53:02 <blsqbot> | 5
14:53:32 <mroman> !blsq "\'\\'\\\\'5\\\\'pe\\'pe\'pe"pe
14:53:33 <blsqbot> | 5
14:54:26 <mroman> fizzie: \' escapes " btw.
14:54:32 <mroman> I'm not sure where that's documented.
14:54:36 <mroman> or if it even is documented
14:57:58 <mroman> !blsq "\'"
14:57:58 <blsqbot> | "\""
14:58:06 <mroman> but it will be printed as \", not as \'
14:58:14 <fizzie> !blsq "\""
14:58:14 <blsqbot> | ERROR: (line 1, column 5):
14:58:14 <blsqbot> | unexpected end of input
14:58:14 <blsqbot> | expecting "\""
14:58:20 <fizzie> Crafty.
14:58:28 <mroman> !blsq 4334ud[]
14:58:28 <blsqbot> | ERROR: Burlesque: ([[) Invalid arguments!
14:58:28 <blsqbot> | 4334
14:58:28 <blsqbot> | ERROR: Burlesque: (ud) Invalid arguments!
14:58:33 <mroman> !blsq 4334{ud[]
14:58:33 <blsqbot> | ERROR: (line 1, column 10):
14:58:34 <blsqbot> | unexpected end of input
14:58:34 <blsqbot> | expecting "%", "g", "s", "S", "m{", "q", "{", "\"", "-", digit, "'", "(", "y" or "}"
14:58:51 <mroman> I like that it can now display more lines :)
15:00:22 <mroman> !blsq ''
15:00:23 <blsqbot> | ''
15:00:33 <mroman> !blsq ''Q
15:00:33 <blsqbot> | '
15:00:35 <mroman> !blsq '"Q
15:00:36 <blsqbot> | "
15:00:38 <mroman> !blsq '\Q
15:00:38 <blsqbot> | \
15:00:47 <mroman> ' doesn't require any escaping
15:01:00 <mroman> !blsq ' L[
15:01:00 <blsqbot> | 'a
15:01:04 <mroman> !blsq ' **
15:01:04 <blsqbot> | 32
15:01:14 <mroman> !blsq ''**
15:01:14 <blsqbot> | 39
15:01:53 <mroman> !blsq 10L[
15:01:53 <blsqbot> | '
15:01:54 <blsqbot> |
15:02:16 <mroman> (^- that's a newline char)
15:03:56 <mroman> !blsq 1 0?/
15:03:57 <blsqbot> | That line gave me an error
15:06:19 <mroman> Well, I gotta go places. Not college, but places.
15:22:41 -!- copumpkin has joined.
15:24:14 -!- mihow has quit (Read error: Connection reset by peer).
15:26:56 -!- mihow has joined.
15:32:48 -!- AndoDaan has joined.
15:36:13 -!- AndoDaan has left.
15:36:31 -!- AndoDaan has joined.
15:42:04 -!- vanila has joined.
15:46:56 -!- AndoDaan has quit (Quit: Leaving).
15:48:51 -!- AndoDaan has joined.
16:03:58 <Vorpal> this is so strange
16:04:23 <Vorpal> This Mac (classic) emulator only works when running under strace or gdb
16:04:30 <Vorpal> Otherwise it just locks up at startup
16:05:05 <vanila> which emulatOR?
16:05:25 <Vorpal> sheepshaver
16:05:47 <vanila> sheepshaver is the most broken thing :///
16:06:09 <Vorpal> vanila, no kidding, are there any alternatives?
16:06:11 <vanila> ( idnotknow any others)
16:06:11 <idris-bot> (input):1:21: error: expected: "!!",
16:06:12 <idris-bot> "$", "$>", "&&", "&&&", "*",
16:06:12 <idris-bot> "***", "+", "++", "-", "->",
16:06:12 <idris-bot> ".", "/", "/=", ":+", ":-",
16:06:12 <idris-bot> "::", ":::", ":=", "<", "<$",↵…
16:06:13 <vanila> sorry
16:06:44 <Vorpal> vanila, Anyway BasiliskII has the same issue
16:06:44 <vanila> because i really wnat to emulate powermac :/
16:07:07 <Vorpal> vanila, anyway it doesn't run all the games I want.
16:07:25 <vanila> extremely slow and painful but classic macs run in mess, minivmac might be better for non powr mac but it's very odd not sure how to even run it
16:08:25 <Vorpal> vanila, mess?
16:08:49 <Vorpal> Is it related to mame?
16:08:51 <vanila> yes
16:09:02 <vanila> its too slow now but one day might be good
16:09:30 -!- olls has joined.
16:09:43 <Vorpal> vanila, Right, I want to play Escape Velocity Override.
16:10:00 <Vorpal> If that isn't doing 30+ FPS it isn't going to play well
16:10:21 <vanila> there was a TC for ev nova that might let you run that on an older version of mac os x
16:10:36 <vanila> it may be easier to emulate that or run it on real hardware depending on what you have
16:10:57 <Vorpal> TC?
16:11:07 <vanila> total conversion - it is official too
16:11:11 <Vorpal> Also I never played nova, I wonder if it was any good
16:11:16 <vanila> its really good
16:12:01 <Vorpal> vanila, Oh? Okay, so how do I emulate OS X? And where do I find EV Nova these days
16:12:32 <vanila> I don't really know, i've never done this but it's just an idea that might be another way to do this
16:14:35 <Vorpal> vanila, I do have an old first model iBook I can use. However, the battery is dead and the power connector is glitchy. Thus if you don't keep the computer perfectly still, the power is going to cut out
16:15:29 <vanila> a shame :( if you can repair that , I think it's the best bet
16:15:37 <Vorpal> Not sure how
16:17:00 <Vorpal> I'm not really good at doing that sort of stuff. Anyway, bbl, food is ready
16:19:01 <fizzie> Vorpal: PearPC, OS X and the classic thing in it. (Not very likely to be any good.)
16:22:52 <J_Arcane> EV Nova was also available for Windows apparenty, and supports the same mods to run the first two games.
16:23:34 <fizzie> And you can buy it for $30 from Ambrosia's web-store.
16:26:25 -!- NATT_SiM has joined.
16:35:18 <mroman> Vorpal: sounds like a race condition .
16:35:36 <mroman> gbd has the property of making race conditions suddenly not buggy anymore
16:36:18 <mroman> !blsq 1ng
16:36:18 <blsqbot> | -1
16:38:30 <mroman> !blsq {1}1+]
16:38:31 <blsqbot> | {1 1}
16:38:33 <mroman> !blsq {1}2+]
16:38:34 <blsqbot> | {2 1}
16:38:36 <mroman> !blsq {1}2[+
16:38:36 <blsqbot> | {1 2}
16:44:27 -!- MDream has changed nick to MDude.
16:44:56 -!- vanila has quit (Quit: Leaving).
16:46:04 <mroman> blsq ) 10rzziq?*m^
16:46:05 <mroman> {0 1 4 9 16 25 36 49 64 81 100}
16:46:08 <mroman> finally blsq has this
16:48:01 <mroman> > scanl1 (<) [1,2,3,4]
16:48:03 <lambdabot> No instance for (GHC.Num.Num GHC.Types.Bool)
16:48:03 <lambdabot> arising from the literal ‘1’
16:57:34 <mroman> blsq ) {1 5 9 10 2 3 1 0}0{.>}LO
16:57:35 <mroman> {1 5 9 10}
16:57:41 <mroman> and that. But I have no name for that one yet.
16:58:11 <mroman> blsq ) {1 5 9 10 2 3 1 0 99}0{.>}LO
16:58:11 <mroman> {1 5 9 10 99}
16:58:24 <mroman> this gives you a sequence of larger getting numbers in a list.
16:58:47 <myname> deletion sort
16:59:05 <mroman> blsq ) {1 5 9 10 2 3 1 0 99}100{.<}LO
16:59:06 <mroman> {1 0}
17:03:11 -!- fungot has quit (Ping timeout: 260 seconds).
17:05:21 -!- shikhout has changed nick to shikhin.
17:08:01 -!- nortti has changed nick to lawspeaker.
17:09:20 -!- lawspeaker has changed nick to nortti.
17:12:26 -!- nortti has changed nick to lawspeaker.
17:13:04 -!- oerjan has quit (Quit: Running after fungot).
17:14:26 -!- lawspeaker has changed nick to nortti.
17:23:48 <Vorpal> <J_Arcane> EV Nova was also available for Windows apparenty, and supports the same mods to run the first two games. <-- ooh
17:24:10 <Vorpal> fizzie, 30 USD for that old of a game is a bit of a stretch though :/
17:25:48 <Vorpal> <mroman> Vorpal: sounds like a race condition. <-- well strace too, but that use ptrace as well. Anyway, this program used to work, on this computer, on an earlier Ubuntu version
17:33:41 -!- MDude has quit (Ping timeout: 240 seconds).
17:34:48 <fizzie> Yes, it's not exactly a bargain price.
17:39:10 -!- NATT_SiM has quit (Remote host closed the connection).
17:39:19 <J_Arcane> Blargh.
17:43:28 -!- vanila has joined.
17:44:32 * J_Arcane found a very nice bit of code for an infix code-block operator in Racket, which is sadly not licensed. Debating bribing the author ...
17:45:03 <vanila> can i see?
17:45:22 <J_Arcane> http://artyom.me/learning-racket-2
17:45:39 <J_Arcane> Round about the end of the first section.
17:45:55 <vanila> The thing is, “define” is too long. Why not use def instead?
17:45:57 <vanila> :/
17:46:05 <J_Arcane> Heh. :D
17:46:17 <J_Arcane> Heresy actually does use def.
17:46:47 <vanila> this is a bad page
17:46:50 <J_Arcane> And while I love scheme/racket wordy function names, they can get a bit tedious (or would if I had to write nearly as much code as in other languages to do anything)
17:47:48 <vanila> damn
17:47:55 <vanila> his : macro is so clever it's making me annoyed
17:53:21 -!- NATT_SiM has joined.
17:55:40 <J_Arcane> vanila: I know, right?
17:55:58 <J_Arcane> I've looked at some other libs and things for it, and none are have as clever or simple.
18:05:42 -!- NATT_SiM has quit (Remote host closed the connection).
18:14:30 -!- MoALTz_ has joined.
18:18:48 -!- MoALTz has quit (Ping timeout: 264 seconds).
18:33:45 -!- paul2520 has quit (Quit: WeeChat 1.0).
18:37:46 -!- shikhin has changed nick to SHIKHIN.
18:38:26 -!- SHIKHIN has changed nick to shikhin.
18:48:05 -!- esowiki has joined.
18:48:10 -!- esowiki has joined.
18:48:10 -!- esowiki has joined.
18:48:42 -!- esowiki has joined.
18:48:46 -!- esowiki has joined.
18:48:47 -!- esowiki has joined.
18:49:16 -!- esowiki has joined.
18:49:20 -!- esowiki has joined.
18:49:21 -!- esowiki has joined.
18:49:51 -!- esowiki has joined.
18:49:55 -!- esowiki has joined.
18:49:56 -!- esowiki has joined.
18:50:41 -!- esowiki has joined.
18:50:46 -!- esowiki has joined.
18:50:46 -!- esowiki has joined.
18:51:32 -!- esowiki has joined.
18:51:36 -!- esowiki has joined.
18:51:37 -!- esowiki has joined.
18:52:21 -!- esowiki has joined.
18:52:26 -!- esowiki has joined.
18:52:26 -!- esowiki has joined.
18:53:11 -!- esowiki has joined.
18:53:16 -!- esowiki has joined.
18:53:16 -!- esowiki has joined.
18:54:01 -!- esowiki has joined.
18:54:06 -!- esowiki has joined.
18:54:06 -!- esowiki has joined.
18:54:51 -!- esowiki has joined.
18:54:56 -!- esowiki has joined.
18:54:56 -!- esowiki has joined.
18:55:41 -!- esowiki has joined.
18:55:42 -!- glogbot has joined.
18:55:46 -!- esowiki has joined.
18:55:46 -!- esowiki has joined.
18:56:34 -!- AndoDaan_ has joined.
18:57:35 -!- AndoDaan has quit (Ping timeout: 240 seconds).
18:57:42 -!- kline has joined.
18:59:18 -!- Bike_ has joined.
18:59:37 -!- fizzie` has joined.
19:00:34 -!- lambdabot has quit (Remote host closed the connection).
19:00:34 -!- fizzie has quit (Ping timeout: 265 seconds).
19:00:37 -!- scounder has quit (Ping timeout: 265 seconds).
19:00:39 -!- Bike has quit (Ping timeout: 265 seconds).
19:00:40 -!- conehead has quit (Ping timeout: 265 seconds).
19:00:40 -!- InvalidCo has quit (Ping timeout: 265 seconds).
19:00:41 -!- int-e has quit (Ping timeout: 265 seconds).
19:00:43 -!- int-e_ has joined.
19:01:03 -!- InvalidCo has joined.
19:02:35 -!- AndoDaan_ has changed nick to AndoDaan.
19:05:13 -!- lambdabot has joined.
19:06:18 -!- conehead has joined.
19:07:28 -!- Bike_ has changed nick to Bike.
19:21:00 <J_Arcane> IT's weird to me how BASIC has so few predicate functions.
19:22:37 -!- AndoDaan_ has joined.
19:23:04 -!- ^v has joined.
19:25:25 -!- AndoDaan has quit (Ping timeout: 264 seconds).
19:26:42 -!- AndoDaan_ has changed nick to AndoDaan.
19:29:40 -!- variable has quit (Ping timeout: 255 seconds).
19:32:39 -!- AnotherTest has joined.
19:35:47 -!- clog has joined.
19:36:30 -!- NotJoel has joined.
19:36:36 -!- variable has joined.
19:36:44 -!- clog_ has quit (Ping timeout: 255 seconds).
19:37:17 -!- ZombieAlive has quit (Remote host closed the connection).
19:41:03 -!- olls has quit (Quit: leaving).
19:41:57 -!- olls has joined.
19:43:43 -!- _Paul has joined.
19:43:51 <_Paul> whaaaaaaaaaaaaaaaaaaaaaaaats up dawgs
19:43:56 <vanila> Hi]
19:44:01 <_Paul> hey nilla
19:44:08 <vanila> I want to implement unlambda
19:44:14 <_Paul> sounds greart
19:52:05 -!- scounder has joined.
19:55:32 <NotJoel> Hi
19:57:07 <elliott> hi
19:57:09 <vanila> unlambda ``r`ci``s`k`c``s``s`ksk`kr.*
19:57:15 <vanila> ,unlambda
19:57:56 <elliott> !unlambda `r`ci``s`k`c``s``s`ksk`kr.*
19:57:58 <EgoBot> No output.
19:58:05 -!- _Paul has left ("Leaving").
19:58:09 <vanila> my intepreter fails on this program
19:58:40 <vanila> I guess it's the c's
19:58:44 <vanila> I implement C wrong
19:59:50 -!- nyuszika7h_ has joined.
20:00:04 -!- nyuszika7h has quit (Killed (sinisalo.freenode.net (Nickname regained by services))).
20:00:05 -!- nyuszika7h_ has changed nick to nyuszika7h.
20:00:19 -!- pikhq_ has joined.
20:00:41 -!- int-e has joined.
20:00:50 -!- viznut_ has joined.
20:02:32 <vanila> !unlambda ``r`cd`.*`cd
20:02:32 <EgoBot> ​\ * \ ** \ *** \ **** \ ***** \ ****** \ ******* \ ******** \ ********* \ ********** \ *********** \ ************ \ ************* \ ************** \ *************** \ **************** \ ***************** \ ****************** \ ******************* \ ******************** \ ********************* \ ********************** \ *********************** \ ************************ \ ************************* \ **************************
20:02:39 <vanila> c is very good
20:04:07 -!- olsner_ has joined.
20:04:32 -!- ^v has quit (Read error: Connection reset by peer).
20:04:33 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/fibo.unl
20:04:36 <vanila> i dont really believe this program
20:04:43 <vanila> <add>
20:04:56 <vanila> a continuation? :/
20:05:08 <vanila> how many extraneous features does this language have
20:05:18 -!- int-e_ has quit (*.net *.split).
20:05:19 -!- pikhq has quit (*.net *.split).
20:05:20 -!- drdanmaku has quit (*.net *.split).
20:05:24 -!- viznut has quit (*.net *.split).
20:05:27 -!- skarn has quit (*.net *.split).
20:05:30 -!- olsner has quit (*.net *.split).
20:05:32 -!- ^v has joined.
20:06:19 <elliott> vanila: that's just metanotation
20:06:52 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/pattern.unl
20:06:55 <vanila> i cant run this either :/
20:06:59 <vanila> well i run it but nithing happens
20:07:07 -!- drdanmaku has joined.
20:07:12 -!- tromp__ has joined.
20:07:29 <elliott> are you stripping comments?
20:07:32 <vanila> yes
20:07:53 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/prime_numbers.unl I have to support c if i want to run this.. but i dont know how to :(
20:07:57 <elliott> it doesn't use c or d at least
20:08:10 -!- viznut has joined.
20:08:10 -!- olsner has joined.
20:09:19 -!- fizzie` has changed nick to fizzie.
20:09:28 -!- AnotherTest has quit (Ping timeout: 258 seconds).
20:09:43 -!- AnotherTest has joined.
20:09:43 <vanila> maybe if i emit CPS terms i can
20:09:53 -!- tromp_ has quit (Ping timeout: 258 seconds).
20:10:00 -!- viznut has quit (Ping timeout: 244 seconds).
20:10:00 -!- olsner has quit (Ping timeout: 244 seconds).
20:10:04 <vanila> http://lpaste.net/114354 so far
20:10:08 <vanila> c is wrong
20:11:01 -!- tromp has quit (Ping timeout: 258 seconds).
20:11:06 -!- tromp_ has joined.
20:11:48 -!- erdic has quit (Ping timeout: 258 seconds).
20:12:14 -!- newsham has quit (Ping timeout: 258 seconds).
20:12:23 -!- erdic has joined.
20:14:50 -!- shikhin has quit (Ping timeout: 258 seconds).
20:15:08 <Bike> can you rewrite this in cps
20:15:15 <Bike> the interpreter itself, i mean
20:15:22 <vanila> okay
20:16:02 -!- AndoDaan_ has joined.
20:16:32 -!- shikhin has joined.
20:17:26 -!- AndoDaan has quit (Ping timeout: 244 seconds).
20:17:32 -!- AndoDaan_ has changed nick to AndoDaan.
20:27:20 <vanila> http://lpaste.net/114354
20:27:31 <vanila> i changed it to CPS but not sure how to fix c case still
20:29:00 <vanila> i mixed up parameter order in definition of c but even corrected it s not right
20:29:32 <vanila> ohh
20:29:35 <vanila> no i fixd it fully now
20:29:37 <Bike> uh, this is the same paste as before
20:29:41 <Bike> ok, good! i guess!
20:29:58 <vanila> http://lpaste.net/114355
20:32:02 <vanila> The `dF function takes an argument Y and evaluates F, giving a function X, and returns the evaluation of `XY.
20:32:10 <vanila> now triangles works but I still can't run primes
20:32:17 <vanila> because i don't implement d
20:32:31 <vanila> i don't really understand it
20:32:53 <vanila> what would it look like in lambda calculus?
20:33:11 <elliott> vanila: `dX doesn't evaluate X
20:33:15 <elliott> despite unlambda being otherwise strict
20:33:29 <elliott> er, `dF doesn't evaluate F rather
20:33:33 <vanila> `dF = (lambda (y) (let ((x f)) (x y)) ?
20:33:34 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: dF: not found
20:33:45 <elliott> almost
20:33:53 <elliott> I think it's memoised
20:33:55 <elliott> not sure though
20:33:58 <elliott> I know there are some subtleties...
20:34:19 <elliott> oerjan would work
20:34:22 <elliott> vanila: note that you have to be able to do, like
20:34:25 <elliott> ``idF
20:34:25 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: `idF: not found
20:34:33 <elliott> so you can't just do d syntactically
20:34:38 <vanila> :(
20:34:38 <elliott> it can also be passed as a parameter or whatever
20:34:40 <vanila> i dont like d
20:34:51 <elliott> *do d
20:35:01 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/prime_numbers.unl
20:35:05 <vanila> holy *** thos uses a LOT of d
20:35:38 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/quine/
20:35:38 <vanila> haha
20:35:45 <vanila> there's more quines in unlambda than there are any other programs
20:37:10 <elliott> have you seen oerjan's self-interpreter?
20:37:44 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/CUAN/unlambda_interpreter.unl
20:37:45 <vanila> is this it?
20:38:20 <elliott> http://oerjan.nvg.org/esoteric/interpreter.unl
20:38:29 <elliott> maybe yours is the same thing but with fewer comments, I don't know
20:38:41 <vanila> wow his code looks really nice
20:38:44 <Taneb> @oeis 1 1 2 8
20:38:55 <lambdabot> 2^(n(n-1)/2).[1,1,2,8,64,1024,32768,2097152,268435456,68719476736,3518437208...
20:39:04 <elliott> it has a nice character table
20:39:08 <Taneb> Hmm, that is not the sequence I have
20:39:16 <elliott> I guess that one you linked can't be complete because it doesn't
20:39:21 <Taneb> Next element is > 2E24
20:39:42 <elliott> uh, I mean ?
20:40:19 <Taneb> I guess what I'm doing is stupid, though
20:40:32 <elliott> vanila: http://oerjan.nvg.org/esoteric/ has an unlambda compiler
20:40:37 <elliott> I think it's hard to properly compile because of d though
20:40:44 <vanila> d is mean
20:41:25 <elliott> well, I guess it's essentially just call-by-name (or -need?), but...
20:46:45 -!- skarn has joined.
20:50:26 -!- olls has quit (Quit: Lost terminal).
20:55:32 <Vorpal> Well fuck sheepshaver, I was trying to use a browser in it. That crashes sheepshaver
20:55:46 <Vorpal> Both Netscape Navigator and Internet Explorer 5
20:55:58 <Vorpal> vanila, ^
21:03:27 -!- MoALTz has joined.
21:04:46 -!- MoALTz_ has quit (Ping timeout: 250 seconds).
21:04:55 <vanila> sheepshaver is so weird howitwants you to identity map low memory area
21:05:05 <vanila> you explicitly cant do that for security reasons...
21:05:09 <vanila> so its dodgy that they want it
21:05:23 <vanila> of course its something weird to do with how the emulatorworks butstill
21:05:26 <vanila> i just dont ilk eit
21:05:28 -!- MoALTz_ has joined.
21:05:33 <vanila> i wishthere was good info about how to emulate powermac
21:05:35 <vanila> it seems impossible
21:05:43 <vanila> at least without being a hardware RE exper
21:05:43 <vanila> t
21:05:51 <Vorpal> vanila, your spacebar is partially broken
21:07:15 <fizzie> I've used SheepShaver "succesfully".
21:07:32 <Vorpal> fizzie, oh yes, for a short while
21:08:12 -!- MoALTz has quit (Ping timeout: 272 seconds).
21:08:33 <elliott> it's probably easiest to just buy an old mac off ebay
21:08:40 -!- ^v has quit (Quit: http://i.imgur.com/Akc6r.gif).
21:09:08 * pikhq_ has a working Famicom now. Wheee.
21:09:11 <elliott> vanila: is it for security? I thought it was to just make sure *NULL segfaults
21:09:29 <Vorpal> elliott, or repair the power connector of my old first model ibook
21:09:57 <elliott> that was to vanila too
21:10:00 <Vorpal> elliott, seen ais recently?
21:10:01 <Vorpal> Oh
21:10:23 <fizzie> There's always qemu-system-ppc too.
21:10:25 -!- MoALTz has joined.
21:10:28 <vanila> I think that kernel data must live there
21:11:26 <fizzie> I bought an old ppc mac off the-Finnish-version-of-eBay (and then sold it again).
21:11:41 -!- AnotherTest has quit (Ping timeout: 240 seconds).
21:12:05 <vanila> 'Allowing processes to map low values increases the security implications of a class of defects known as "kernel NULL pointer dereference" defects'
21:12:20 <elliott> Vorpal: less than two months ago
21:12:25 <elliott> 2014-09-24
21:12:26 <vanila> I thought there was more to it than that, but I guess it's just what you said
21:12:29 <elliott> don't you keep logs? :p
21:12:39 -!- MoALTz_ has quit (Ping timeout: 255 seconds).
21:12:39 <vanila> ais was here yesterday
21:12:47 <elliott> oh
21:12:48 <vanila> and I want to talk to zzo
21:13:01 <elliott> okay yeah he's just been using different nicks
21:13:17 <elliott> vanila: I guess that's getting the kernel to dereference a NULL pointer unwittingly?
21:13:22 <elliott> and then it panics or something?
21:13:30 <elliott> weird
21:13:53 <Vorpal> elliott, that is the gist about the 0-page protection yes
21:13:56 -!- MoALTz_ has joined.
21:15:53 -!- AndoDaan has quit (Quit: Leaving).
21:16:26 -!- MoALTz has quit (Ping timeout: 272 seconds).
21:30:33 <vanila> I want to design a simple lambda language to practice writing a compiler
21:31:07 <vanila> not sure how to design it, I thought about giving it simple types and data type definitions, but work.. :/
21:31:09 -!- Patashu has joined.
21:35:06 <b_jonas> vanila: do you want a compiler specifically, not an interpreter or something?
21:39:43 -!- AnotherTest has joined.
21:44:09 -!- AnotherTest has quit (Ping timeout: 255 seconds).
21:48:24 -!- AndoDaan has joined.
21:49:32 <paul2520> vanila: I'd recommend checking out Build Your own LISP: http://buildyourownlisp.com/
21:50:08 -!- ZombieAlive has joined.
21:51:10 <vanila> thanks
22:19:38 -!- newsham has joined.
22:21:32 <Taneb> I feel like I ought to learn OpenCL or something
22:22:37 <nys> lisp in small pieces
22:22:41 <nys> is a good book
22:25:31 <int-e> Chapter 1. The opening parenthesis.
22:27:55 <Phantom_Hoover> nys, the GPGPU framework, not the lisp
22:28:42 <nys> no no i was talking in reference to paul2520's earlier recommendation
22:28:43 <elliott> there... isn't a lisp called OpenCL, as far as I know
22:28:49 <elliott> nys was replying to vanila
22:28:54 <int-e> @metar LOWI
22:28:54 <lambdabot> LOWI 152220Z AUTO 06007KT 020V100 9999 -SHRA FEW050 BKN060 12/07 Q0998
22:28:57 <nys> yes
22:29:19 <elliott> ideally someone would write an open-source common lisp implementation that runs on GPUs called OpenCL for maximum confusion
22:31:33 <vanila> I find lisp in small pieces hard
22:31:44 <nys> i suppose it is
22:31:45 <Bike> it's all the denotational semantics.
22:31:48 <nys> it's nice and deep though
22:32:08 <Bike> anyone here have experience with weird japanese text encodings
22:37:40 <vanila> what should i call my lambda calculus language
22:37:46 <vanila> pure, no mutation
22:38:47 <Taneb> Eel
22:38:56 <vanila> thanks :)
22:38:59 <nys> Fat Calculus
22:40:29 <fizzie> While I was fiddling with things, enabled https:// for esolangs.org (at least experimentally).
22:40:33 -!- Primal has joined.
22:40:51 <Primal> Ok mobileirc is awful
22:41:34 <elliott> fizzie: nice!
22:41:50 <fizzie> elliott: Next, a Gopher interface.
22:42:08 <elliott> btw we really need a new featured language
22:42:22 <Bike> i vote whatever vanila's doing
22:42:49 <Primal> Batch
22:43:00 <Primal> Jk
22:43:06 <elliott> shubshub, is that you
22:43:11 <vanila> I would help to make a gopher version of esolangs
22:43:17 <vanila> that is a nice idea
22:43:19 <vanila> read only
22:43:29 <elliott> fizzie: nice overkill RSA key
22:43:34 <Primal> Someone make an app or something
22:44:07 <elliott> does startcom not have an SHA-2 intermediate cert?
22:44:13 <elliott> :/
22:44:15 <fizzie> Hmm.
22:44:31 <elliott> also, you have SSL3 enabled
22:44:37 <elliott> *v3
22:45:00 <elliott> do you know about https://wiki.mozilla.org/Security/Server_Side_TLS
22:45:22 <elliott> it's offering like 3DES and stuff right now :p
22:45:22 <fizzie> Yes, I copied the zem.fi SSL/TLS settings from there.
22:45:25 <vanila> $ zsync http://esolangs.org/dump/esolang.xml.zsync
22:45:29 <vanila> is this up to date
22:45:45 <fizzie> I'll see about fixing the esolangs nginx config too.
22:45:50 <elliott> fizzie: it's updated recently though, since poodle
22:45:54 <vanila> ok if i managed to get this as a gopher will it be hosted?
22:45:56 <elliott> well I guess 3DES is actually only disabled on the "modern" one there, sigh
22:46:27 <elliott> fizzie: also it looks like you're on an old openssl
22:46:51 <FireFly> TIL about zsync. seems useful
22:46:58 <FireFly> `thanks vanila
22:46:59 <HackEgo> Thanks, vanila. Thanila.
22:47:07 <elliott> or at least SSL labs claims you have exploitable CVE-2014-0224 (?!) and no TLS_FALLBACK_SCSV (but just disabling SSLv3 is better)
22:47:40 <fizzie> elliott: I haven't really been keeping the rest of the server updated (it's kind of outside the scope of my mandate), but I guess I could install security updates from the repository, at least.
22:47:59 <elliott> I don't think Gregor cares about that server :p
22:48:39 <vanila> any thoughts about gopher
22:48:52 <elliott> can you do gopher over TLS
22:48:59 <vanila> i dont know
22:48:59 <FireFly> I'm sure zzo38 would approve of gopher access
22:49:10 <elliott> also all of this is kind of irrelevant without HSTS of course, but presumably that comes after
22:49:58 <vanila> yes
22:49:59 <fizzie> I was thinking more like a MediaWiki extension that'd host a live Gopher (read) view, not a snapshot.
22:50:02 <fizzie> Not sure if there is one.
22:50:06 <vanila> I like zzo's gopher page
22:50:12 <fizzie> It sounds like the kind of thing someone would have written, but perhaps not.
22:50:25 <fizzie> (If there is, it's probably horrible.)
22:50:36 <vanila> fizzie, I was thinking of writing such a converter so that it could be hosted with a normal gopher server
22:51:48 <Sgeo> Uh. Why is Google telling me that AIM may harm my browsing experience?
22:52:13 <fizzie> I don't know anything about how Gopher servers deal with dynamic content, but if they do, it could query the live pages via the MediaWiki API, of course.
22:53:07 <vanila> hmm
22:53:12 <vanila> that sounds better because it's up to date
22:53:16 <vanila> but its harder too
22:54:06 <Primal> I need a new Python compiler that can still loop in JavaScript++
22:54:47 <Primal> Any ideas
22:55:52 <elliott> thankfully the wiki has been replaced by a 502 Bad Gateway page so it's really easy to make a gopher version
22:56:50 <fizzie> aptitude had stalled in the middle of a PHP update, because it wanted to ask me a question.
22:57:04 -!- AnotherTest has joined.
22:57:23 <elliott> I wonder if the wiki was vulnerable to shellshock because of these upgrade practices :p
22:57:42 <fizzie> Hmm.
22:57:43 -!- AnotherTest has quit (Remote host closed the connection).
22:57:54 <fizzie> It's still 502 Bad Gateway even after finishing. Curious.
22:58:10 <elliott> maybe php-fpm didn't restart properly or something
22:58:46 <Primal> Yeah it's still a bad gateway
22:59:20 <fizzie> "connect() to unix:/var/run/php5-fpm.sock failed (13: Permission denied)"
23:00:03 <int-e> hehe. "*tl;dr* It's complicated."
23:00:50 <FireFly> Huh, gopher seems pretty simple
23:01:16 <fizzie> Oh, the default listen.mode for the socket has been changed from 0666 to 0660.
23:01:39 -!- Ethereal_ has joined.
23:01:50 <fizzie> Should be back in business.
23:02:08 <Primal> Let me check
23:02:26 <fizzie> (If not, then that's tough luck, since I'll be away for a few minutes.)
23:02:27 <Primal> Yuh
23:02:53 <Primal> It works
23:04:45 <Primal> My Unreliable past is a weird name for a language
23:04:47 <vanila> so i have to parse XML and then translate mediawiki syntax to gopher...
23:04:51 <vanila> seems too hard :(
23:05:55 <FireFly> "translate mediawiki syntax to gopher"?
23:05:59 <FireFly> Hm
23:06:32 <elliott> fizzie: also, you don't seem to have forward secrecy(? maybe? SSL Labs says something about "with some browsers" but also that it doesn't work with their "reference browsers".)
23:09:20 <elliott> I really don't know why it would be saying that...
23:09:56 <Primal> I need more money for a new VPN but .. I need this money for the fobs
23:10:15 <elliott> it's using ECDHE for me
23:11:13 -!- Ethereal_ has quit (Quit: I Burnt my food).
23:12:44 <Sgeo> Installing new Pidgin even though Pidgin keeps dying. Maybe someone's attacking me :/. Even if not and it's just Pidgin sucking majorly, why keep running insecure stuff
23:13:47 <vanila> im sory i dont think I will do this
23:18:05 <int-e> elliott: is it saying that because some non-DH configurations are accepted? Like AES256-SHA...
23:18:48 <elliott> int-e: I guess. it would be weird to say that there's no PFS with "reference browsers" if those reference browsers can't do DH, but I guess maybe they can and just don't negotiate it with the current settings?
23:18:58 <int-e> ( http://sprunge.us/PMYV )
23:19:01 <idris-bot> (input):1:5: error: expected: "!!",
23:19:02 <idris-bot> "$", "$>", "&&", "&&&", "*",
23:19:02 <idris-bot> "***", "+", "++", "-", "->",
23:19:02 <idris-bot> ".", "/", "/=", ":+", ":-",
23:19:02 <idris-bot> "::", ":::", ":=", "<", "<$",↵…
23:19:08 <int-e> idris-bot: sorry.
23:19:34 <elliott> yes, the cipher configuration is a mess.
23:20:01 <int-e> surely 90% of those combinations have no right to exist?
23:20:02 <Bike> broken debuggers: the best or the bestest
23:20:22 <elliott> fizzie: https://forum.startcom.org/viewtopic.php?f=15&t=15742 SHA-2 intermediate certificates exist, happily
23:21:01 <Primal> U.S. xabber or jappix if you are looking for an xmpp client
23:21:08 <Primal> Use•
23:29:28 <Primal> Brb
23:29:41 -!- Primal has quit (Quit: Page closed).
23:29:59 <fizzie> elliott: Crafty. All the links in their instructions pointed at the SHA-1 hashed versions.
23:30:40 <elliott> fizzie: I don't think you understand. TLS is meant to be painful.
23:33:08 <elliott> incidentally you're welcome to give me root if that would be less annoying than me bothering you about this :p
23:34:01 <fizzie> "The size of the prime number p constrains the size of the pre-master key PMS, because of the modulo operation. A smaller prime almost means weaker values of A and B, which could leak the secret values X and Y. Thus, the prime p should not be smaller than the size of the RSA private key."
23:35:11 <fizzie> (For the DH parameter thing.)
23:35:42 <elliott> oh, the problem is you have a 4096-bit RSA key?
23:35:46 <fizzie> Yes.
23:35:58 <fizzie> openssl dhparam 4096 actually said, and I quote, "This is going to take a long time".
23:36:11 <elliott> 4096 bits probably isn't buying you anything but a very slight slowdown for a year-long key if you have PFS anyway
23:36:42 <fizzie> Yes, but I can't switch it to anything shorter until the certificate expires.
23:36:46 * elliott nods
23:37:03 <elliott> certificates are such a mess
23:38:09 <elliott> so wait, what is it doing right now then?
23:38:23 -!- ^v has joined.
23:40:08 <fizzie> I don't know. I didn't configure a DH parameter file. Maybe it didn't advertise any EDH ciphers?
23:40:34 <fizzie> "openssl dhparam" has now filled one screen with .s and +s.
23:40:45 <elliott> I get "The connection is encrypted and authenticated using AES_128_GCM and uses ECDHE_RSA as the key exchange mechanism.", fwiw.
23:40:52 <fizzie> Hm.
23:41:21 <fizzie> Well, it finished, anyway, so I'll give it the generated parameters.
23:41:30 <elliott> I hope these aren't as slow to use as they are to generate.
23:42:43 -!- Dulnes has joined.
23:43:36 <Dulnes> The cookies I've burned them
23:46:56 <Dulnes> Hello
23:48:36 <elliott> fizzie: I believe you can fix the SHA-2 thing without getting a new certificate, since that part isn't signed, by the way
23:49:27 <fizzie> elliott: Yes, that should be fixed now.
23:49:50 <fizzie> elliott: ssllabs test says overall A (subscores 100, 95, 100, 90).
23:50:02 <elliott> ah, cache was getting in the way
23:50:50 <fizzie> "Forward Secrecy: Yes (with most browsers) ROBUST" now too.
23:51:03 <fizzie> I didn't enable OCSP stapling, because the nginx version was too old for that.
23:51:34 <fizzie> And didn't toggle on HSTS yet either, since, well. The canonical address is still the http:// one, anyway.
23:51:40 <elliott> I think probably nobody cares about OCSP stapling
23:51:47 <elliott> since I don't think anyone does revocation checking right now?
23:51:58 <elliott> fizzie: heh, and I was just typing: thanks for the great work! ok, now how long until I can bug you to make it official and add preloaded HSTS? :p
23:52:18 <Dulnes> Well.. I know a guy who still does
23:52:23 <elliott> I wonder why mozilla's cipher configuration includes CAMELLIA suites.
23:52:29 <InvalidCo> whoops
23:52:36 <elliott> weird japanese clients, maybe?
23:52:37 <InvalidCo> I didn't realize that dwarf had turned into my pet
23:53:11 <Dulnes> palemoon and waterfox are nice to play with
23:53:40 <elliott> I wonder what you need to do to get A+
23:53:55 <elliott> oh, maybe it checks HSTS
23:54:30 <InvalidCo> "the book was coated with contact poison!"
23:54:31 <InvalidCo> what?
23:54:57 <b_jonas> InvalidCo: that's one of the failure to read effects
23:55:01 <InvalidCo> ah
23:55:05 <fizzie> Also only got 90 in the "cipher strength" category, possibly because I went with the "intermediate" list.
23:55:21 <InvalidCo> apparently was trying to read the spellbook of cancellation
23:55:36 <InvalidCo> wait, this isn't nethack
23:55:36 <InvalidCo> :D
23:55:38 <InvalidCo> whoops
23:55:48 <InvalidCo> s/net/#net/
23:56:12 <myname> this seems to happen regularly
23:57:27 <quintopia> apparently the intro comp media class as georgia tech involves the entire class writing programs in Chef
23:57:38 <elliott> fizzie: yeah, going with the modern one would be nice but would rule out firefoxes older than feb 2014, chromes older than sep 2012, IEs older than oct 2013, opera older than july 2013, safaris older than june 2013, etc.
23:57:44 <quintopia> i'm sitting across from the poor sap who has to grade them
23:57:44 <elliott> which isn't a great combination with https-only
23:58:08 <elliott> still... why camellia?
23:58:11 -!- Patashu has quit (Ping timeout: 240 seconds).
23:58:19 <elliott> I should figure out why they have that in there.
23:59:19 <fizzie> elliott: Let's see if the 'pert adds DNSSEC glue too. (I asked them to update the nameserver list -- switching VPSes -- and added a postscript about optionally doing that. For the one person with a DNSSEC resolver. And of course those whose DNS it will break as a collateral damage due to overly long DNS replies or whatnot.)
23:59:34 <elliott> the, uh, 'pert? oh.
23:59:38 <b_jonas> quintopia: huh what...
23:59:42 <elliott> fizzie: will you support DANE???
23:59:46 <b_jonas> chef?
23:59:50 <elliott> a whole half a person would benefit from that
←2014-11-14 2014-11-15 2014-11-16→ ↑2014 ↑all