00:00:07 <Taneb> Or planning to do or whatever
00:06:38 <zzo38> Maybe to write more of my Z-machine compiler, is what I might do today
00:06:48 <zzo38> Or play Pokemon card, or both
00:07:37 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
00:13:20 <pikhq> Bike: Define "better".
00:13:27 -!- not^v has joined.
00:13:33 -!- ^v has quit (Read error: Connection reset by peer).
00:15:12 <Bike> pikhq: is margarine really better than I Can't Believe It's Not Butter(TM)?
00:16:59 <newsham> bike: https://www.youtube.com/watch?v=0MsbvGmLaU4
00:23:21 <olsner> I don't believe I can't believe it's not butter is not butter after watching that clip
00:24:13 -!- Phantom_Hoover has joined.
00:24:21 <olsner> (if you believe you know what I'm talking about please explain, because I don't)
00:37:48 * oerjan finally got at least to a 1024
00:38:17 <Bike> fuck you, man :(
00:38:26 <int-e> is everybody playing this game?
00:38:34 <oerjan> you've been having trouble getting to that too?
00:38:44 <Bike> at least i got a 2048 in flappy logarithmic 2048
00:38:50 <Bike> "woo, you pased eleven pillars"
00:39:21 <olsner> I think my record is either 512 or 1024 (in the original 2048, not in the derivatives)
00:40:03 <oerjan> Bike: i think i fell into a nice rhythm this time
00:40:08 <int-e> 512 was easy, 1024 was hard, 2048 was impossible until I adapted a good strategy (which I can't claim credit for, I watched a colleague play.).
00:40:12 -!- not^v has quit (Read error: Connection reset by peer).
00:40:21 <oerjan> and yeah the original.
00:40:40 <int-e> Somebody wrote a computer program that gets 2048 90% of the time.
00:41:05 <olsner> I found the doge-2048 easier than the original though, maybe the trick is to watch the colors and not the numbers
00:41:24 <int-e> http://gabrielecirulli.github.io/2048/ http://www.2048tile.co/ (which I believe are the same game, but the latter is full of ads)
00:42:37 <int-e> And I hope that's the original one.
00:44:39 -!- Sprocklem has quit (Ping timeout: 245 seconds).
00:44:44 <oerjan> if i add the pieces other than the 1024 at game over, i get 526 i think
00:48:20 <newsham> pet peeve: store websites that try to be amazon, but wont tell you whats actually in stock in an actual store down the street from you.
00:48:33 <newsham> no walmart, i dont give a fuck about what you can fedex me.
00:53:41 <Bike> i just got my highest score yet (only up to 512 though) but it took systematism
00:53:45 <Phantom_Hoover> <int-e> 512 was easy, 1024 was hard, 2048 was impossible until I adapted a good strategy (which I can't claim credit for, I watched a colleague play.).
00:54:14 <olsner> left up right down, p. good strategy
00:58:10 <kmc> <Bike> is rankine really better than fahrenheit
00:59:05 <Bike> i don't actually remember which one rankine is. fahrenheit but based at absoltue zero?
01:00:52 <Bike> wikipedia tells me that the gas constant, as measured in cubic feet psi per rankine-pound-mole, is almost eleven
01:01:17 <Bike> i'm like 80% sure somebody wrote this table as a joke
01:03:56 -!- ket1v has joined.
01:04:43 <oerjan> <olsner> left up right down, p. good strategy <-- that actually works quite a while, i think. maybe as far as 256.
01:05:10 <oerjan> although by the time it stops working, your field will be so messed up it's unsaveable without extreme luck.
01:07:38 <oerjan> hm not quite to 256 on first try
01:07:57 <oerjan> but i have 128, 64, 32 and 3*16
01:08:20 <oerjan> so it will probably manage occasionally
01:08:29 -!- ket1v has quit (Ping timeout: 240 seconds).
01:10:34 <oerjan> 2nd try had 128 + 64 + 2*32
01:12:13 <shachaf> oerjan: any strategy will work for quite a while
01:12:26 <shachaf> imo try to lose with as low a score as possible
01:13:47 <oerjan> 3rd try got above 256, then as well as 128 + 2*64
01:17:03 -!- Phantom_Hoover has quit (Quit: Leaving).
01:17:53 <oerjan> shachaf: ok i managed before 128 on first try
01:19:21 <shachaf> how is the score computer anyway
01:19:28 <shachaf> why isn't it just the sum of the numbers on the board or something
01:20:05 <monotone> I thought it was the sum of the tiles you joined together. Admittedly I didn't actually check my theory.
01:20:37 <int-e> http://int-e.eu/~bf3/tmp/2048-2.png :)
01:27:27 <oerjan> shachaf: ok with a slightly better misere strategy i managed score 276
01:28:12 -!- tromp has quit (Remote host closed the connection).
01:28:45 -!- tromp has joined.
01:29:25 -!- ket1v has joined.
01:33:01 -!- tromp has quit (Ping timeout: 252 seconds).
01:35:01 <oerjan> http://i.imgur.com/EQzOd1v.png
01:40:53 -!- ket1v has quit (Remote host closed the connection).
01:46:52 -!- ket1v has joined.
01:58:30 -!- ket1v has quit (Remote host closed the connection).
02:00:41 <zzo38> Do they keep track of misere scores?
02:11:46 <oerjan> <quintopia> the lack of gregor and gregorbots is subtly motivating me to write a bfjoustbot <-- EgoBot is still here
03:03:03 <quintopia> oerjan: the lack of gregor is the bigger deal. also, his unwillingness to install the fixed point scoring
03:10:35 -!- conehead has quit (Quit: Computer has gone to sleep.).
03:21:50 <shachaf> zzo38: Do you like the chess variations in <http://sigbovik.org/2014/proceedings.pdf>?
03:22:43 -!- ket1v has joined.
03:23:11 <zzo38> This file is damaged.
03:24:28 <zzo38> I don't know. My computer tells me it is damaged.
03:24:31 -!- shikhin has quit (Ping timeout: 264 seconds).
03:24:51 <zzo38> I don't quite believe them, but that is what it says.
03:25:03 <shachaf> Which PDF viewer are you using?
03:25:10 <shachaf> I am using Google Chrome. It works for me.
03:26:52 <Sgeo> Fixed point scoring for waht?
03:36:27 -!- tromp has joined.
03:37:38 -!- nooodl_ has quit (Ping timeout: 246 seconds).
03:43:11 <newsham> another country heard from
03:53:36 -!- ^v has joined.
04:12:49 <kmc> i wanna take acid and play with d3.js demos for hours
04:13:48 -!- ket1v has quit (Remote host closed the connection).
04:14:24 -!- ket1v has joined.
04:18:21 <kmc> http://rudradevbasak.github.io/16384_hex/
04:19:07 -!- nisstyre has quit (Ping timeout: 265 seconds).
04:19:22 -!- nisstyre has joined.
04:19:24 -!- nisstyre has quit (Changing host).
04:19:24 -!- nisstyre has joined.
04:21:43 <kmc> shachaf: woah, SIGBOVIK preprint
04:33:14 -!- conehead has joined.
04:42:19 -!- idris-ircslave has quit (Quit: Terminated).
04:55:26 -!- Sellyme_ has joined.
04:57:55 -!- contrapumpkin has joined.
04:58:47 -!- copumpkin has quit (Read error: Connection reset by peer).
05:00:11 -!- Sellyme has quit (*.net *.split).
05:00:13 -!- Sellyme_ has changed nick to Sellyme.
05:00:48 -!- nisstyre has quit (Quit: WeeChat 0.4.3).
05:03:14 -!- Sellyme has quit (Excess Flood).
05:05:15 <Bike> i was thinking of writing a sigbovik paper, too
05:05:43 <Sgeo> (Disclosure: I am an employee of Cablevision. My opinions are my own, and do not necessarily reflect the opinions of Cablevision)
05:05:54 <Sgeo> I feel like I should be proud of my ISP http://blog.netflix.com/2014/03/internet-tolls-and-case-for-strong-net.html
05:06:09 <Sgeo> "Some major ISPs, like Cablevision, already practice strong net neutrality and for their broadband subscribers, the quality of Netflix and other streaming services is outstanding."
05:07:46 <elliott_> did you actually just write that disclosure on irc
05:08:03 -!- nisstyre has joined.
05:08:32 <Bike> what a coincidence, my opinions don't necessarily reflect cablevision's either
05:08:34 <Sgeo> elliott_: I think I'm supposed to? Well, at least on other social networks
05:08:43 <zzo38> Perhaps company policy requires you to make such a disclosure?
05:08:52 <zzo38> I don't know how the company policy works.
05:11:01 <Sgeo> zzo38: it does, pretty much. Not sure if I'm supposed to talk about that though. I _think_ I can
05:11:59 <Bike> is that thought owned by Cablevision
05:12:08 -!- tromp has quit (Remote host closed the connection).
05:12:43 -!- tromp has joined.
05:14:01 <shachaf> disclosure: my opinions are mostly just cobbled together from the opinions of thousands of others. i've hardly had an original thought in my life.
05:15:21 <Sgeo> "If it was never seen, never known, and never thought of, then you will not find it in the Book no matter how far you look, though, and I mean no offence, the chances of you ever having an original thought such as that are remote."
05:16:44 -!- tromp has quit (Ping timeout: 245 seconds).
05:16:51 <Bike> the turner diaries
05:17:08 <Sgeo> zzo38: http://qntm.org/library
05:19:55 -!- Sellyme has joined.
05:32:41 -!- password2 has joined.
05:36:57 -!- passwordBOT has joined.
05:39:10 <Sgeo> [insert joke about factorials here]
05:39:27 <Sgeo> [insert sheep here]
05:40:04 <password2> mmm, it was probably not a wise choice making my bf memory gor from -127 to +127 if my cell values gor 0-255
05:40:50 <Sgeo> http://www.froup.com/tr/tr.pl?92
05:40:59 -!- passwordBOT has quit (Ping timeout: 240 seconds).
05:45:13 <Sgeo> http://www.froup.com/tr/tr.pl?110
05:50:16 -!- ^v has quit (Quit: Leaving).
06:03:16 <int-e> what a waste of time
06:05:43 -!- oerjan has quit (Quit: Nite).
06:08:32 -!- Sorella has quit (Quit: It is tiem!).
06:14:34 <int-e> http://int-e.eu/~bf3/tmp/2048-3.png became http://int-e.eu/~bf3/tmp/2048-4.png
06:17:16 -!- conehead has quit (Quit: Computer has gone to sleep.).
06:21:00 <kmc> 2048 is the mind-killer
06:22:34 <pikhq> Politics is the 2048
06:22:56 <Jafet> kmc will not 2048.
06:23:34 <pikhq> kmc prefers playing aleph-null
06:23:55 <int-e> that requires in infinite grid
06:24:23 <int-e> I *will* sleep now. Good night.
06:26:41 <kmc> so, how 'bout them Turing machines, eh? never know when they're gonna stop.
06:27:05 -!- passwordBOT has joined.
06:27:21 <pikhq> I'm pretty sure the cardinality of the set of brainfuck variants actually is aleph-null.
06:27:29 <kmc> is it a combination brainfuck variant / 2048 game / web framework?
06:28:00 <password2> it adds relative increase cell and decrease
06:28:40 <password2> +bf:++++++++++>++++++++++++<[->}-<]
06:28:44 <pikhq> So, what, > that goes right a number of times based on the current cell?
06:29:01 <pikhq> That does make array-like things much, much easier.
06:29:10 <password2> no } increases the cell the pointer is pointing to
06:29:21 <passwordBOT> 0 2 0 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
06:29:23 <Jafet> Don't invest in brainfuck derivatives.
06:29:38 <kmc> buy buy buy sell sell sell
06:29:52 <pikhq> What is d brainfuck/dx, anyways?
06:30:01 <kmc> d brain / d fuck
06:30:25 <kmc> let's get all fancy ∂brain/∂fuck
06:30:36 <pikhq> Oh, nice, a partial one.
06:31:12 <Jafet> The jacofuckyoubian
06:32:01 <password2> a bit , didn't expect someone would actually whip out the correct symbol so fast
06:32:19 <pikhq> kmc: What's the compose for that?
06:33:36 <kmc> Most ever Brainfuckiest Fuck you Brain fucker Fuck
06:33:41 <kmc> why the muttering
06:33:58 <password2> now to think of neat possibilities of my bfpointer thingy
06:34:52 -!- passwordBOT has quit (Quit: Good).
06:41:01 -!- conehead has joined.
06:57:09 -!- ket1v has quit (Remote host closed the connection).
07:03:23 -!- MindlessDrone has joined.
07:06:32 -!- MoALTz has joined.
07:13:48 <password2> err , i see now my modification to bf is very limited
07:15:14 <password2> maybe if i make the looping relative too
07:23:01 <password2> the syntax for moving a value got a bit simpler
07:27:38 -!- Bike has quit (Quit: nine hour).
08:04:53 -!- douglass_ has quit (Ping timeout: 264 seconds).
08:05:04 -!- douglass_ has joined.
08:29:11 <Sgeo> http://www.froup.com/tr/tr.pl?538
08:41:20 <Jafet> The following sentence is false.
08:41:25 <Jafet> The preceding sentence is true.
08:57:54 -!- ket1v has joined.
08:59:12 -!- AnotherTest has joined.
09:02:09 -!- ket1v has quit (Ping timeout: 245 seconds).
09:03:26 -!- conehead has quit (Read error: Connection reset by peer).
09:20:20 <fizzie> Are there solvers (or a name) for a generalized TSP where the distances between nodes can depend on the index of that edge on the path? (And with free start/end points and no need to return to the origin.)
09:32:15 <fizzie> (I guess that last condition is easy to reduce to the standard one by adding one vertex.)
09:37:25 <Jafet> That sounds vaguely PSPACE-hard.
09:41:20 <lexande> isn't it still only NP-hard? at least the decision problem of whether you have a path of cost < n still only requires a polynomial-time-checkable witness
09:53:01 <Jafet> Suppose d(e, i) ≈ e/2^i, then paths can be arbitrarily long
10:10:07 <oklopol> however, how are those numbers given, it's probably very dangerous to give them via polytime turing machines
10:11:19 <fizzie> I'm not even sure what e is in "d(e, i) ≈ e/2^i".
10:15:00 <oklopol> it's the edge on the left and its weight on the right, maybe
10:16:20 <fizzie> (Also not sure about arbitrarily long, since i \in {1, ..., N}.)
10:17:15 <oklopol> couldn't it be the fastest solution to wait for exponentially many steps at the first node (because up to that point, everything is very expensive)
10:17:25 <oklopol> or does tsp have to be a hamiltonean path
10:17:38 <fizzie> I think it usually does.
10:17:51 <fizzie> (Traditionally, a Hamiltonian cycle, even.)
10:17:59 -!- Melvar has quit (Ping timeout: 240 seconds).
10:18:20 <oklopol> i guess in the euclidean case it's fine
10:18:31 <oklopol> because if you repeat a node, you could bypass it by triangle inequality
10:22:13 <fizzie> Wiki says that if you have a non-metric case without the triangle inequality, and allow for revisiting cities, you can replace that by doing regular TSP in the complete graph where d(A, B) is the shortest path between A and B in the original graph.
10:22:55 <fizzie> (Presumably not so applicable for the case with time in it.)
10:23:02 <oklopol> so okay, for tsp it's very reasonable
10:23:07 <oklopol> but it's not reasonable for this version
10:23:41 <fizzie> My underlying practical problem needs a Hamiltonian path, anyway.
10:23:46 <oklopol> unless you already had an application in mind, i guess even the ugliest problem could be called reasonable if it's actually needed
10:25:17 <Deewiant> @tell ais523 Jettyplay will happily encode something more than 4 (or is 2 the cutoff?) GiB in size but only writes a 0-byte file. I thought it'd be an easy fix but evidently the 4 GiB limit is due to the AVI format...
10:26:36 <fizzie> Internet says the problem is called "Time Dependent Traveling Salesman Problem", which isn't terribly surprising.
10:27:26 <Jafet> That's a really long ttyrec
10:27:27 <oklopol> i don't understand why hamiltonean
10:27:37 <oklopol> is your application a secret?
10:28:04 <fizzie> No, but it's too silly to speak of.
10:28:39 <oklopol> can you think of a non-silly application?
10:29:18 <fizzie> This TDTSP paper talks of "a number of important applications", but does not go into all that much detail.
10:29:30 <fizzie> Except mentioning that the Traveling Deliveryman Problem is a special case.
10:30:18 <oklopol> oh important applications, those are pretty important
10:30:42 <fizzie> Also applicability of their algorithm to Vehicle Routing Problem variants.
10:30:42 -!- passwordBOT has joined.
10:31:09 <fizzie> http://www.optimization-online.org/DB_FILE/2010/12/2872.pdf "THe Time Dependent Traveling Salesman Problem: Polyhedra and Algorithm", Abeledo et al., if you want to go chasing references or something.
10:33:20 <oklopol> may i'll just believe them
10:35:40 <fizzie> My application is that I have a set of N aligned photos that I want to blend so that I take a narrow vertical column from each of them, and order the photos so that it minimizes the sum of the distances between adjacent photos, as considered in the region around the column, which I think is an instance of that.
10:38:00 <fizzie> (Probably not one of the "important" ones.)
10:40:32 <Jafet> How is that time-dependent? The distance between adjacent photos is fixed.
10:40:43 <int-e> Jafet: "around the column"
10:40:57 <int-e> and column number = time.
10:41:49 <oklopol> why is that not an instance of the usual version? what is time here?
10:41:58 <Jafet> If the distance depends on more photos in either direction, then it is path-dependent, not time-dependent
10:42:13 <int-e> iow, depending on the column number, you compare different areas of the photos.
10:44:14 <fizzie> Yes, that was exactly what I thought. (Though you could certainly generalize it further by involving more than just the adjacent photos.)
10:44:26 <oklopol> fizzie: but why do you want a hamiltonean path?
10:44:58 <oklopol> do you have as many pictures as columns?
10:45:29 <fizzie> I was going to just set the column width based on that.
10:45:37 <Jafet> I suppose fizzie wants to have exactly as many columns as pictures.
10:46:13 <int-e> oklopol: I imagine (wild guess, probably wrong) it's for a website or similar displaying images, an index of the photos that can be rotated by dragging or the like.
10:46:39 <fizzie> It's not for anything as useful as that, I just wanted to see what it looks like.
10:46:53 <fizzie> (The photos are all of the same thing taken at different times.)
10:47:07 <Jafet> For a small number of images, you could just use brute force.
10:47:40 <int-e> another kind of time lapse
10:47:40 <Jafet> Those are normally presented in chronological order, though.
10:48:16 <fizzie> I'll probably have about a hundred of them, and 100! is a big number. (But I could use some generic optimization things, sure.)
10:48:23 <int-e> but maybe the pictures are taken at random times of the day :)
10:48:53 <int-e> And the timestamps are lost.
10:49:01 <Jafet> You could reduce this to weighted set cover by weighting each set {left_i, right_j} with the transition cost of left_i to right_j. The solution has to cover all {left_i} and all {right_j}.
10:49:52 <fizzie> There is a chronological order, but chronologically adjacent photos aren't especially similar here.
10:49:54 <Jafet> (Weighted set cover is NP-complete by reduction to integer programming.)
10:50:11 <int-e> Jafet: but ypu'll likely get a bunch of cycles that way.
10:50:48 <Jafet> That's true. I guess you can add ILP constraints to restrict those.
10:52:22 <Jafet> Apparently the correct name is exact set cover.
10:53:20 <Jafet> Why do we have so many names, anyway. They all have trivial reductions to ILP.
10:53:24 <int-e> (I would have made the same mistake, hence the smile.)
10:53:36 -!- nooodl_ has joined.
10:58:42 -!- ket1v has joined.
10:59:41 <int-e> Jafet: I'm wondering about a (practically irrelevant) detail here. Assume we have an LP with integer variables but real coefficients. How big can an encoding of that using only integer coefficients be?
11:00:42 <int-e> (in practice we are usually happy with rational coefficients and the question becomes easy to answer)
11:03:32 -!- ket1v has quit (Ping timeout: 252 seconds).
11:04:03 <Jafet> As we all know, real numbers are not real
11:05:02 <int-e> Ah, I can answer this; even with a single equation (to inequalities) the encoding can require arbitrarily large numbers, because with real coefficients, a single equation can encode a single integer point on a plane (or in Z^n if we wish).
11:05:23 <Jafet> Since the equations are linear, though, they are going to define a polytope. So you probably only need to consider something like the field extension Q[√x] for whatever real x you're using.
11:05:42 -!- elliott_ has quit (Quit: Reconnecting).
11:05:50 -!- elliott has joined.
11:06:21 <Jafet> So if you have an oracle for arithmetic in Q[√x1][√x2]...
11:06:56 <int-e> Note that I completely skirted the representation and computability issues.
11:07:27 <int-e> it was a stupid question because ILP can't express x >= sqrt(2)*y
11:07:47 <Jafet> What is the cost function?
11:08:02 <Jafet> (The best cost function for that would obviously be x - sqrt(2)*y)
11:10:48 <int-e> Ah. In my background we're just concerned with feasibility, so there is no cost function. It could be x - sqrt(2)*y and there could be additional constaints a <= x <= b to make it interesting (otherwise the answer will be 0).
11:11:10 <int-e> that is, (x,y) = (0,0) and cost = 0.
11:12:57 -!- Melvar has joined.
11:16:13 <password2> did I just accidentally edit the noticE?
11:17:11 <int-e> If you mean the topic, then no, you did not.
11:17:19 <int-e> Otherwise I'm confused.
11:17:47 <password2> because i wanted to copy a url and i cut it , now my topic shows it without that url
11:18:56 <int-e> Just don't press enter in the topic field :)
11:19:20 -!- password2 has left ("Leaving").
11:19:25 -!- password2 has joined.
11:19:59 <int-e> In any case some kind soul would happily restore it for you if you mess up, I'm sure. So don't worry :)
11:25:38 <password2> I just don't like to create unnecessary effort for people
11:27:25 <password2> mmm , i see someone actually built a hoverboard
12:05:13 -!- passwordBOT has quit (Quit: Good).
12:11:04 <Vorpal> elliott, do you know of any other distro with a package system similar to NixOS, specifically the different versions in different places bit?
12:11:26 <Vorpal> Prebuilt packages preferably
12:11:31 <elliott> gobolinux. you probably don't want to use it. use nixos instead
12:11:49 <Vorpal> Doesn't nixos use source packages?
12:12:18 <elliott> it has cached binary packages
12:12:24 <elliott> obviously if you customise packages they will hvae to be built
12:12:30 <elliott> because no binary with a matching hash will exist
12:12:50 <elliott> well you won't get any binaries on arm I don't think.
12:12:52 <Vorpal> I forgot which one is which
12:13:07 <elliott> https://nixos.org/wiki/Raspberry_Pi
12:13:36 <Vorpal> elliott, yeah lack of prebuilt will be a killer for RPi sadly
12:13:47 <elliott> you can cross-compile the binaries from any other linux host
12:13:51 <elliott> that should be very easy with nix
12:14:08 <Vorpal> Hm, I had issues cross-compiling in general for non-trivial stuff
12:14:25 <Vorpal> The best thing I managed was distcc to a cross toolchain, which makes configure and such slow still
12:16:46 <Vorpal> have you used nixos yourself as a primary OS on a computer (rather than for just playing around with)?
12:17:15 <elliott> no but I did a lot of research into doing so a few times
12:17:38 <elliott> you don't want my real-world distro usage, I have also used windows as a primary OS within the last year
12:18:01 <Vorpal> So, what did you find? Last time I checked I found the existing package repo... somewhat lacking. Sure I could package a lot of software myself, but that takes time, and so does ensuring that it is up-to-date.
12:18:46 <Vorpal> I don't expect it to compare to debian or ubuntu, but hopefully most of what I want should just be there.
12:19:18 <elliott> the package base is okay for a niche distro, it's not that hard to package stuff (and you can just submit it upstream), shrug
12:22:08 <Vorpal> Yeah NixOS on RPi seems annoying
12:24:55 <Vorpal> elliott, so, out of interest, what is wrong with gobolinux?
12:25:20 <elliott> why use something even more niche than nixos that has almost none of the benefits
12:25:45 <elliott> and do you really want $HOME to start with /Users
12:26:09 <elliott> /System/Links/Headers ~~so beautiful~~~
12:26:16 <Vorpal> Looks like OS X at that point
12:29:16 -!- passwordBOT has joined.
12:29:46 -!- boily has joined.
12:29:52 <password2> i wonder what eso lang i should try next
12:29:55 -!- metasepia has joined.
12:30:37 <Vorpal> password2, since you use xchat you could also have restored your client's copy of the topic by doing a plain "/topic" without any parameters
12:30:53 <Vorpal> password2, which ones have you tried?
12:31:07 -!- shikhin has joined.
12:31:15 <password2> i have tried brainfuck , and even added onthere instruction
12:32:13 <Vorpal> What about underload, or befunge? Both are quite different languages from bf at least
12:34:11 * boily shouts “AUBERGINE!” :D
12:34:48 <Vorpal> boily, need to check that one out myself
12:36:47 <password2> i think i'll play around witf bf a bit more first
12:38:40 <passwordBOT> 0 |0| 3 7 14 21 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
12:45:44 <boily> (meanwhile, oh fungot. I was paying my ISP bill on the banking site. clicked the confirm button, then 500 internal server error. oooooh fsck.)
12:45:44 <fungot> boily: agora alice c64 ct darwin discworld europarl ff7 fisher ic irc* jargon lovecraft nethack pa speeches ss wp
12:50:12 <Vorpal> boily, lets just the database is ACID
12:51:09 <Vorpal> It certainly should be
12:51:20 <Taneb> https://sphere.chronosempire.org.uk/~HEx/8402/ reverse 2048
12:51:23 <boily> I relogged in, then checked the transaction history. the attempt was absent, so I tried it again, and it went through.
12:57:13 -!- passwordBOT has quit (Ping timeout: 240 seconds).
12:58:43 -!- password2 has quit (Ping timeout: 265 seconds).
12:59:35 -!- ket1v has joined.
13:02:28 <fungot> ket1v: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica ...
13:02:50 -!- password2 has joined.
13:03:58 -!- ket1v has quit (Ping timeout: 240 seconds).
13:04:28 <int-e> I really like how mtgox found heaps of coins in the electronic equivalent of the back of a drawer :)
13:06:44 <Vorpal> int-e, oh? Recently? Last I heard of the story, they went bankrupt
13:07:43 <int-e> sometime this weak
13:07:52 <Vorpal> Yeah found a news article
13:08:10 <Vorpal> Also scary that they could have missed it in the first place
13:12:12 <int-e> I wonder whether they found any black lotus cards or similarly expensive cards, too.
13:13:07 <boily> ~duck blacker lotus
13:13:25 * boily notes that the cuttlefish should fetch info from gatherer too...
13:13:57 <int-e> (I never bothered to research how a mtg card exchange became - at some point in time - the biggest online market for bitcoins, but the idea is ridiculous to me.)
13:14:30 <boily> the idea is very interesting.
13:15:01 <boily> also, this land's price is currently floating at around a hundred bucks → http://magiccards.info/query?q=misty&v=card&s=cname
13:15:24 <int-e> boily: You know what MtGoX originally stood for, right?
13:15:36 <int-e> (not sure about capitalization)
13:15:49 <boily> the power nine or something like that?
13:16:13 <int-e> It was an MtG Online eXchange.
13:16:49 * boily is enlightened by the sudden revelation *angel and dæva chorus ♪*
13:28:33 -!- nooodl has joined.
13:30:43 <Slereah_> Any idea why this bootloading attempt does not load boots?
13:30:44 <elliott> int-e: it was really just a reused name, afaik.
13:31:02 <elliott> it never really was an actual MTG exchange, to my knowledge?
13:31:07 <elliott> at least not one beyond prototype stage
13:31:11 <elliott> it just got scrapped and reused for bitcoins later.
13:31:19 -!- nooodl_ has quit (Ping timeout: 245 seconds).
13:31:35 <Slereah_> Just trying to load in memory at 0x8000:0x0000 some code to make the screen white
13:36:35 <boily> elliott: don't go and break the effect of my background choir hth
13:54:06 <Slereah_> Apparently boot sector is sector 1, not 0
14:00:18 <Slereah_> Code still doesn't work though.
14:00:26 <Slereah_> Second sector starts at 0x200 right?
14:01:15 -!- jconn has quit (*.net *.split).
14:01:19 -!- ion has quit (*.net *.split).
14:01:19 -!- EgoBot has quit (*.net *.split).
14:01:19 -!- olsner has quit (*.net *.split).
14:01:19 -!- newsham has quit (*.net *.split).
14:01:19 -!- samebchase has quit (*.net *.split).
14:01:19 -!- shachaf has quit (*.net *.split).
14:01:19 -!- int-e has quit (*.net *.split).
14:01:19 -!- kmc has quit (*.net *.split).
14:01:19 -!- Gracenotes has quit (*.net *.split).
14:01:19 -!- ineiros has quit (*.net *.split).
14:01:19 -!- mroman has quit (*.net *.split).
14:01:26 -!- shachaf has joined.
14:01:33 -!- Gracenotes has joined.
14:01:35 -!- mroman has joined.
14:01:39 -!- samebchase has joined.
14:01:39 -!- newsham has joined.
14:01:41 -!- olsner has joined.
14:01:47 -!- EgoBot has joined.
14:02:01 -!- ion has joined.
14:05:36 -!- sebbu has quit (Ping timeout: 335 seconds).
14:05:53 -!- sebbu has joined.
14:06:22 -!- ineiros has joined.
14:06:27 -!- sebbu has quit (Changing host).
14:06:27 -!- sebbu has joined.
14:06:43 -!- int-e has joined.
14:07:14 -!- elliott_ has joined.
14:07:53 -!- kmc has joined.
14:10:41 -!- mysanthrop has joined.
14:15:50 -!- myname has quit (Write error: Broken pipe).
14:16:51 -!- elliott has quit (Remote host closed the connection).
14:16:56 -!- Phantom_Hoover has joined.
14:27:13 <int-e> Slereah_: how are you testing this?
14:27:36 <Slereah_> Floppy image put in virtual box
14:28:18 * int-e found with dosbox that the image has to have the right size. 180kiB works.
14:32:55 <Slereah_> http://pastebin.com/ajGQ42ii < all workin'
14:33:35 * int-e ended up with http://sprunge.us/iiQB
14:35:06 <int-e> urgh. s/%/#/g in there. stupid last minute changes
14:37:17 <Slereah_> I guess I could put my Pong code afterwards
14:38:46 <int-e> I did figure out the sector number mistake quite soon, but the floppy image was too small, and dosbox didn't complain about it. It loaded the first 512 bytes as the boot sector, but the disk reading call would fail. Took me a while to figure that out.
14:41:44 -!- tromp has joined.
14:42:28 <int-e> Another mystery is that the code stops working if I load the second sector to 0x7000:0 ... huh.
14:51:53 <int-e> but that appears to be dosbox' fault.
14:53:53 -!- augur has joined.
15:00:26 -!- ket1v has joined.
15:02:08 -!- Sorella has joined.
15:04:15 <int-e> ooh. damn, I should initialise the stack before using it. dosbox apparently used 0x7000:0x100 ...
15:05:21 -!- ket1v has quit (Ping timeout: 265 seconds).
15:12:14 <int-e> Hah. http://sourceforge.net/p/dosbox/code-0/3808/tree//dosbox/trunk/src/dos/dos_programs.cpp has a comment on it. It says: '/* set up stack at a safe place */'. That's what I thought, too... "load sector to a safe place".
15:15:14 -!- Bike has joined.
15:28:05 <int-e> Somewhat Secure SHell?
15:32:30 <password2> more like , stop talking bout things i dont know
15:34:34 <password2> this room continually surprises me bu the conctend being discussed
15:39:42 -!- Bike has quit (Quit: NINE HOUUUUUR).
16:06:42 -!- shikhout has joined.
16:09:31 -!- shikhin has quit (Ping timeout: 264 seconds).
16:09:32 -!- shikhout has changed nick to shikhin.
16:10:04 <fizzie> Oh, regarding yesterday's problem with MATLAB and an invertible matrix with det(m) == 0; someone had switched things from double to single precision. The problem went away after switching back to double.
16:10:26 <fizzie> There was even a comment saying something to the tune of "single precision gives same results but faster".
16:10:59 -!- tromp has quit (Remote host closed the connection).
16:11:34 -!- tromp has joined.
16:15:38 -!- tromp has quit (Ping timeout: 240 seconds).
16:19:00 -!- tromp has joined.
16:23:20 <oklopol> did you tell him he's an idiot
16:28:41 <fizzie> He had already left for the day, so I'll have to wait until Monday for that.
16:34:32 -!- elliott_ has changed nick to elliott.
16:57:30 -!- Bike has joined.
17:00:11 -!- yorick has joined.
17:01:26 -!- ket1v has joined.
17:05:36 <newsham> fizzie: dont you get big inaccuracies when det(m) is very close to zero?
17:06:11 -!- ket1v has quit (Ping timeout: 265 seconds).
17:10:57 -!- Bike has quit (Read error: Connection reset by peer).
17:11:11 <fizzie> It's not very close to zero if I use double-precision processing elsewhere in the code.
17:11:30 <fizzie> Well, depending on what "very close" means; it's 0.0185 for the test case where I noticed the problem.
17:11:42 <fizzie> (That's not so close.)
17:12:25 <fizzie> (It's not really my code, so I don't know all the details of it.)
17:30:05 <Vorpal> ouch, the sudoers man page is really terrible. Way too much work to read to figure out a simple "how do I let this user run exactly this command" for example. Wasn't there a recent xkcd about that iirc?
17:30:38 <Vorpal> Well I'm resorting to "I remember doing this once before, lets find out on which system and then copy & adapt"
17:31:14 <int-e> Vorpal: are you reading xkcd? ( https://xkcd.com/1343/ ... image title)
17:31:34 <Vorpal> int-e, ah yeah that one. Fairly recent iirc?
17:32:45 <Vorpal> int-e, I blame my hazy memory on having a really bad cold from monday evening until thusday. Probably norovirus (I believe it is called in English)
17:34:04 -!- Frooxius has quit (Quit: *bubbles away*).
17:34:20 <int-e> Vorpal: My thought was "this is the second complaint about the sudoers manpage that I've seen in the last couple of days" and then I followed that thread.
17:34:40 <Vorpal> int-e, right. But it is actually terrible
17:36:52 -!- Frooxius has joined.
17:37:53 <fizzie> It's terrible, but "skip directly to EXAMPLES" is a workable strategy when dealing with it. (And even there, one has to skip over all those alias definitions.)
17:38:57 <Vorpal> fizzie, as long as there is an example that works for your case
17:39:05 <Vorpal> otherwise, you are in a LOT of problems
17:40:10 <fizzie> Sure, but there certainly are some examples you can adapt for "let one user run one command". (Okay, the behaviour w.r.t. command line arguments is slightly hard to derive from the examples.)
17:40:33 <Vorpal> fizzie, that is the thing I need though :/
17:41:02 <fizzie> Well, it's not entirely impossible to guesstimate something that works from the "pete" example.
17:41:27 <fizzie> Or at least when combined with the "joe" one.
17:41:50 <fizzie> (Actually, just "joe" directly.)
17:42:05 <Vorpal> Also the fucking host part, I never figured that one out really
17:42:05 <fizzie> "joe ALL = /usr/bin/su operator -- The user joe may only su(1) to operator."
17:43:22 -!- password2 has quit (Quit: Leaving).
17:45:35 <fizzie> The BNF-based documentation could probably be improved a lot by just rearranging it more logically.
17:46:56 <fizzie> As it is, it needs quite a lot of spelunking to even figure out that one of those actual interesting lines -- like that "joe ALL = /usr/bin/su operator" -- is an instance of (the rather badly named) "User_Spec".
17:49:26 <fizzie> EBNF, and there's quite a lot of text too, it's just all meshed up.
17:49:43 <Phantom_Hoover> hahaha they define ENBF, then define the syntax by means of ebnf
17:49:59 <fizzie> Yes, that's I guess what the xkcd joked about.
17:50:43 <fizzie> Given all that formalism, it's quite funny that it doesn't have a start symbol that'd define what an entire "sudoers" file consists of.
17:52:41 -!- AnotherTest has quit (Ping timeout: 245 seconds).
17:53:03 <fizzie> (That part is specified in prose. "The sudoers file is composed of two types of entries: aliases (basically variables) and user specifications (which specify who may run what). There are four kinds of aliases: User_Alias, Runas_Alias, Host_Alias and Cmnd_Alias." And then you can start getting deeper based on those four terms, and guess that since User_Spec is defined as the first line of the ...
17:53:08 <fizzie> ... "User specification" section, that's the syntax for a single entry.)
17:55:15 -!- MoALTz has quit (Read error: Connection reset by peer).
17:55:59 -!- MoALTz has joined.
17:56:45 <fizzie> Alias ::= 'User_Alias' User_Alias (':' User_Alias)* | ... reusing the name of the nonterminal as part of the concrete syntax might also have not been the best choice, especially when not all occurrences of the concrete string are quoted.
18:24:40 <Vorpal> elliott, well I played around a bit with the nixos virtual appliance thing they have on their web-site. Still not quite there for me. Spent 45 minutes on trying to make my fontconfig settings take effect (KDE settings, ~/.fonts.conf, ~/.Xresources, asking NixOS IRC channel). About to give up
18:26:47 -!- tromp has quit (Remote host closed the connection).
18:26:50 <Vorpal> elliott, still, I love the concept of the OS, will probably revisit it in the future
18:27:23 -!- tromp has joined.
18:29:34 <Vorpal> Phantom_Hoover, a really interesting linux distro. Ask elliott about the package manager. Kind of in 2 other conversations already so...
18:30:40 <elliott> I happen to be in two conversations, too!
18:31:01 -!- tromp has quit (Remote host closed the connection).
18:31:17 -!- tromp has joined.
18:36:11 <int-e> of course nixos has a website which may already answer your questions
18:38:15 <elliott> int-e: yeah but this channel is the premium interactive version of the web.
18:39:04 <int-e> I'm just doing my part by stating the obvious. :-P
18:52:02 -!- idris-ircslave has joined.
18:54:49 <int-e> Phantom_Hoover: don't worry, your bill will be in the mail.
18:57:58 -!- nisstyre_ has joined.
18:59:56 -!- tromp has quit (Remote host closed the connection).
19:00:29 -!- tromp has joined.
19:02:12 -!- ket1v has joined.
19:04:47 -!- tromp has quit (Ping timeout: 246 seconds).
19:06:38 -!- ket1v has quit (Ping timeout: 240 seconds).
19:09:41 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
19:49:28 -!- MindlessDrone has quit (Quit: MindlessDrone).
19:52:14 -!- ais523 has joined.
19:55:03 -!- Phantom_Hoover has joined.
20:03:35 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
20:30:59 <zzo38> I wrote some made-up-Pokemon-card. One of them is: [TRAINER] Play an evolution card in opponent's trash onto a card on his side which it can evolve from.
20:31:50 -!- conehead has joined.
20:33:29 <zzo38> Another one is: Put a card from your hand or trash or an energy card attached to one of your pokemon with your side cards. (Cannot be used if you already have 6 side cards.)
20:33:47 <zzo38> Do you like these thiings?
20:36:31 <zzo38> First one is similar to POKEMON FLUTE but it is for evolution cards instead. It might prevent opponent from attacking/retreating due to lack of energy cards, or affect their powers, or can change their weakness/resistance.
20:38:00 <ais523> the first one seems weak to me
20:38:12 <ais523> many decks don't really play evolution cards
20:38:14 <ais523> and even if they do, they may not end up in trash
20:38:22 <ais523> it might work if you could search your opponent's deck for one
20:38:52 <ais523> Deewiant: (re your message) yes, that's believable, and unfortunate
20:38:58 <zzo38> OK, although remember these are designed for the old game, not the modern one.
20:39:07 <ais523> people used evolutions even less in the old game
20:39:22 <ais523> one of the best decks was one that just played as many high-HP basics as possible
20:39:43 <zzo38> Although yes, allowing to search in their draw pile as well as trash would certainly make it more powerful.
20:40:11 <ais523> "side cards" = Prize cards?
20:40:21 <zzo38> ais523: Yes it is another name for that
20:40:32 <zzo38> Not playing many evolution cards also would seem good for drafts and stuff like that
20:40:57 <ais523> that second trainer seems very weak, especially because it doesn't do anything else
20:41:02 <ais523> like, not only do you go down a Prize
20:41:10 <ais523> you also have to play the card itself
20:41:18 <ais523> if it also drew you a card, then that would be less bad
20:41:31 <ais523> and maybe it'd be used by a deck that wanted the effect
20:43:33 <zzo38> Actually it can help in the right circumstancse if you don't also draw a card. One thing you can get the card back, but maybe you also want to get rid of the card from where it currently is In the right circumstances it can be of much help, if your strategy is around getting the card you need, otherwise you may lose time and the opponent may gain time.
20:43:34 <Deewiant> ais523: Relatedly, doing the encode fully in memory is somewhat uncomfortable with multi-gigabyte outputs like that (especially with Java's annoying -Xmx setting requirement to actually use available RAM)
20:44:00 <ais523> Deewiant: I didn't actually realise that people would produce multiple-gigabyte outputs with jettyplay
20:46:10 <ais523> anyway, jettyplay allows pluggable container formats
20:46:11 <zzo38> POKEMON FLUTE can be helpful to get opponent's cards that have powers that benefit you. It can also help if it is a card with a high retreat cost, use GUST OF WIND to activate it and stop them from retreating, especially if you have card that can attack bench pokemon cards.
20:46:12 <Deewiant> Oh, it also crashed quite quickly when using 720p without ZMBV, hang on
20:46:13 <ais523> just AVI is the only one I implemented
20:46:23 <ais523> and yeah, that's pretty much inevitable
20:46:28 <ais523> that's just running out of memory to store uncompressed video
20:46:37 <ais523> really, the raw option's only there because I implemented it first for testing
20:46:38 <Deewiant> No, I mean like on the first frame or so
20:46:50 <Deewiant> Hang on while I try to reproduce
20:46:50 <zzo38> This one with evolution cards, you could evolve opponent's cards to increase their retreat cost, but also to affect their weakness/resistance, for example EEVEE evolving into many different card with different weakness/resistance.
20:47:17 <Deewiant> ais523: java.lang.IllegalArgumentException: Pixel stride times width must be less than or equal to the scanline stride
20:47:34 <Deewiant> On frame 2, actually, according to the dialog
20:47:43 <Deewiant> Or then it's frame 1 and it updated the dialog early
20:48:18 <zzo38> Some cards although ordinarily don't help much, can be extremely good in the correct circumstances.
20:49:40 <ais523> zzo38: and such cards normally aren't worth the opportunity cost of drawing them
20:50:15 <zzo38> Depending on the rest of the cards in the deck, they can be; I have had these experiences a lot.
20:50:29 <Deewiant> ais523: Irrelevant of resolution actually too, I only tried 720 earlier
20:50:39 <zzo38> Also if you are playing some kind of draft formats or other ways the cards are randomized, you may get it anyways.
20:50:49 <ais523> Deewiant: oh that makes more sense, it's probably something simple then
20:50:55 <zzo38> The second card I listed can be very good if you are able to knock out opponent's pokemon card on your turn.
20:51:13 <zzo38> Knocking out bench pokemon card is best so that they cannot freely switch to another active card.
20:52:36 <zzo38> Of course evolution cards are more difficult to play in a drafted game, unless you modify the rules so that they can still be played effectively.
20:59:46 <zzo38> Some cards have some uses which are situational and some are common. Some cards have some uses which have situations which are common enough. Some cards which are good in some situations, can be good with situations involving your own cards, which is good, but then there may also be situations based on opponent's cards, which helps even more.
21:03:00 -!- ket1v has joined.
21:03:10 <zzo38> DIGGER can be good in a situation where your attacks do more damage with more damage on you.
21:04:11 -!- jconn has joined.
21:07:18 -!- ket1v has quit (Ping timeout: 240 seconds).
21:08:53 -!- ^v has joined.
21:11:05 <zzo38> GOOP GAS ATTACK can be good to get rid of opponent's powers, but it is especially good if you also have some cards with your own powers which affect both players; you can then cancel those powers too.
21:14:31 <int-e> so can you build a turing machine with them?
21:16:02 <int-e> (this question was inspired by http://www.toothycat.net/~hologram/Turing/ )
21:16:11 <zzo38> Yes I have seen that.
21:19:48 <zzo38> I have a Team Rocket strategy guide for Pokemon Card. It is really terrible; they totally neglect defense and sacrifice, and miss a lot of other things. They say CHALLENGE! is really good, but I don't think so.
21:20:58 <zzo38> One thing they don't mention is that if you have DARK VAPOREON [Lv.28], and your opponent has a card with high retreat cost and high attack cost, it can very good. Especially, if you also have GENGAR [Lv.38] then you can move damage onto their bench pokemon cards instead.
21:22:01 <zzo38> I think it might be a bit more useful if DARK VAPOREON [Lv.28]'s second attack did less damage, though.
21:29:30 <zzo38> GUST OF WIND is a very good card in a lot of situations (just ensure you don't waste it). You can eliminate the opponent's strong attacking card (possibly before they get a chance to attack, evolve, or use a power). You can select a card with high attack and retreat costs in order to gain time. If you are close to winning, you can even use it to activate a card you can easily knock out. And yet there are many more possibilities!
21:31:17 <zzo38> It is also often a good idea to not knock out opponent's active pokemon cards, although this depends a lot on the circumstances.
21:44:20 -!- Phantom_Hoover has joined.
21:44:28 -!- Phantom_Hoover has quit (Changing host).
21:44:28 -!- Phantom_Hoover has joined.
21:52:41 <zzo38> I suppose it should not be surprising that the strategy guide neglects defense; another book I have says the same thing about books about chess. However, it still also fails to mention various other things too, so that isn't entirely it.
21:55:39 <fizzie> Since I talked about that photo thing here, here's a quick average of all 25 photos I have so far (with histogram equalization to make it a bit less bland): https://dl.dropboxusercontent.com/u/113389132/Misc/20140322-align.jpg
21:55:49 <fizzie> Will play around with the column thing later.
21:59:18 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
22:00:16 <Sgeo> zzo38: do you have to be JOHN to play GUST OF WIND
22:07:07 -!- shikhout has joined.
22:10:07 -!- shikhin has quit (Ping timeout: 264 seconds).
22:10:09 -!- shikhout has changed nick to shikhin.
22:16:33 -!- DaedricBunny has joined.
22:16:46 <fizzie> Also: https://dl.dropboxusercontent.com/u/113389132/Misc/20140322-align2.jpg
22:30:56 -!- pikhq has quit (Ping timeout: 246 seconds).
22:31:16 -!- pikhq has joined.
22:47:19 * Sgeo goes to watch The Best of Both Worlds
22:50:27 <zzo38> Why does the Windows font viewer use "The quick brown fox jumps over the lazy dog" for some fots and "Jackdaws love my big sphinx of quartz" for some fonts?
22:50:48 <zzo38> (Both sentences are pangrams, but I don't know why it has two)
22:53:33 -!- tromp has joined.
22:58:09 -!- MoALTz has quit (Quit: Leaving).
22:59:30 -!- Phantom_Hoover has joined.
23:00:30 -!- oerjan has joined.
23:03:51 -!- ket1v has joined.
23:08:39 -!- Bike has joined.
23:08:46 -!- ket1v has quit (Ping timeout: 268 seconds).
23:12:46 -!- Slereah_ has quit (Read error: Connection reset by peer).
23:13:27 -!- Slereah_ has joined.
23:34:18 -!- Bike has quit (Ping timeout: 240 seconds).
23:38:42 <fizzie> zzo38: Have you discovered some kind of pattern in the fonts where it chooses one or the other, e.g. about their format?
23:47:35 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
23:53:20 <zzo38> fizzie: I tried to, but didn't find any.