00:00:03 <shachaf> well, i imagine that you can do something like for any Alexandrov space, more generally
00:00:38 <shachaf> since every Alexandrov space corresponds to a preörder and every preörder has a dual
00:21:20 -!- Phantom_Hoover has joined.
00:22:47 -!- nooodl has quit (Quit: Ik ga weg).
00:22:57 -!- nooga has quit (Read error: Operation timed out).
00:25:03 -!- augur has quit (Remote host closed the connection).
00:32:33 -!- Sgeo has joined.
01:30:14 -!- copumpkin has joined.
01:41:18 -!- augur has joined.
01:41:27 -!- augur has quit (Remote host closed the connection).
01:41:53 -!- augur has joined.
01:42:49 -!- augur_ has joined.
01:43:07 -!- augur has quit (Read error: Connection reset by peer).
01:44:11 -!- Bike has joined.
02:20:36 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
02:34:31 -!- yorick has quit (Remote host closed the connection).
02:54:47 -!- Sorella has quit (Quit: It is tiem!).
03:09:05 -!- Sprocklem has joined.
03:19:12 -!- impomatic has quit (Read error: Connection reset by peer).
03:59:06 -!- gdjfgh has quit.
04:02:20 * ski . o O ( <http://en.wikipedia.org/wiki/Pontrjagin_duality> )
04:02:33 <ski> oerjan : have you seen "Topology Via Logic" by Steven Vickers ?
04:03:45 <Sgeo> "It will either prove or disprove, or maybe neither."
04:03:50 <Sgeo> http://www.cnn.com/2014/01/15/showbiz/justin-bieber-vandalism-probe/index.html?hpt=hp_t3
04:05:27 <oerjan> that should about cover it, then
04:10:43 -!- mauke has quit (Disconnected by services).
04:10:53 -!- mauke has joined.
04:13:12 -!- preflex_ has joined.
04:14:27 -!- preflex has quit (Ping timeout: 260 seconds).
04:14:35 -!- preflex_ has changed nick to preflex.
04:32:22 -!- Sellyme has quit (Excess Flood).
04:33:03 -!- Sellyme has joined.
04:40:07 -!- Sprocklem has quit (Ping timeout: 260 seconds).
04:55:26 <ais523> ski: is that the same Steve Vickers who works at my department?
04:55:40 <ais523> if so, it's weird seeing people casually namedrop my colleagues, /again/
04:58:43 <ski> ais523 : i don't know what department you're at, but it's the Vickers at <http://www.cs.bham.ac.uk/~sjv/papersfull.php#TVL>
05:04:07 <oklopol> maybe RE = open fits better because if you have a union of RE sets such that the turing machines giving them are enumerated by a turing machine, then the union is RE
05:05:33 <oerjan> @tell shachaf see oklopol in logs
05:09:44 <ais523> hmm, my Turing-completeness-without-player-action proof for Magic: the Gathering is almost complete
05:10:10 <ais523> any Magic fans (b_jonas?) here happen to know a way to cause an arbitrary creature to stop being a creature, while keeping its rules text around?
05:10:44 <oklopol> when you discuss for example the full shift S^\Z for a finite set S (that is, the cantor space), you want to talk about open sets which are somehow computable, and closed sets which are somehow computable
05:10:55 <oklopol> what you do is say you have an "RE open set"
05:11:19 <oklopol> which means that a turing machine outputs words w such that every word with central pattern w is in your set
05:11:27 <oklopol> and you have coRE closed sets
05:11:40 <oklopol> by outputting words w such that no word with w in the center is in your set
05:13:26 <oklopol> well i don't know what percentage of humans would say that, but you certainly could say that
05:13:48 <ais523> actually I think you can just turn it into a copy of a licid until end of turn, but that doesn't really work for what I want as then it gets enchant creature, which I don't want either
05:13:59 <ais523> basically the idea is that it needs to be able to survive infinitely many -1/-1 effects
05:16:00 <Sgeo> o.O at Unexpected Results
05:20:50 <Sgeo> "You can concede a game while Platinum Angel on the battlefield. A concession causes you to leave the game, which then causes you to lose the game (Once you concede, you no longer control a Platinum Angel, so its ability can't prevent you from losing the game)."
05:21:05 <Sgeo> Why did there need to be a ruling on this?
05:21:18 <Sgeo> As in, did someone concede then dispute the concession?
05:22:59 -!- oerjan has quit (Quit: You're asking whether geeks did something ridiculously pedantic?).
05:23:18 <ais523> yeah, I like oerjan's explanation for this
05:24:51 <quintopia> ais523: do you want to review the language we discussed the other day before I post it to the wiki
05:25:51 <Sgeo> ais523: what's oerjan's explanation?
05:26:04 <ais523> Sgeo: see the quit message
05:26:09 <ais523> "(Quit: You're asking whether geeks did something ridiculously pedantic?)"
05:26:24 <ais523> quintopia: not sure I remember, what does NOMW stand for?
05:28:21 <ais523> you may as well just post it, and then I can look at it there
05:28:25 <quintopia> ais523: also are you working on refining the MTG TM? remove the requirement that a player has an option?
05:28:37 <ais523> I'm trying to remove phasing altogether
05:28:45 <ais523> via inventing an esolang that maps more directly than a TM does
05:29:02 <ais523> actually, I guess just giving these creatures all creature types would work
05:29:16 <quintopia> can you give a creature all creature types?
05:29:27 <quintopia> artificial evolution only changes the creature type right
05:29:35 <ais523> it was a mechanic in lorwyn block
05:29:47 <ais523> I'm pretty sure at least one card granted it to arbitrary creatures
05:30:57 <ais523> actually there's an entire cycle
05:31:13 <ais523> also amoeboid changeling, which has it as a tap ability
05:31:26 <coppro> I could only remember the red one, plus nameless inversion
05:32:26 <quintopia> you clearly know mtg much better than me
05:32:31 <ais523> anyone know how many creature types there are offhand? this construction needs like 200
05:38:38 <Sgeo> http://www.reddit.com/r/magicTCG/comments/1vc7ot/if_a_future_civilization_found_a_complete_set_of/
05:39:40 <oklopol> so can you design a magic deck such that with probability 1, no matter what the players do, a particular computation emerges
05:39:43 <ais523> Sgeo: the "what happens if you play Floral Spuzzem and you're using that Un-card that removes errata" has become something of an inside joke
05:39:48 <oklopol> i guess they can just say "okay i give up"?
05:40:01 <oklopol> but other than that, how likely is it that you can do that
05:40:03 <ais523> I believe the current opinion is that the Floral Spuzzem gets a loss for slow ply
05:40:15 <ais523> oklopol: they could also not play any spells
05:40:21 <ais523> there aren't any cards that force you to play them
05:40:30 <ais523> no could there be, reasonably, it'd be impossible to enforce
05:44:15 <ais523> because nobody but the person holding them would know that they'd drawn them
05:44:29 <oklopol> there are many card games with such rules though
05:44:50 <oklopol> it feels really wrong but works in practice
05:45:06 <ais523> the Magic designers are really worried about that sort of cheating, for some reason
05:45:20 <ais523> they were worried even about designing cards that gave you a bonus if you played them immediately upon drawing them
05:45:37 <ais523> or cards that let you look at a set of cards from your deck without changing their order
05:46:00 <ais523> both of which are enforceable by the opponent, but only if they pay attention to how your hands are moving
05:46:31 -!- conehead has quit (Quit: Computer has gone to sleep.).
05:46:55 <oklopol> so, i need to do some linear algebra homework, invert matrices and stuff
05:47:58 <oklopol> also yesterday my gf was watching and kept saying "why did you write 99 there btw?"
05:48:14 <oklopol> the worst part is she was always right and it was supposed to be something else
05:48:49 <oklopol> i'm sure i have much to teach these people
05:49:56 <oklopol> like, "why did you write 99?"
05:50:13 <oklopol> see what i did was realize that i am interested in the span of these vectors
05:50:26 <oklopol> and i know that if i sum these vectors like this, then the span does not change, but
05:50:34 <oklopol> "NO NO i mean 90 + 5 is 95, not 99"
05:51:13 <Bike> http://www.smbc-comics.com/index.php?db=comics&id=2039#comic
05:52:43 <oklopol> (she does know matrices though, i just lecture to everyone equally)
05:58:59 <Sgeo> Both Factor and Rebol look backwards sometimes, and forwards other times
05:59:11 <Sgeo> I'm pretty sure both prefix-only and postfix-only suck
06:03:48 <quintopia> ais523: http://esolangs.org/wiki/Not_The_Main_Worb
06:03:58 <ais523> quintopia: I like the name :-)
06:04:41 <ais523> also, that's a really good reason to have image-based source code
06:06:42 <ais523> quintopia: it's not Turing-complete unless you give it infinite memory somehow
06:07:05 <ais523> say by tiling the plane with it, with different tiled copies being taken to have different labels
06:07:22 <ais523> you said you could construct TC families
06:08:09 <ais523> I do think that construction at the bottom proves it NP-hard to determine whether a stable state exists for any given NTMW program, though
06:08:15 <quintopia> ais523: which implies that one is allowed to generate circuits with a limited TM, yes?
06:08:32 <ais523> quintopia: yeah but TCness doesn't work like that
06:08:52 <ais523> being able to produce arbitrary circuits is proof of NP difficulty, though
06:09:04 <quintopia> ais523: I'll change it to LBA then. happy?
06:09:08 <ais523> what's more interesting is whether that's PSPACE-complete
06:09:18 <ais523> quintopia: no because it doesn't take input, that's part of the definition of LBA
06:09:41 <quintopia> ais523: we can define input size as "number of raw sources and sinks"
06:09:50 <ais523> there's a whole different set of complexities for things with finite memory, like that finite state machine
06:09:53 <ais523> that's code size, not input size
06:10:09 <quintopia> ais523: they function as input to the circuit
06:10:30 <quintopia> why do i keep pinging you no one else is talking
06:10:45 <ais523> because they act deterministically, there's no input being received from outside
06:10:52 <ais523> and if there /were/, you still wouldn't have an LBA
06:11:03 <quintopia> there's no input being received from outside when a TM starts with its input on the tape also
06:11:04 <ais523> because the amount of data storage you have is not proportional to the size of the input you've received
06:11:16 <ais523> Turing machines also aren't LBAs, though
06:11:33 <ais523> the LBA restrictions of Turing machines require you to implement a separate input mechanism
06:11:41 <quintopia> i'm just saying your objection to calling parts of the code "input" is invalid
06:12:03 <ais523> I'm saying you can't arbitrarily define some lines of code (or pixels of code, in this case) to be input
06:12:07 <ais523> because that makes no sense
06:12:18 <ais523> input isn't part of the code, by definition
06:12:41 <quintopia> why can't i draw a line across the image and say "everything on this line is input, everything else is code"
06:12:52 <ais523> now, you could define a preprocessor that compiled input into circuitry, then you probably would have an LBA
06:12:53 <oklopol> "<ais523> being able to produce arbitrary circuits is proof of NP difficulty, though" ?
06:13:15 <oklopol> maybe the circuit is something different than i imagine
06:13:48 <ais523> oklopol: because if you can create an arbitrary circuit out of AND/OR/NOT/delay (even random delay), you can create a binary counter and a check circuit
06:13:51 <ais523> and use it to solve 3SAT
06:13:56 <quintopia> or even pick out a set of pixels and say "these particular pixels are not code. they are input."
06:14:29 <oklopol> but usually if you can make arbitrary evaluate-only circuits, you get P completeness (you get any computation, but it's deterministic), or if the circuit has loops, you get PSPACE-hardness (because you can run any computation in the polynomial space that the logspace transducer gave you in the form of the circuit)
06:14:35 <ais523> quintopia: because it doesn't make sense, for a particular program, for its data storage to be proportional in the size of the input, because the input has a fixed size
06:14:41 <ais523> being a specific set of pixels
06:14:51 <oklopol> because you can also check QSAT
06:14:57 <ais523> oklopol: good, I was wondering if it was also PSPACE-hard
06:15:05 <ais523> but didn't know how to prove it
06:15:29 <ais523> quintopia: PSPACE-hardness is one of the highest computational classes an FSM can have, so I think the language is fulfilling its purpose pretty well
06:15:52 <oklopol> yeah circuits don't get higher than that
06:15:55 <quintopia> ais523: okay imagine that the circuit family definition expands in a particular way, so that every time you add one pixel of input, you add a fixed number of other pixels that serve that pixel's input?
06:16:18 <ais523> quintopia: then it's an LBA, agreed
06:16:27 <quintopia> i thought we were talking about circuit families all along here and now i find out you were talking about one particular circuit
06:17:04 <ais523> this is the "input via preprocessor" idea I explained above
06:17:33 <quintopia> you could call it that. or you could just call it "the definition of the circuit family"
06:19:00 <quintopia> anyway, what if adding a pixel of input required adding a circuit proportional in size to the depth of the circuit you're adding to and increasing the depth by 1, so that the circuit size grows quadratically with input size? what is that?
06:19:48 * oklopol mumbles something about logspace-uniform circuit families
06:20:19 <ais523> wow, so it seems that the strongest known "P≠NP"-like result in the normal chain of complexity classes is "NL≠PSPACE"
06:20:23 * oklopol has been reading too much complexity theory lately
06:20:28 <ais523> that's… not a very strong result
06:20:28 <quintopia> ais523: so should i just say it can "probably" allow the creation of PSPACE-complete circuit families?
06:20:40 <ais523> quintopia: you can say that it's PSPACE-hard
06:20:56 <ais523> actually, I think it's trivially PSPACE-complete
06:21:29 <oklopol> ais523: the thing is in the usual chain, there's only a "space steps" exponential jumps apart, and "time steps" exponential jumps apart.
06:21:30 <ais523> given that oklopol proved it PSPACE-hard and it obviously doesn't take more than nondetrminstic polynomial space to simulate
06:21:41 <oklopol> the thing is, we cannot compare space and time in a nontrivial way
06:22:03 <ais523> well PSPACE is know to be a subset of EXPTIME
06:22:22 <oklopol> however, we have stronger separation results than just that an adding an exponential amount of space gives you new languages
06:22:47 <ais523> I'm not very good at complexity theory
06:22:55 <ais523> like, I know the very basics, just not much more than that
06:23:21 <oklopol> well i know more than the basics, but it's more of a hobby for me
06:23:34 <coppro> ais523: I thought we at least had P != PSPACE
06:23:52 <oklopol> (although we're probably gonna publish some gadget stuff soon)
06:23:55 <ais523> coppro: not according to Wikipedia, apparently that's still an open problem
06:23:58 <coppro> and definitely have P != EXPTIME, right?
06:24:00 <oklopol> coppro: we don't have that
06:24:08 <oklopol> P != EXPTIME because both are time classes
06:24:12 <ais523> yeah, you have P≠EXPTIME
06:24:31 <ais523> P=PSPACE would be even more surprising than P=NP, of course
06:24:33 <quintopia> i know how to prove Savitch's Theorem and...I know Dick Lipton's theorem about how the polynomial hierarchy collapses, but I don't recall the proof since it's been so long since I was in his class.
06:24:48 <oklopol> P vs NP? no one knows because they are different types of classes. P vs PSPACE? no one knows because they are different types of classes. (except for trivial relations)
06:25:00 <oklopol> similarly for quantum stuff, but now there are no comparisons...
06:25:28 <oklopol> interestingly, randomness can be compared in nontrivial ways
06:25:37 <oklopol> (you come up with fun ways to approximate sizes of sets)
06:26:10 <coppro> There exists an oracle separation between BQP and BPP at least
06:26:17 <coppro> which is actually kinda surprising
06:26:36 <oklopol> yeah oracle separations exist for many things
06:27:53 <oklopol> i don't really know what BQP is
06:28:01 <coppro> BPP except with a quantum computer
06:28:12 <oklopol> but it's coming up soon in my complexity theory bedtime stories book
06:28:23 <coppro> huh, P_{CTC} (deterministic computations with closed timelike curves) = PSPACE
06:29:12 <oklopol> oh i guess that's basically what i said about circuits with self-loops
06:29:13 <quintopia> at the very least, it's obvious that P_{CTC} contains NP
06:30:22 <ais523> draughts is EXPTIME-complete? that's surprising
06:30:56 <quintopia> oklopol: i'm not your CAS. invert your own damn matrices.
06:31:03 <ais523> isn't that just linear programming?
06:31:29 <ais523> you could do it without matrices
06:31:34 <ais523> oklopol: Computer Algebra System
06:31:38 <ais523> like Maple or Mathematica
06:32:03 <oklopol> HE DOESN'T EVEN SPECIFY THE FIELD, HOW ARE THE STUDENTS SUPPOSED TO GUESS
06:32:16 <quintopia> ais523: doing elimination of large systems by hand is annoying. so is taking determinants. so is taking inverses. so, no, use a CAS.
06:32:19 <ais523> that particular problem, I'd guess it's rational numbers
06:32:47 <ais523> quintopia: that thing isn't large enough that I'd be annoyed at having to use elimination or substitution on it
06:32:49 <oklopol> quintopia: there's no fourth equation
06:32:53 <coppro> it's clearly the finite field of order 23
06:32:56 <oklopol> that's why this is problem 6 already.
06:33:38 <quintopia> oklopol: so your answer is supposed to be another equation?
06:34:02 <oklopol> "find solutions" probably means "give the solution set in a more explicit form"
06:34:11 <lambdabot> oerjan said 1h 28m 37s ago: see oklopol in logs
06:34:22 <oklopol> i just said something about RE vs open
06:34:30 <Bike> a good cas oughta be able to do that shit for ya
06:34:39 <Bike> or matlab or whateverthefuck
06:34:59 <Bike> making people do matrix computation by hand is pointless and cruel.
06:35:50 <quintopia> oklopol: the solution set should be a line right??????????????????????????????????
06:36:35 <oklopol> the typical solution will be 1-dimensional, yes
06:36:53 <oklopol> i have no idea about this particular one yet
06:37:35 <oklopol> it cannot be unique, but it can be also two or three dimensional in theory
06:39:37 <quintopia> but those equations LOOK LIKE THEY ARE PROBABLY INDEPENDENT
06:40:06 <quintopia> why are you reading this book oklopol
06:40:07 <Bike> this reminds me that i didn't know until a few months ago that solving systems of polynomials is A Hard Problem
06:40:23 <Bike> because, i don't know, i'm an idiot probably
06:40:29 <oklopol> if someone asks you if a matrix is invertible, you can always safely say yes
06:40:41 <Bike> i knew a fun fact about that...
06:41:01 <quintopia> oklopol: is the zero matrix invertible???????????????????
06:41:16 <oklopol> but you will encounter that during your lifetime with 0 probability, even if you live forever.
06:41:25 <shachaf> oerjan: Right, so it's more complicated than "finite union", I guess.
06:41:28 <oklopol> assuming you only look at real matrices
06:41:29 <Bike> i think it was like, if you have the set of matrices with normal random entries around zero, then the average matrix is singular and the singular matrices are measure zero
06:41:59 <quintopia> you meant "if someone asks you if a randomly selected real matrix is invertible..."
06:42:00 <Bike> i guess it probably doesn't matter if they're normal
06:42:22 <Bike> every time you analyze a joke a frog dies, quintopia, and they are an important part of many ecosystems.
06:43:01 <quintopia> was that a joke?????? oklohelpol!!!!
06:44:32 <Bike> the joke being that the singular matrices are generally measure zero, but you're still going to run into them because mathematicians don't sample uniformly, they look for the most pathological bullshit possible :p
06:44:48 <oklopol> yes, like the all zero matrix
06:44:57 <oklopol> where do they come up with that stuff!
06:45:00 <Bike> the bullshit matrix
06:45:34 <oklopol> ...it seems i have yet to solve my equation
06:52:46 <oklopol> (i got {(7,-1,3,0) + u(28,-9,1,1) | u \in \R\} as the set of solutions)
07:15:06 -!- b_jonas has quit (Remote host closed the connection).
07:26:31 -!- FreeFull has quit.
07:35:26 <Sgeo> Screw concatenativity, I want mixfix
07:36:54 <Sgeo> 1 > 2 if [ "Math broke" print ] [ "Math didn't break" print ]
07:37:29 <Sgeo> Words could take in both a past stack and a future stack
07:38:21 <myname> Sgeo: without the print it would actually look nice
07:39:11 <Sgeo> Hmm, what's wrong with the print? But this should be valid too
07:39:23 <Sgeo> 1 > 2 if [ "Math broke" ] [ "Math didn't break" ] print
07:40:02 <Sgeo> Hmm... if print took a future argument instead, this would kind of not work
07:40:16 <Sgeo> print 1 > 2 if [ "Math broke" ] [ "Math didn't break" ]
07:40:38 <Sgeo> print might be satisfied by the result of 1 > 2, which I don't want
07:40:59 <Sgeo> Thank you myname
07:41:12 <myname> just make booleans unprintable :p
07:41:20 <myname> who needs that anyways
08:03:55 <mroman> You could use paratheses
08:04:06 <mroman> print (1 > 2 if ["Math broke"] ["Math didn't break"]
08:04:41 <mroman> where stuff in parantheses are treated as regular arguments in any other language
08:04:53 <mroman> that way you can write 5 6 + (concatenativ/stackish)
08:05:07 <mroman> or + (5 6) or possibly even + (5) (6)
08:06:00 <mroman> i.e you delay the function and evaluate the stack of the expression in the parantheses first
08:06:03 <mroman> and then run the function
08:07:49 <fizzie> Uh, any ideas? I have this gstreamer command line that used to work; it involves a multifilesink location='frame%08 err, never mind, I see the problem.
08:08:54 <fizzie> (Was supposed to be frame%08d.jpg but had written frame%08.jpg instead -- it generated the files "frame (nil)g", "frame 0x1g", "frame 0x2g", and so on.)
08:17:12 -!- nooga has joined.
08:19:49 <fizzie> "giblib warning: couldn't load font yudit/12, attempting to fall back to fixed. giblib warning: failed to even load fixed! Attempting to find any font. feh ERROR: Couldn't load font to draw a message" huh
08:20:05 -!- mtve has joined.
08:33:58 <quintopia> lol, i just realized that the Karma page on the wiki still says it is not known whether it is Turing-complete
08:34:50 <Bike> fizzie: does it actually say 'feh'
08:34:52 <quintopia> and I just found a file on my computer from March 2009 that contains an implementation of BCT in Karma
08:44:53 -!- nooga has quit (Ping timeout: 248 seconds).
08:57:44 <ski> oklopol : sounds ok
08:58:32 <ski> (the RE and coRE stufF)
08:58:45 <shachaf> ski: Hmm, did you see the question I asked earlier about that?
09:00:09 <mauke> Bike: feh is a program
09:00:41 <mauke> % = start of format specifier
09:01:02 <mauke> 0 = set the zero flag
09:01:14 <mauke> 8 = set minimum field width to 8
09:01:26 <mauke> . = equivalent to .0, i.e. set precision to 0
09:02:01 <mauke> j = length modifier (the following format refers to an intmax_t/uintmax_t)
09:02:14 <mauke> p = void * pointer
09:03:15 <mauke> I'm pretty sure 0, .0, and j are invalid/undefined with p
09:03:50 <mauke> and passing an int where a void * is expected is also undefined behavior
09:04:12 <mauke> but what you end up with is integers printed as memory addresses as if by %8p
09:04:32 <mauke> oh, and the g is just a g
09:08:56 -!- nooga has joined.
09:10:24 <shachaf> mauke: Is if (argv[0] && argv[0][0]) { Prog = argv[0]; } a standard idiom for you?
09:13:20 <shachaf> I don't know. Does that exist anywhere else?
09:44:55 <oklopol> ski: i mainly meant that i've more often seen arithmetical hierarchy terminology used for this stuff
09:45:26 <oklopol> so like \Pi^0_1 instead of coRE
09:45:54 <oklopol> (not that there's a difference in meaning)
09:48:15 <ski> shachaf : no
09:51:40 <shachaf> it was https://plus.google.com/+ShachafBen-Kiki/posts/Bt4CxtttYxP
09:51:52 <shachaf> I don't know how much of that was brought up in the discussion here.
09:55:43 <ski> shachaf : perhaps you could repeat the question ?
09:55:46 * ski . o O ( re "split", <http://en.wikipedia.org/wiki/Section_%28category_theory%29> )
09:56:04 <coppro> someone point me at a language that we don't know if it's TC or not
09:56:47 <ski> a retract situation consists of one morphism `s : A -> B' and one morphism `r : B -> A', satisfying `r . s = id_A'
09:57:08 <ski> `r' is said to be a retraction (of `s'), and `s' is said to be a section (of `r')
09:57:12 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: r': not found
09:57:42 <ski> if we define `i = s . r', then `i . i = s . r . s . r = s . id_A . r = s . r = i', so `i' is then an idempotent
09:58:24 <ski> (in a retract situation, `r' is also known as a "split epi(morphism)/epic" and `s' as a "split mono(morphism)")
09:59:21 <ski> given any idempotent `i', if we can find `r' and `s' such that `r . s = id' (a retract situation), and `s . r = i', then `i' is called a split idempotent, we say that `i' splits
09:59:31 <ski> shachaf : perhaps that helps ?
10:01:19 <shachaf> That's yet another thing that's called "split", not the epi or the mono but their composition (in the other direction)
10:01:50 <shachaf> Oh, maybe it's being used as an adjective.
10:02:20 <shachaf> The epi and mono are "split" from the idempotent.
10:02:46 <ski> (that's as far as i understand it)
10:02:50 <shachaf> Anyway, the question in a rough form is in the Google+ post above.
10:03:27 <shachaf> I will probably delete the post later, as I do.
10:03:43 <shachaf> Someone commented so I'd feel a bit bad deleting it. :-(
10:06:24 <ski> shachaf : did you see Martín Escardó's "Synthetic topology of data types and classical spaces" in 2004-11 at <http://www.cs.bham.ac.uk/~mhe/papers/> ?
10:08:47 <ski> "Semidecidable property ≈ open set" is mentioned in chapter 1
10:09:25 <shachaf> Oh, well, I've heard that analogy before, sure.
10:10:58 <shachaf> Is that exactly what I'm trying to get at here?
10:13:08 <ski> "Given a family `p_i(x)' of properties, in order to observe that the disjunction `\/_i p_i(x)' holds it suffices to observe that one of the disjuncts `p_i(x)' holds. Hence arbitrary disjunctions of observable properties are observable. ..(some qualifications to that).."
10:13:56 <ais523> btw, everyone: http://esolangs.org/wiki/StackFlow
10:14:06 <ais523> not bad for 2 and a bit days' work :-)
10:14:23 <shachaf> Qualifications described in a later chapter in that PDF, I guess.
10:14:42 <shachaf> Thanks for the link, I'll probably read it tomorrow or something.
10:15:16 <coppro> ais523: what is "sufficiently large that the language remains Turing complete"? doesn't that need to be unbounded?
10:15:38 <ais523> coppro: no, it just needs to be large enough to implement at least one interpreter for a Turing complete language
10:16:19 <ais523> and I have one that uses 5 stacks, 20 symbols, 41 rules
10:16:33 <ais523> the program it interprets is input using the initial contents of the stack, which interpreters can't put bounds on
10:16:52 <coppro> also, I love the markdown
10:17:13 <oklopol> "This is a program implements a cyclic tag system in StackFlow."
10:17:37 <ais523> btw, it's untested because I don't have an interp yet
10:18:27 <ais523> incidentally, is StackFlow the first esolang designed for literate programming? it may be
10:19:21 <ais523> coppro: I'm not sure what made me think of Markdown-based syntax, but when I did, the opportunity was too good to pass up
10:28:41 * shachaf goes to sleep, should have gone 2 hours ago
10:29:39 -!- Bike has quit (Ping timeout: 272 seconds).
10:50:46 -!- MindlessDrone has joined.
10:51:04 <Sgeo> (About Time Walk)
10:51:06 <Sgeo> "Combos well with a deck."
10:51:44 -!- MindlessDrone has quit (Client Quit).
11:02:59 -!- MindlessDrone has joined.
11:16:53 -!- impomatic has quit (Ping timeout: 248 seconds).
11:56:38 -!- peapodamus has quit (Read error: Connection reset by peer).
11:58:41 -!- peapodamus has joined.
12:04:07 -!- Taneb has quit (Read error: Operation timed out).
12:04:11 -!- ski has quit (Read error: Operation timed out).
12:04:24 -!- ski has joined.
12:04:33 -!- Taneb has joined.
12:05:34 -!- qlkzy_ has changed nick to qlkzy.
12:10:19 <Taneb> In Mergesort, does it particularly matter how you divide the list?
12:11:22 <ais523> Taneb: no, although particularly stupid divisions (like always dividing into 1 element, and the rest) hurt the computational class
12:11:39 <ais523> *computational complexity
12:12:10 <Taneb> Okay, so just alternating elements is fine?
12:13:20 <ais523> a good way to think about it is this: if a sort algorithm cares what order the list was in originally, it's probably broken
12:14:04 <Taneb> I have a Haskell function called "notQuickSort", which, true to its name, is not quicksort
12:15:02 <ais523> there are a lot of functions that aren't quicksort
12:15:09 <fizzie> ais523: I wouldn't call a (comparison-)sort function that's O(n log n) worst-case but O(n) best-case for an already-sorted list "broken".
12:15:29 <ais523> fizzie: it's O(n²) worst case
12:15:41 <ais523> I wouldn't call that "caring", in that case
12:16:00 <ais523> in the sense of the output being different
12:16:16 <Taneb> ais523, I was aiming to write Quicksort, but I wasn't including the pivot except in the singleton list case
12:16:32 <ais523> so the pivots just disappear?
12:18:13 <Taneb> Yeah, it was me being stupid. It's fixed now, though
12:32:03 -!- ais523 has quit.
12:52:30 -!- nooga has quit (Ping timeout: 245 seconds).
12:55:54 -!- nooga has joined.
12:58:48 -!- boily has joined.
12:58:59 -!- metasepia has joined.
13:14:31 -!- yorick has joined.
13:16:32 -!- Sgeo has quit (Read error: Connection reset by peer).
13:28:50 -!- Sorella has joined.
13:58:01 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
14:13:33 <mauke> here's a php puzzle: http://paste.scsys.co.uk/291607?tx=on
14:38:55 <fizzie> "virtual bool aku::PhnReader::next_frame(): Assertion `m_current_frame >= m_cur_phn.start' failed."
14:42:54 <boily> fungot: aku PhnReader'nglui Cthulhu fhtagn?
14:42:54 <fungot> boily: m-- m-1 doesn't work?
14:43:15 <boily> fungot: no it doesn't, the m_cur_phn.start fails the assertion.
14:43:15 <fungot> boily: how big was the code
14:43:18 <fungot> boily: and sorry, no :) or :(? it would be
14:44:01 -!- epicmonkey has joined.
14:44:08 -!- epicmonkey has quit (Read error: Connection reset by peer).
14:44:10 <fizzie> (It's not all that big.)
14:44:27 <fizzie> I think it doesn't like a gap in this file, is all.
14:54:39 -!- impomatic has quit (Ping timeout: 272 seconds).
14:54:50 -!- mrhmouse has joined.
15:01:42 -!- conehead has joined.
15:06:24 -!- conehead has quit (Remote host closed the connection).
15:07:03 -!- conehead has joined.
15:15:17 -!- quintopia has quit (Ping timeout: 248 seconds).
15:15:34 -!- tertu has joined.
15:18:24 -!- nooodl has joined.
15:24:27 <FireFly> fungot: you should fix the code
15:24:36 <fungot> FireFly: give the short example, or a file:// url which refers to my pseudofilesystem.
15:25:22 <FireFly> I'm fairly sure fungot is sentinent
15:25:23 <fungot> FireFly: so cl and elisp have it, but i discovered that the library system is intended to provide a way to make
15:25:42 <FireFly> oh yeah, lisp is popular in AI research
15:32:35 <boily> FireFly: it is a well known fact that fungot is sentient. blame fizzie.
15:32:35 <fungot> boily: in soviet russia, crow scares you!!
15:33:31 <FireFly> fungot: in regular #esoteric, you scare me
15:33:32 <fungot> FireFly: and it was for computer science, and forcer isn't just going to ask you earlier. i mean even if he explains lots of dynamic libraries :) you already call two internal lambdas, just make your cffi and use mpi...
15:35:41 -!- Bike has joined.
16:17:53 -!- Sellyme has quit (Excess Flood).
16:20:34 -!- Sellyme has joined.
16:21:44 -!- Sellyme has quit (Excess Flood).
16:23:03 -!- Sellyme has joined.
16:28:20 -!- Sellyme has quit (Excess Flood).
16:29:34 -!- Sellyme has joined.
16:35:15 <fizzie> fungot: What, exactly, do you believe?
16:35:15 <fungot> fizzie: the compiler is happy, " as well" you only need one! :)
16:42:08 * boily gives fungot a loch ness plushie
16:42:09 <fungot> boily: well as i can't use a serialiser while it's already in use)
16:43:50 <mrhmouse> what are you serializing, fungot?
16:43:50 <fungot> mrhmouse: i'd like to have the thread stay alive and accept future messages, i get pretty fnord
16:44:45 <boily> pretty fnords. heh ^^
16:52:46 -!- Phantom_Hoover has joined.
16:52:51 -!- Phantom_Hoover has quit (Changing host).
16:52:51 -!- Phantom_Hoover has joined.
16:53:57 <boily> `ello Phantom_Hoover
16:54:29 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/bin/ello
17:00:45 <HackEgo> tesheaf gaya reprovemon egellorguer kalm gester dire warpe ostv fedeipenne
17:01:26 <mrhmouse> I feel like "tesheaf" should produce "teshellaf"
17:02:31 <mrhmouse> Ah, I see the issue. It's operating on the list all at once
17:03:22 <mrhmouse> `run for word in $(words 10); do ello $word; done
17:03:32 <HackEgo> holtykkellabellon \ appomellorstag \ fimello \ gaffenello \ albrevello \ lebranello \ wasschedello \ clepellosit \ damello \ acarello
17:04:11 <fizzie> fungot: It can get pretty fnord if you're trying to serialize a thread, yes.
17:04:12 <fungot> fizzie: am almost done downloading it. ( which i don't use it any more
17:04:51 <mrhmouse> fungot: then why are you downloading it?
17:04:51 <fungot> mrhmouse: it's _not_ a useful observation if the means of a lambda would mean having a regular expression srfi defined actual regular expressions, do it well
17:22:09 <mrhmouse> Are there any esoteric templating/markup languages?
17:22:30 <mrhmouse> Specifically, ones that translate to HTML?
17:26:02 <FireFly> fungot: downloading srfis, eh+
17:26:02 <fungot> FireFly: actually i think it's just so that i don't
17:26:14 <FireFly> fungot: oh, okay, not downloading SRFIs then
17:26:14 <fungot> FireFly: yay. i will correct this anyway, it should only work on a ppc, then? lists?
17:26:41 <FireFly> Pretty sure you're talking about srfi-1, fungot
17:26:41 <fungot> FireFly: just that when i tried
17:27:22 <HackEgo> Usage: words [-dhNo] [DATASETS...] [NUMBER_OF_WORDS] \ \ options: \ -l, --list list valid datasets \ -d, --debug debugging output \ -N, --dont-normalize don't normalize frequencies when combining \ multiple Markov models; this has the effect \ of making larger dataset
17:27:44 <HackEgo> valid datasets: --eng-1M --eng-all --eng-fiction --eng-gb --eng-us --french --german --hebrew --russian --spanish --irish --german-medical --bulgarian --catalan --swedish --brazilian --canadian-english-insane --manx --italian --ogerman --portuguese --polish --gaelic --finnish --norwegian --esolangs default: --eng-1M
17:28:04 <FireFly> `run words --german-medical 10 | xargs -n 1 ello
17:28:11 <HackEgo> beridiertellome \ akzellon \ gregelseptellom \ pertemototelloniumcin \ retrinmesierello \ silandsello \ lungseinfallello \ pellonsen \ helllyopiege \ krebello
17:28:16 <Vorpal> mrhmouse, I haven't heard of any. Not sure how you could go about creating an interesting one either. Interesting idea though.
17:39:32 -!- atrapado has joined.
17:44:18 <oklopol> party on mothafucka party on
17:44:20 <mauke> The party is now on.
17:45:14 <int-e> what was the ocular pattern for myndzi?
17:45:35 <int-e> (or firework, or whatever that flowery thing is supposed to be)
17:45:43 -!- FreeFull has joined.
17:46:15 <mauke> U+1F440 (f0 9f 91 80): EYES [👀]
17:46:21 <metasepia> --- Possible commands: dice, duck, echo, eval, fortune, metar, ping, yi
17:46:33 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: party: not found
17:46:40 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: celebrate: not found
17:48:08 <HackEgo> You get NOTHING! You LOSE! Good DAY sir!
17:49:07 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: rot0: not found
17:49:52 <int-e> `cat `type -p fsck`
17:49:56 <HackEgo> cat: `type -p fsck`: No such file or directory
17:50:47 <FreeFull> `ln -s /usr/bin/cat /hackego/bin/rot0
17:50:49 <HackEgo> ln: invalid option -- ' ' \ Try `ln --help' for more information.
17:50:53 <FreeFull> `run ln -s /usr/bin/cat /hackego/bin/rot0
17:50:55 <HackEgo> ln: creating symbolic link `/hackego/bin/rot0': No such file or directory
17:51:04 <FreeFull> `run ln -s /hackego/bin/rot0 /usr/bin/cat
17:51:05 <HackEgo> ln: creating symbolic link `/usr/bin/cat': Read-only file system
17:51:10 <int-e> `run cat `type -p fsck`
17:51:50 <int-e> methinks that will not terminate
17:52:12 <mauke> ln operand order is like cp
17:52:32 <int-e> `run ls -l /hackego/bin/rot256
17:53:06 <HackEgo> ls: cannot access /hackego/bin/rot256: No such file or directory
17:53:17 <int-e> `run ls -l `type -p rot256`
17:53:19 <HackEgo> lrwxrwxrwx 1 5000 0 9 Sep 25 13:06 /hackenv/bin/rot256 -> /bin/echo
17:53:33 <fungot> \o| c.c \o/ ಠ_ಠ \m/ \m/ \o_ c.c _o/ \m/ \m/ ಠ_ಠ \o/ c.c |o/
17:53:33 <myndzi> | c.c.c | ¯|¯⌠ `\o/´ | c.c.c | `\o/´ ¯|¯⌠ | c.c.c |
17:53:34 <myndzi> /| c.c |\ /< | | /< c.c >\ | |\|/< c.c /|
17:53:39 <int-e> FreeFull: tsk, you had the wrong directory.
17:54:25 <HackEgo> You get NOTHING! You LOSE! Good DAY sir!
17:56:00 <HackEgo> /hackenv/bin:/opt/python27/bin:/opt/ghc/bin:/usr/bin:/bin
17:56:08 <FreeFull> I have rot0 in there but HackEgo errors when I try to run it
17:56:12 -!- monotone_ has joined.
17:56:18 <int-e> `wl en de hippopotamus
17:56:23 <HackEgo> bash: rot0: command not found
17:56:30 <FreeFull> `run /hackenv/bin/rot0 /dev/urandom
17:56:32 <HackEgo> bash: /hackenv/bin/rot0: No such file or directory
17:56:34 <HackEgo> lrwxrwxrwx 1 5000 0 12 Jan 16 17:54 bin/rot0 -> /usr/bin/cat
17:56:36 -!- monotone has quit (Ping timeout: 252 seconds).
17:57:36 <FreeFull> Made it echo rather than cat anyway
17:58:23 -!- monotone_ has changed nick to polytone.
17:58:42 <HackEgo> Traceback (most recent call last): \ File "/hackenv/bin/wl", line 52, in <module> \ q = query(continue_id) \ File "/hackenv/bin/wl", line 44, in query \ response = urllib2.urlopen(url).read() \ File "/opt/python27/lib/python2.7/urllib2.py", line 126, in urlopen \ return _opener.open(url, data, timeout) \ File "/opt/python27/lib/
17:59:32 <HackEgo> My hovercraft is full of eels.
17:59:49 <int-e> (It was meant to do that.)
17:59:57 <int-e> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/bin/wl is a fun read
18:00:54 <int-e> FireFly: you have to find any wikipedia page with empty 'languages' block.
18:03:33 <HackEgo> You get NOTHING! You LOSE! Good DAY sir!
18:03:55 <int-e> `wl sv en "Pero micca"
18:03:56 <FireFly> `run wl sv en "Pero micca"
18:03:56 -!- conehead has quit (Quit: Computer has gone to sleep.).
18:03:58 <HackEgo> You get NOTHING! You LOSE! Good DAY sir!
18:03:59 <HackEgo> Traceback (most recent call last): \ File "/hackenv/bin/wl", line 52, in <module> \ q = query(continue_id) \ File "/hackenv/bin/wl", line 44, in query \ response = urllib2.urlopen(url).read() \ File "/opt/python27/lib/python2.7/urllib2.py", line 126, in urlopen \ return _opener.open(url, data, timeout) \ File "/opt/python27/lib/
18:04:15 <int-e> ``````````````````````````````````````````````````
18:04:16 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: `````````````````````````````````````````````````: not found
18:06:20 <HackEgo> /hackenv/bin:/opt/python27/bin:/opt/ghc/bin:/usr/bin:/bin
18:08:11 <HackEgo> env: PATH: No such file or directory
18:08:20 <FireFly> I guess you have to grep it
18:08:26 <FireFly> oh, right, it takes a command to run
18:09:16 <boily> int-e: huh? a special `thank case?
18:11:15 <int-e> `perl -e'$,=",";print @ARGV,"\n"' $PATH
18:11:17 <HackEgo> Scalar found where operator expected at -e line 1, near "'$,=",";print @ARGV,"\n"' $PATH" \ (Missing operator before $PATH?) \ syntax error at -e line 1, next char $ \ Execution of -e aborted due to compilation errors.
18:11:22 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: ̀thanks: not found
18:11:40 <int-e> `` perl -e'$,=",";print @ARGV,"\n"' $PATH
18:11:42 <HackEgo> /hackenv/bin:/opt/python27/bin:/opt/ghc/bin:/usr/bin:/bin,
18:13:26 <int-e> boily: Apparently, though I have no idea why.
18:13:42 <int-e> `` cat $(type -p thankjames)
18:14:33 <mauke> `` echo ${PATH//:/ }
18:14:36 <HackEgo> /hackenv/bin /opt/python27/bin /opt/ghc/bin /usr/bin /bin
18:16:08 -!- conehead has joined.
18:16:45 <HackEgo> bash: -c: line 0: unexpected EOF while looking for matching ``' \ bash: -c: line 1: syntax error: unexpected end of file
18:18:01 <HackEgo> bash: than: command not found
18:20:40 <HackEgo> info: Writing node (dir)Top... \ File: dir,Node: TopThis is the top of the INFO tree \ \ This (the Directory node) gives a menu of major topics. \ Typing "q" exits, "?" lists all Info commands, "d" returns here, \ "h" gives a primer for first-timers, \ "mEmacs<Return>" visits the Emacs manual, etc. \ \ In Emacs, you can click mouse bu
18:24:14 <HackEgo> man: can't open the manpath configuration file /etc/manpath.config
18:25:01 <FireFly> I guess it didn't like that
18:25:04 <HackEgo> Vim: Warning: Output is not to a terminal \ [1;24r[?25h[?8c[?25h[?0c[27m[24m[0m[H[J[?25l[?1c[2;1H[1m[34m~ [3;1H~ [4;1H~
18:25:06 <int-e> `cat /hackenv/share/maze.c
18:25:08 <HackEgo> main() { asm("xor %edi, %edi\n" "inc %edi\n" "mov %rsp, %rsi\n" "go: movl $0xb195e2, (%rsi)\n" "rdtsc\n" "and $1, %al\n" "add %al, 2(%rsi)\n" "mov %edi, %eax\n" "xor %edx, %edx\n" "mov $3, %dl\n" "syscall\n" "jmp go"); }
18:25:28 <FireFly> hackenv is cwd I'm fairly sure
18:26:45 * boily unsticks HackEgo with his trusty mapole
18:27:09 -!- Slereah has joined.
18:27:31 -!- FreeFull has quit (Ping timeout: 260 seconds).
18:27:57 -!- Slereah_ has quit (Ping timeout: 252 seconds).
18:28:30 -!- atrapado has quit (Ping timeout: 252 seconds).
18:28:30 -!- glogbackup has quit (Ping timeout: 252 seconds).
18:38:01 <metasepia> CYUL 161800Z 21006KT 15SM FEW030 FEW120 M01/M08 A2996 RMK CU1AC2 CU TR SLP147
18:38:56 <ion> http://0-media-cdn.foolz.us/ffuuka/board/tg/image/1336/15/1336155252526.jpg
18:41:34 -!- atrapado has joined.
18:51:01 <Slereah> I should go back to dorf fortress
18:51:33 <HackEgo> Tanea plays Minecrafs, Dware Fortresr, and lives in Hexhal.
18:52:54 <boily> `run sed -i s/Hexhal/Yorj/ wisdom/tanea
19:04:32 <boily> (I like the sound of “Yorj”. /jɔʁʒ/)
19:05:46 <Slereah> "Oh this looks like a great place to build a fortress!"
19:10:42 <boily> ~eval "With an updated cabal..."
19:13:44 -!- FreeFull has joined.
19:18:26 -!- augur_ has quit (Remote host closed the connection).
19:18:52 -!- augur has joined.
19:19:15 -!- FreeFull has quit (Ping timeout: 252 seconds).
19:23:33 -!- augur has quit (Ping timeout: 272 seconds).
19:26:10 -!- FreeFull has joined.
19:34:03 <int-e> "Minecrafs", is there a point to that typo?
19:34:24 <HackEgo> You do not have the clearance necessary to view this entry.
19:42:12 -!- FreeFull has quit (Ping timeout: 252 seconds).
19:47:39 -!- Phantom_Hoover has quit (Ping timeout: 276 seconds).
19:50:44 -!- Bike has quit (Ping timeout: 272 seconds).
19:52:08 -!- nchambers has joined.
19:56:39 -!- SirCmpwn has quit (Ping timeout: 260 seconds).
19:58:05 <FireFly> the same point as with the other typns in that line
19:58:12 -!- Bike has joined.
20:05:17 -!- augur has joined.
20:05:38 -!- heroux has quit (Ping timeout: 264 seconds).
20:10:01 <boily> typns typns typns typns typns... ♪
20:10:06 -!- heroux has joined.
20:20:28 -!- nchambers has changed nick to DTSCode.
20:30:42 -!- DTSCode has changed nick to DTSCode-Work.
20:35:35 <mrhmouse> boily: I can't help but read that as a Patapon chant
20:35:45 <metasepia> Patapon is a video game published for the PlayStation Portable handheld game console combining gameplay features of a rhythm game and a god game.
20:36:19 <mrhmouse> You do chants to spur your little force of warriors on in battle, or to give them commands such as "retreat".
20:37:38 <boily> I don't have a PSP :(
20:37:59 <boily> I want to control little characters by rhythming unpronounçable syllables!
20:38:50 -!- Sprocklem has joined.
20:39:47 <mrhmouse> boily: I don't own a PSP either, I've only seen the game and thought it interesting.
20:40:59 * boily scands “tnuctip tnuctip tnuctip tnuctipun”
20:51:52 -!- SirCmpwn has joined.
20:53:08 <FireFly> I don't own a PSP either, but I have an annoyingly long list of PSP games that I'd like to play
20:53:33 * FireFly swats Nippon Ichi for not making more games for Nintendo consoles
21:11:22 -!- MindlessDrone has quit (Quit: MindlessDrone).
21:14:41 -!- CADD_ has joined.
21:14:59 -!- CADD_ has quit (Client Quit).
21:16:11 -!- Bike has quit (Ping timeout: 265 seconds).
21:19:39 -!- FreeFull has joined.
21:24:10 <boily> ooookay... something weird happened to the metasepia logs. now everything is in bold light blue...
21:25:18 <FireFly> `run echo -e '\e[1;31mhi boily'
21:25:42 <boily> it worked. bold red.
21:26:00 <boily> `run echo -e '\e[0mnormality!'
21:26:13 <FireFly> Maybe you want to strip control sequences
21:26:25 <boily> maybe I really want to strip that stuff.
21:26:35 <boily> oh well. time to disappear in the Great Cold Outside.
21:26:41 -!- boily has quit (Quit: REFRIGERATED CHICKEN).
21:26:42 -!- metasepia has quit (Remote host closed the connection).
21:26:59 <mauke> `run echo $'\004aare you using irssi?\004a'
21:27:54 * FireFly wonders what ^D would be mnemonic for
21:28:10 <FireFly> oh, maybe some kind of extended colour thing?
21:28:20 <FireFly> considering it's one after ^C
21:28:57 <FireFly> Most of the other IRC formatting control-codes are mnemonic, so I figure ^D would be too
21:29:03 -!- Bike has joined.
21:30:14 <mauke> `run echo $'are you using \004</irssi\004g?'
21:31:30 <FireFly> I wish I was, now, since I'm curious what I'm missing out on
21:32:02 <mauke> \004</ is bold red
21:33:00 <int-e> well, I'm seeing red ^D for 004 :)
21:34:10 <int-e> using irssi with http://int-e.eu/~bf3/tmp/colorhack.pl
21:35:14 <Bike> "The eval function works with text variables to implement a powerful text macro facility." hey Gregor did your evil eval paper mention matlab because i've got some choice quotes here
21:38:08 <Taneb> IT IS TIME TO TRY TO BAKE A CAKE
21:39:05 <Bike> A class of functions, called "function functions," works with nonlinear functions of a scalar variable. That is, one function works on another function.
21:56:25 * ski . o O ( `@(x,v) x .* v + w' )
21:56:40 -!- oerjan has joined.
21:58:50 <shachaf> ski: thanks for the reference to that paper yesterday, looks relevant to something or other that i'm probably trying to figure out
21:59:01 <shachaf> will know more after reading it
22:05:10 -!- atrapado has quit (Quit: Leaving).
22:07:11 <shachaf> "Weak Limited Principle of Omniscience" is a p. hesitant name
22:07:34 <ski> i don't recall whether i asked you whether you had seen Steven Vicker's "Topology Via Logic"
22:07:40 <ski> shachaf : hehe
22:08:32 -!- nchambers has joined.
22:08:37 <ski> also "Lesser Limited Principle of Omniscience"
22:08:44 <ski> (not the same thing)
22:08:46 <shachaf> ski: I have seen it -- we talked about "topological systems" as related to Chu spaces at one point.
22:09:10 <shachaf> (I don't remember who recommended it to me before, but thanks for the recommendation if it was you.)
22:11:40 -!- DTSCode-Work has quit (Ping timeout: 245 seconds).
22:12:03 <ski> <http://en.wikipedia.org/wiki/Limited_principle_of_omniscience>,<http://en.wikipedia.org/wiki/Weak_counterexample#Brouwerian_counterexamples>,<http://ncatlab.org/nlab/show/principle+of+omniscience>,<http://plato.stanford.edu/entries/mathematics-constructive/>
22:12:21 <ski> also LEM/PEM is called PO (Principle of Omniscience) by Bishop
22:16:35 <ski> (and Markov's Principle, MP, is considered an Omniscience Principle (though not a Brouwerian one, but a Markovian one). so reducing something to it amounts to a Markovian (counter-)example, rather than a Brouwerian (counter-)example)
22:17:34 <ski> shachaf : did the Escardó paper mention WLPO, or which ?
22:17:55 <shachaf> http://www.cs.bham.ac.uk/~mhe/papers/universe-indiscrete.pdf did
22:18:23 <shachaf> Which is related to a different question I was asking at one point.
22:19:13 <ski> hm, interesting, haven't read this one before
22:22:17 <ski> <http://www.infoq.com/presentations/nimrod>,<http://www.reddit.com/r/programming/comments/1vdo3l/nimrod_a_new_approach_to_metaprogramming/>
22:27:55 -!- atrapado has joined.
22:29:47 <shachaf> whoa, Andrej Bauer responded to my post
22:30:48 <shachaf> "There is another, fancy aprroach to this, called synthetic topology. Instead of doing all this computable stuff we just do ordinary topology, but we do it in intuitionistic logic. Then we can interpret such intuitionistic topology in any topos. When we pick the effective topos, we get back exactly computable topology."
22:31:53 -!- nchambers has quit (Ping timeout: 252 seconds).
22:32:58 <Taneb> I was going to make a cake
22:33:17 <oerjan> so is there a Greater Principle of Ignorance
22:39:07 <oerjan> apparently there's a Hawking's Principle of Ignorance.
22:39:23 <HackEgo> words --esolangs 5000 | grep -o "\w\+fuck\w*" || espletive
22:40:30 <mrhmouse> is 5000 too high? it (potentially) reduces the number of invocations, but I wasn't sure how much work `words is doing
22:40:57 <FireFly> never heard about that language
22:41:25 <FireFly> `run for f in $(seq 10); do espletive; done | xargs
22:41:28 <mrhmouse> Hrm. Seems to be stuck there. Earlier I got oozlybubblefuck, which doesn't appear on the wiki.
22:41:41 <HackEgo> blangintfuck brainfuck madbrainfuck befucks minifuck hydrainfuck alpainfuck infuck drainfuck rainfuck norfuck brainfuck
22:42:21 <Bike> a good program.
22:42:34 -!- augur has quit (Read error: Connection reset by peer).
22:42:41 <Bike> maybe it should have infix fucks (infux) too
22:42:54 -!- augur has joined.
22:42:59 <mrhmouse> Bike: it should. "\w\+fuck\w*"
22:43:14 <FireFly> it matched befucks at least
22:45:41 <Bike> oh missed that.
22:47:46 <oerjan> `run sed -i 's/espletive/exec espletive/' bin/espletive
22:47:59 <HackEgo> brainfuck \ norfuck \ brainfucks
22:49:07 <oerjan> you cannot exec while you're still running other commands in the pipeline :P
22:49:25 <oerjan> you _shouldn't_ be doing that, should you
22:50:43 <int-e> `` while ! words --esolangs 5000 | grep -o "\w\+fuck\w*"; do :; done
22:51:15 -!- augur has quit (Remote host closed the connection).
22:51:36 -!- Bike has quit (Ping timeout: 252 seconds).
22:53:07 <int-e> `` sed -i 's/fuck/${1-fuck}' bin/espletive
22:53:09 <HackEgo> sed: -e expression #1, char 16: unterminated `s' command
22:53:24 <int-e> `` sed -i 's/fuck/${1-fuck}/' bin/espletive; cat bin/espletive
22:53:28 <HackEgo> words --esolangs 5000 | grep -o "\w\+${1-fuck}\w*" || espletive
22:54:07 <HackEgo> crainfuck thatefuck celiumbrainfuck minifuck brainfuck ballfuck memfuck
22:54:31 <HackEgo> cat: repeat: No such file or directory
22:54:45 <HackEgo> args=$* \ for f in $(seq $1); do ${args[@]:1} ; done | xargs
22:54:53 <int-e> `` sed -i 's/$/"$@"/' bin/espletive; cat bin/espletive
22:54:57 <HackEgo> words --esolangs 5000 | grep -o "\w\+${1-fuck}\w*" || espletive"$@"
22:55:08 <int-e> `` sed -i 's/"/ "/' bin/espletive; cat bin/espletive
22:55:12 <HackEgo> words --esolangs 5000 | grep -o "\w\+${1-fuck}\w*" || espletive"$@"
22:55:35 <int-e> `` sed -i 's/"\$/ "$/' bin/espletive; cat bin/espletive
22:55:39 <HackEgo> words --esolangs 5000 | grep -o "\w\+${1-fuck}\w*" || espletive "$@"
22:56:03 <int-e> no funk? or did I mess this up completely now?
22:57:07 <nooodl> i could imagine no \w+funk
22:57:35 <mrhmouse> someone with better sed skills than me should change that + to a *
22:58:09 <oerjan> `run words --esolangs 5000 | wc
22:58:14 <int-e> `` sed -i 's/\+/*/' bin/espletive; cat bin/espletive
22:58:14 <oerjan> `run words --esolangs 5000 | wc
22:58:21 <oerjan> `run words --esolangs 5000 | wc
22:58:21 <HackEgo> words --esolangs 5000 | grep -o "\w\*${1-fuck}\w*" || espletive "$@"
22:58:37 <int-e> `` sed -i 's/\\*/*/' bin/espletive; cat bin/espletive
22:58:41 <HackEgo> *words --esolangs 5000 | grep -o "\w\*${1-fuck}\w*" || espletive "$@"
22:58:51 <mrhmouse> int-e: methinks \* is not the same as *
22:59:00 <oerjan> ok if it only has one line, wtf does it sometimes print more than one hit
22:59:02 <int-e> mrhmouse: it's not.
22:59:07 <int-e> `` sed -i 's/\\\*/*/' bin/espletive; cat bin/espletive
22:59:11 <HackEgo> words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" || espletive "$@"
22:59:29 <mrhmouse> oerjan: you mean when we were running it in loops?
22:59:47 <HackEgo> infuck wordfuck migolfuck brainfuck memfuck golfuck norfuck
22:59:54 <oerjan> mrhmouse: no, it did that sometimes even when run once
22:59:56 <int-e> can words --esolangs 5000 fail?
23:00:14 <mrhmouse> int-e: it can, hence the recursive tail
23:00:41 <int-e> `` for i in $(seq 10); do words --esolangs 5000 > /dev/null || echo fail!; done
23:00:43 <mrhmouse> well, `grep can fail... I don't actually know if `words can fail.
23:01:16 <int-e> I'm still confused how espletive can produce several words in a single run.
23:02:15 <mrhmouse> Didn't know about that `` shortcut for `run...
23:04:27 <HackEgo> words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" || espletive "$@"
23:04:34 <mrhmouse> It's because `words sometimes produces multiple lines
23:04:46 <mrhmouse> and sometimes more than one line matches the regex
23:04:53 <oerjan> `run words --esolangs 5000 | wc
23:04:53 <int-e> `` sed -i 's/-o/-o -m 1/' bin/expletive; cat bin/espletive
23:04:55 <HackEgo> sed: can't read bin/expletive: No such file or directory \ words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" || espletive "$@"
23:05:05 <oerjan> `run words --esolangs 5000 | wc
23:05:07 <int-e> `` sed -i 's/-o/-o -m 1/' bin/espletive; cat bin/espletive
23:05:11 <HackEgo> words --esolangs 5000 | grep -o -m 1 "\w*${1-fuck}\w*" || espletive "$@"
23:05:19 <oerjan> mrhmouse: i haven't managed to make it do that...
23:05:39 <FireFly> `run words --esolangs 5000 | wc
23:05:55 <FireFly> What happened with my output
23:06:15 <int-e> `` echo a b c | grep -o '[^ ]'
23:06:23 <FireFly> `run words --esolangs 50 | wc
23:06:25 <int-e> `` echo a b c | grep -o -m 1 '[^ ]'
23:06:35 <int-e> so -m doesn't help. grmbl.
23:07:20 <int-e> `` sed -i 's/||/| head -1 ||/' bin/espletive; cat bin/espletive
23:07:23 <oerjan> oh -o takes more than one hit?
23:07:24 <HackEgo> words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" | head -1 || espletive "$@"
23:07:49 <mrhmouse> from man grep: Print only the matched (non-empty) parts of a matching line, with each such part on a separate output line.
23:09:37 <mrhmouse> this proved more amusing than I thought it would...
23:10:31 <oerjan> (head might always succeed)
23:11:45 <mrhmouse> maybe after the recursive call?
23:12:22 <oerjan> the number of nested processes isn't going to reduce by this...
23:12:28 <int-e> `` sed -i 's/$/ | head -1/' bin/espletive; cat bin/espletive
23:12:32 <HackEgo> words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" || espletive "$@" | head -1
23:13:08 -!- atrapado has quit (Ping timeout: 252 seconds).
23:13:45 <int-e> We *could* turn it into a loop.
23:13:52 -!- mrhmouse has quit (Quit: Leaving.).
23:15:08 <int-e> `` sed -i 's/\(.*\)||.*|/while ! \1; do :; done |/' bin/espletive; cat bin/espletive
23:15:12 <HackEgo> while ! words --esolangs 5000 | grep -o "\w*${1-fuck}\w*" ; do :; done | head -1
23:16:16 <shachaf> Now that Andrej Bauer commented on my post I don't want to delete it anymore. :-(
23:17:21 <int-e> oerjan: are you happier now?
23:18:32 <oerjan> sorry, all out of infernal cash
23:19:55 <FireFly> Now I want to make forthagonal
23:20:13 <oerjan> that does sound promising
23:21:24 <FireFly> `run for i in $(seq 5); do espletive gon; done | xargs
23:21:40 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: : not found
23:21:51 <HackEgo> shogonatorycombie thagonal pogonata pogonata thagonal
23:22:10 <HackEgo> redcodan iota zt xsm q-ball /// c-lon rever2pi bytejump nump redgree kayak tree aargfak p'' tri divzeros bf-pda tg that trits objector mine adjust tlwnn
23:22:54 <int-e> Oh well. Validation of arguments is left as an exercise to the first one to complain about this feature.
23:25:16 <oerjan> `` espletive '[0-9]\{2\}' # surely this is a feature
23:25:42 <int-e> Why am I not surprised?
23:26:35 <oerjan> int-e: istr searching for "sex" on the wiki and the only mention was ironically on Taneb's user page. although that was whole-word only.
23:29:37 <oerjan> well, not as a substring either on either of language or joke language lists
23:30:07 -!- augur has joined.
23:30:48 -!- Phantom_Hoover has joined.
23:32:57 -!- peapodamus has quit (Read error: Connection reset by peer).
23:34:49 -!- peapodamus has joined.
23:57:57 -!- augur has quit (Ping timeout: 248 seconds).
23:59:50 -!- Sprocklem has quit (Ping timeout: 252 seconds).