00:17:05 <doesthiswork> one interesting thing about natural languages is that less surprising information is placed towards the beginning of a sentence, and more surprising towards the end
00:17:53 <coppro> and is that actually a property of languages or simply a human propensity to express surprise in a, well, surprising manner
00:17:55 <doesthiswork> this might be an attempt to smooth out the bit of information conveyed per secend
00:18:26 <coppro> or simply because of the amount of buildup required to deliver a surprise means that changing the order of the sentence won't move the surprise
00:19:06 <Phantom__Hoover> doesthiswork, or it might be one of the three alternate explanations just presented
00:19:24 <Bike> is there data involved here
00:19:34 <oerjan> is this true of japanese which has a lot in the opposite order of western languages
00:24:39 <doesthiswork> oerjan: japanese is in topic comment order is indeed different from english's subject predicate order
00:25:06 <doesthiswork> but what I"m talking about are the situations where there is more than one phasing availible
00:27:30 <doesthiswork> like "the cub scouts will hold the car wash despite the rain" vs "despite the rain the car wash will still be held by the cub scouts" vs "the car wash will still be held by the cubscouts despite the rain"
00:28:09 <doesthiswork> they communicate the same thing but seem to imply a different preceding conversation
00:29:10 <Phantom__Hoover> well, except in that i'd expect it to be phrased "despite the rain the cub scouts will still hold the car wash"
00:32:06 <doesthiswork> were you expecting the latter to sound less natural?
00:33:32 <Phantom__Hoover> well like you've given an example but you've done nothing to convince me that this supposed effect favours the former phrasing
00:34:08 <doesthiswork> different phrasings would be favored in different situations
00:34:10 <Bike> no, seriously, does this effect actually exist
00:37:21 <doesthiswork> bike: it will take me awhile to find you a citation because it was in my textbook
00:42:38 <kmc> great, nvidia GL library seems to map itself with rwx permissions
00:43:20 <Phantom__Hoover> doesthiswork, (sorry, but with linguistics especially people keep presenting dumb armchair speculation)
00:43:45 <Bike> information content of language is an interesting thing but rather easy to fuck up
00:44:00 <Sgeo> http://www.reddit.com/r/PHP/comments/1l7baq/creating_a_user_from_the_web_problem/
00:44:58 <doesthiswork> I can find a prescriptive recomendation to put old knowledge first, but I haven't found proof that that is what people do in natural speach
00:45:23 <Bike> yeah, it sounds like the kind of thing that would take a big corpus and a lot of boring interpretation by human researches
00:45:34 <Bike> researches* not that linguists don't do that, i mean
00:45:51 -!- Koen_ has quit (Quit: Koen_).
00:47:17 <Bike> the most baffling part of that link is that there are serious, helpful responses
00:48:02 <doesthiswork> I believe that people put the old information first when ordering clauses, as much as the syntax of their language allows.
00:51:32 <doesthiswork> because a professor of linguistics assumed so, and because it seems a reasonable thing for people to do for a variety of reasons.
00:52:54 <doesthiswork> yes, and I don't have the statistical tools to prove it
00:54:23 <doesthiswork> english has quite fixed word order, so some orderings that would work in spanish are not natural english sentences
00:56:20 -!- mnoqy has joined.
00:58:11 <Sgeo> Now this is just depressing, in a 'failing to learn' way
00:58:11 <Sgeo> http://www.reddit.com/r/PHP/comments/1l7baq/creating_a_user_from_the_web_problem/cbwj0r2
00:58:22 <Bike> you're actually reading it
00:58:51 <Sgeo> Maybe I should try this http://www.dvwa.co.uk/
01:06:11 <doesthiswork> bike: no proof here either but references to people who first assumed it http://web.mit.edu/newsoffice/2012/applying-information-theory-to-linguistics-1010.html
01:07:41 <Bike> it's positively lovely how bad these sorts of articles always are
01:07:49 <Bike> "information theory — the discipline that gave us digital communication"
01:08:25 <Bike> i think most people make the one and kind of give up after that
01:08:28 <Bike> it's just, too obvious.
01:09:18 <kmc> are there any Coq papers in PNAS
01:10:10 <Bike> i don't think th- oh, i get it
01:11:49 <Bike> « our “open-source” large-scale (∼1 L/h) 129Xe polarizer for clinical, preclinical, and materials NMR and MRI research» i love the future
01:12:59 <Bike> um i have no idea.
01:13:28 <Bike> http://www.pnas.org/content/110/35/14150.abstract
01:13:41 <Bike> there's a paper called "Core foundations of abstract geometry" that isn't math. awesome.
01:13:53 <kmc> is it about Core Foundation
01:15:59 <Bike> "In this study, we performed continuous electroencephalography in rats undergoing experimental cardiac arrest " jesus christ
01:16:44 <Bike> those cute little rats!!
01:18:00 <Bike> are you suggesting i give ancaps heart attacks to see what their brains do
01:18:58 -!- Nisstyre has quit (Quit: Leaving).
01:19:01 -!- carado has quit (Ping timeout: 246 seconds).
01:19:17 -!- carado has joined.
01:22:39 <oerjan> i'd imagine ancaps would be more fun to put in a large labyrinth with various resources.
01:23:05 <oerjan> assuming it's the abbreviation i think it is
01:25:01 <oerjan> this guy can head the project http://girlgenius.wikia.com/wiki/Doctor_Mittelmind
01:29:35 <Bike> mittelmind is one of my favorite characters
01:36:47 -!- augur has quit (Remote host closed the connection).
01:43:46 -!- Bike has quit (Ping timeout: 240 seconds).
01:44:57 -!- copumpkin has quit (Remote host closed the connection).
01:45:32 -!- copumpkin has joined.
01:51:06 -!- Phantom__Hoover has quit (Remote host closed the connection).
02:08:33 -!- oerjan has quit (Quit: leaving).
02:19:15 <kmc> https://bugs.launchpad.net/ubuntu/+source/keepassx/+bug/1214844 *facepalm*
02:21:24 -!- augur has joined.
02:21:36 <^v> Any characters in your password that are not found in codepage 1252 are silently squashed to '?'
02:21:49 <^v> thats fucking hilarious
02:22:59 -!- trout has changed nick to variable.
02:24:11 * ^v slaps variable with a trout
02:24:53 <^v> (({("XSRX;GSA@uGS0Xd;]"):gsub("(.)",function(c) return string.char(49+((string.byte(c)*1337)%75)):gsub("_"," ") end)})[1].." "):rep(20)
02:24:53 <^v> <-- put this into a lua prompt
02:25:04 <^v> i maed some obufuscated lua cod
02:26:19 -!- nooodl has quit (Ping timeout: 264 seconds).
02:40:37 -!- Bike has joined.
02:42:18 <Bike> "Skimming an old report on Snobol" at least someone on the internet understands me.
02:43:41 -!- monotone_ has changed nick to polytone.
03:06:06 -!- jconn has quit (Ping timeout: 264 seconds).
03:09:47 <Sgeo> http://ideone.com/jw9i6s
03:09:50 <Sgeo> Didn't seem to do much
03:10:19 <Bike> whaddya mean, it says returned zero right there
03:11:28 <Bike> perfectly respectable thing to return
03:13:58 <Bike> hm, gsub apparently treats string patterns as, well, literal strings, rather than interpreting them as regexes.
03:14:01 <Bike> is that the problem?
03:15:03 -!- anarckhia has joined.
03:42:37 <shachaf> kmc: https://plus.google.com/108532745084733145449/posts/gzLPnqH8HHH
03:48:28 -!- carado has quit (Ping timeout: 246 seconds).
04:42:37 -!- c33333 has joined.
04:43:36 <c33333> Looking for esoteric, nirgle, fusion or mikeh
05:01:12 -!- doesthiswork has quit (Quit: Leaving.).
05:53:37 -!- lekrel has joined.
05:53:56 <lekrel> I've decided to put up an interesting challenge. It has more to do with cryptography but also related to programming. Here is what you ought to do in order to complete it: 1. Generate a 192-bit triple DES ciphertext using your alias as the key and plaintext for the encryption, 2. Write a Malbolge program that would output the ciphertext
05:54:01 <lekrel> 3. PM me the key (a.k.a your alias) along with the malbolge program source
05:54:04 <lekrel> You'll need an interpreter/compiler for the malbolge part. Good luck.
05:54:06 -!- ^v has quit (Read error: Connection reset by peer).
06:10:45 -!- jconn has joined.
06:19:40 -!- c33333 has quit (Ping timeout: 264 seconds).
07:03:46 <kmc> shachaf: :D
07:05:12 <kmc> it was actually built in the 30s though
07:06:10 -!- doesthiswork has joined.
07:20:43 <olsner> which means that during the 1960s, it was (already) built
07:23:27 <kmc> "For security reasons, the effect of MALLOC_CHECK_ is disabled by default for set-user-ID and set-group-ID programs. However, if the file /etc/suid-debug exists (the content of the file is irrelevant), then MALLOC_CHECK_ also has an effect for set-user-ID and set-group-ID programs."
07:23:51 <Fiora> huh, why would that be?
07:24:06 <Fiora> disabling it for SUID I guess?
07:25:05 <kmc> I guess so that setuid programs can be compiled with checking enabled and an attacker can't disable it as part of exploiting a security hole
07:25:41 <Fiora> but didn't that say it was disabled for setuid?
07:25:52 <kmc> the environment variable is disabled
07:26:04 <kmc> sorry, I did not paste the whole context
07:26:23 <Fiora> and the environment variable lets you toggle it?
07:26:40 <kmc> this is from mallopt(3); programs can call this function to change malloc behavior, but also some of the settings have env vars
07:27:12 <kmc> but for setuid programs the env vars are ignored
07:28:39 <kmc> Openwall has no setuid binaries ^_^
07:29:12 <kmc> they got the Linux kernel to provide an ICMP socket type so that ping would not be setuid
07:29:21 <kmc> it's a security-focused linux distribution
07:31:55 <lekrel> I've decided to put up an interesting challenge. It has more to do with cryptography but also related to programming. Here is what you ought to do in order to complete it: 1. Generate a 192-bit triple DES ciphertext using your alias as the key and plaintext for the encryption, 2. Write a Malbolge program that would output the ciphertext
07:32:00 <lekrel> 3. PM me the key (a.k.a your alias) along with the malbolge program source
07:32:05 <lekrel> You'll need an interpreter/compiler for the malbolge part. Good luck.
07:38:51 <coppro> Does anyone know a data structure for representing transitive closures?
07:39:41 <coppro> e.g. I want to represent that a < b and b < c and efficiently find that a < c
07:41:19 <coppro> I would ideally like constant time, up to the underlying (tree, hash, etc.) data structure
07:44:41 <kmc> what about just storing the transitive closure as an adjacency matrix
07:47:33 <kmc> https://thoughtstreams.io/glyph/string-trepanation/
08:05:05 <elliott> lekrel: are you mafingre/etc.?
08:05:39 <elliott> hmm same country but different ISP
08:17:34 -!- lekrel has quit (Quit: Page closed).
08:18:06 -!- doesthiswork has quit (Quit: Leaving.).
08:19:04 <fizzie> kmc: dbus-driven mechanisms to give ping CAP_NET_RAW hth
08:31:21 -!- epicmonkey has joined.
08:43:34 -!- SingingBoyo has joined.
08:47:29 <fizzie> "Too many workers requested. The job requires 44, but the cluster with profile 'Triton_latest_local' supports maximum of 4."
09:21:08 <Deewiant> Stay away from Triton, I've been trying to get practically anything to run there for weeks.
09:23:25 <kmc> attempt no landings there
09:24:44 -!- asie has joined.
09:26:42 -!- SingingBoyo has quit (Quit: Leaving).
09:31:11 <fizzie> All these other wo^H^Hclusters are okay, though?
09:32:54 <fizzie> Also I don't understand: I ran the MDCS thing with matlabpool(44) and -- according to squeue -- it tried to allocate 44 *nodes*; I tried again with matlabpool(4) and it ran on one node with 4 workers; so I went with 8 and 12, and those mapped into 8 and 12 nodes again.
09:33:26 <fizzie> I guess maybe the "nodes" field of squeue has different meanings for pending and running jobs.
09:33:53 <fizzie> Based on http://sprunge.us/XFRW anyway.
09:34:13 <fizzie> (This is going to take ages parallelized to 8 tasks only.)
09:35:48 <fizzie> It runs with --exclusive, don't know if that makes a difference.
09:35:49 <Deewiant> Maybe the matlab stuff does strange things.
09:36:07 <fizzie> It's certainly very flaky.
09:36:38 <fizzie> I think a lot of people are running MATLAB jobs outside MDCS, which, as far as I know, is explicitly against the license terms.
10:33:26 -!- nooodl has joined.
10:45:41 -!- MindlessDrone has joined.
11:08:49 -!- Koen_ has joined.
11:24:06 -!- asie has quit (Remote host closed the connection).
11:24:24 -!- asie has joined.
11:29:06 -!- Sgeo has quit (Read error: Connection reset by peer).
11:37:56 -!- asie has quit (Remote host closed the connection).
11:45:53 -!- carado has joined.
11:46:26 -!- asie has joined.
11:54:31 -!- epicmonkey has quit (Ping timeout: 260 seconds).
11:55:32 -!- asie has quit (Remote host closed the connection).
12:19:26 -!- oklopol has quit (Ping timeout: 264 seconds).
12:52:31 -!- conehead has quit (Quit: Computer has gone to sleep.).
12:58:18 -!- anarckhia has quit (Ping timeout: 264 seconds).
13:02:06 -!- boily has joined.
13:03:15 <Roujo> Say I want to put in a multi-line script into HackEgo... How do I do that?
13:03:26 <Roujo> Can you echo multi-line files?
13:04:24 <boily> `run echo -e "My hovercraft\nIt is full of eels."
13:04:25 <HackEgo> My hovercraft \ It is full of eels.
13:04:45 <boily> Roujo: yes, it works. the «\» is a newline indicator.
13:06:10 <boily> also, there's a command to slurp up a script from the web or something, but I can't remember it.
13:06:22 <boily> Gregor: is it still possible to do something like that ↑?
13:10:06 -!- epicmonkey has joined.
13:10:24 <Roujo> Hmmm... My changes won't go through =/
13:10:32 <HackEgo> Cats are cool, but should be illegal.
13:10:51 <Roujo> `run echo Cats are cool, and should be illegal. > wisdom/cat