00:01:48 -!- atrapado has quit (Quit: Bye).
00:20:25 <oerjan> olsner: what's a swedish rabbit who thinks everyone should listen to a bit of carmina burana
00:21:20 <kmc> hmmmmm i have some bacon wrapped filets mignon in my freezer hmmmmmmmmmmmmmmmmmmm
00:22:19 <kmc> 2 lazy 2 cook
00:24:07 * oerjan thinks olsner might be idle.
00:25:22 <kmc> i want to know the answer!
00:25:40 <oerjan> even though you don't know swedish?
00:26:12 <kmc> that's a risk i'll just have to take
00:27:09 <shachaf> whoa, whoa, whoa, did you know that monomorphisms and epimorphisms are both a form of injectivity thing
00:27:17 <shachaf> ie injective and surjective functions
00:27:42 <shachaf> in particular "f is injective" means "(f .) is injective" and "f is surjective" means "(. f) is injective"
00:29:41 <shachaf> (except this works in any category, not just (->), otherwise it would be a silly definition)
00:30:10 -!- prooftechnique has joined.
00:30:31 <shachaf> and when they're surjective you get something stronger (but backwards)
00:30:34 <kmc> shachaf: it disturbs me that two of your three commas are bolded
00:31:02 <shachaf> do bold spaces disturb you
00:31:13 <shachaf> the whole phrase is bolded. "whoa, whoa, whoa" is atomic
00:31:49 <shachaf> anyway if (f .) is surjective then f has a left inverse?? and (. f) -> right inverse? or something like that
00:31:50 <Sgeo> I'm a Cablevisionary!
00:32:56 <Sgeo> (Note: I'm not actually that excited about the cutesy name for Cablevision employees)
00:33:03 <nooga> I just discovered that my classic, 81 Mercedes C123 coupe has its left wind welded into the ramp
00:35:01 <shachaf> i keep thinking that you are nooodl and then i realize that you are not nooodl so who even are you
00:35:32 <shachaf> but nooodl would never say a thing like that. what's a wind anyway
00:35:51 <shachaf> i can't grep logs. no logs
00:36:04 <nooga> http://tunes.org/~nef/logs/esoteric/ ?
00:36:31 <nooga> wget them all and then can and grep
00:36:31 -!- glogbackup has quit (Remote host closed the connection).
00:37:45 <shachaf> nooodl: we just don't know
00:38:17 <nooga> shachaf: you'll discover that I haunted this # for a long time
00:39:14 <nooga> oh just grep the logs and infer who am I from it
00:42:15 <nooga> maybe this isn't the best idea
00:43:22 <nooga> anyway, who's noodl?
00:43:46 <oerjan> <shachaf> anyway if (f .) is surjective then f has a left inverse?? and (. f) -> right inverse? or something like that
00:44:10 <oerjan> getting identity and getting everything seems equivalent
00:45:54 <oerjan> <Sgeo> (Note: I'm not actually that excited about the cutesy name for Cablevision employees) <-- do you have a company anthem twh
00:47:42 <oerjan> shachaf: nooga is the pole, that's easy to remember.
00:48:01 <nooga> oerjan never fails
00:48:03 <shachaf> we've had other poles in here
00:49:28 <nooga> shachaf: tbh i don't know who are you, you must be new here then
00:50:14 <HackEgo> noooodl is the correct spelling
00:51:16 <shachaf> imo add a special case to bin/? for no+dl
00:51:32 <oerjan> there is one, but maybe not in the sense you want
00:51:46 <shachaf> you mean the one for output?
00:52:16 <HackEgo> #!/bin/sh \ topic=$(echo "$1" | lowercase | sed "s/ *$//") \ topic1=$(echo "$topic" | sed "s/s$//") \ cd wisdom \ if [ \( "$topic1" = "ngevd" \) -a \( -e ngevd \) ]; \ then cat /dev/urandom; \ elif [ -e "$topic" ]; \ then cat "$topic" | rnooodl; \ elif [ -e "$topic1" ]; \ then cat "$topic1" | rnooodl; \ else echo "$1? ¯\
00:52:20 <shachaf> it's no{3,9}dl or something
00:52:26 <HackEgo> perl -pe 's/nooodl/"n@{[o x(3+rand 7)]}dl"/ge'
00:52:39 <nooga> where are those prety graphs derived from the logs
00:52:51 <nooga> AFAIR fizzie made them
00:55:08 <nooga> hah, can't find them
00:55:08 <oerjan> `run sed -i '2s!s/!s/no\+dl/nooodl/;s/!' bin/?
00:55:23 -!- glogbackup has joined.
00:55:27 <HackEgo> #!/bin/sh \ topic=$(echo "$1" | lowercase | sed "s/no+dl/nooodl/;s/ *$//") \ topic1=$(echo "$topic" | sed "s/s$//") \ cd wisdom \ if [ \( "$topic1" = "ngevd" \) -a \( -e ngevd \) ]; \ then cat /dev/urandom; \ elif [ -e "$topic" ]; \ then cat "$topic" | rnooodl; \ elif [ -e "$topic1" ]; \ then cat "$topic1" | rnooodl; \ el
00:55:54 <oerjan> `run sed -i '2s!s/!s/no\\+dl/nooodl/;s/!' bin/'?'
00:56:06 <HackEgo> noooooooodl is the correct spelling
00:56:23 <HackEgo> nooooooodl is the correct spelling
00:56:25 <nooga> can I `run perl -e "fork while fork" ?
00:56:28 <shachaf> `? noooooooooooooooooooooooooooooooooooooooooodl
00:56:30 <HackEgo> noooooooodl is the correct spelling
00:56:40 <ais523> is the correct spelling random?
00:56:54 <ais523> is it ever the same as the original query?
00:57:43 <shachaf> It's between 3 and 9 os, I think?
00:57:48 <HackEgo> perl -pe 's/nooodl/"n@{[o x(3+rand 7)]}dl"/ge'
01:01:52 <HackEgo> #!/bin/sh \ topic=$(echo "$1" | lowercase | sed "s/no\+dl/nooodl/;s/ *$//") \ topic1=$(echo "$topic" | sed "s/s$//") \ cd wisdom \ if [ \( "$topic1" = "ngevd" \) -a \( -e ngevd \) ]; \ then cat /dev/urandom; \ elif [ -e "$topic" ]; \ then cat "$topic" | rnooodl; \ elif [ -e "$topic1" ]; \ then cat "$topic1" | rnooodl; \ e
01:02:18 <shachaf> `run sed -i '2s/no/noo/' bin/\?
01:02:31 <HackEgo> noooooooodl is the correct spelling
01:02:35 <nooga> looks like something is broken
01:02:41 <shachaf> because one o is just broken
01:03:12 <oerjan> `learn nooodles are the invention of the chinese. they were brought to europe by marco polo, a distant ancestor of taneb.
01:03:33 <HackEgo> noooooodles are the invention of the chinese. they were brought to europe by marco polo, a distant ancestor of taneb.
01:03:36 <nooodl> i've set up highlights on /no+dl/ before. people have called me nodl before as a result! history repeats itself
01:04:18 <shachaf> oerjan: what's with not capitalizing wisdom entries
01:04:41 <oerjan> shachaf: rnoodle isn't case insensitive
01:05:32 <HackEgo> nooga hate OS X. NOOGA SMASH. Hug not allowed.
01:05:46 <ais523> programmers use caps conventions for namespacing all the time
01:06:02 <ais523> so in the future, we can make `? ALLCAPS do something completely unrelated to wisdom, for instance
01:06:09 -!- Sorella has quit (Read error: Connection reset by peer).
01:06:15 <shachaf> I'm talking about the text, not the file name.
01:06:35 <ais523> that's namespaced too, so if it's written in sentencecase, you know it's sarcastic, or something
01:06:47 <oerjan> ais523: ? _is_ case insensitive.
01:07:16 <ais523> sure, have you ever seen an uppercase '?'?
01:07:30 <HackEgo> nooodl is the correct spelling
01:07:32 -!- glogbackup has quit (Remote host closed the connection).
01:07:55 <shachaf> `run sed -i 's/n/N/;s/ t/ T/;s/taneb/Taneb/' wisdom/nooodles
01:07:56 <HackEgo> sed: can't read wisdom/nooodles: No such file or directory
01:08:01 <shachaf> `run sed -i 's/n/N/;s/ t/ T/;s/taneb/Taneb/' wisdom/noodles
01:08:03 <HackEgo> sed: can't read wisdom/noodles: No such file or directory
01:08:04 <HackEgo> <oerjan> YOU are out of order.
01:08:34 <HackEgo> wisdom/nooodl wisdom/nooodle
01:08:36 <ais523> just as a hint, if an op goes and edits something to say you shouldn't do it
01:08:46 <ais523> you should probably take the hint
01:09:58 <ais523> oerjan: don't deny it, this is hilarious
01:09:59 * oerjan adds to the what? queue
01:10:42 <ais523> I have no idea of the context
01:10:44 <HackEgo> nooga hate OS X. NOOGA SMASH. Hug not allowed.
01:10:54 <oerjan> ais523: i cannot undeny things i don't understand hth
01:10:58 <shachaf> `run sed -i 's/n/N/;s/ t/ T/;s/taneb/Taneb/' wisdom/noodle
01:11:00 <HackEgo> sed: can't read wisdom/noodle: No such file or directory
01:11:05 <shachaf> `run sed -i 's/n/N/;s/ t/ T/;s/taneb/Taneb/' wisdom/nooodle
01:11:36 -!- Sorella has joined.
01:11:42 <ais523> oerjan: I was interpreting HackEgo's reponse as you disabling HackEgo entry
01:12:20 <ais523> I'm tired and incoherent
01:12:53 -!- ais523 has left ("I'm too incoherent for meaningful conversation").
01:13:11 <oerjan> i don't actually have the power to do that, except by kicking HackEgo
01:14:07 <HackEgo> Nooodles are the invention of the chinese. They were brought to europe by marco polo, a distant ancestor of Taneb.
01:14:28 <oerjan> shachaf: you didn't fix rnooodl to match
01:15:05 <shachaf> rnooodl is case-sensitive.
01:15:33 <oerjan> `learn Nooodles are the invention of the chinese. They were brought to Europe by Marco Polo, a distant ancestor of Taneb.
01:15:35 <HackEgo> /hackenv/bin/learn: line 4: wisdom/: Is a directory \ I knew that.
01:15:41 <oerjan> `learn Nooodles are the invention of the chinese. They were brought to Europe by Marco Polo, a distant ancestor of Taneb.
01:15:56 <shachaf> oh, i guess those are names too
01:16:09 <oerjan> `learn Nooodles are the invention of the Chinese. They were brought to Europe by Marco Polo, a distant ancestor of Taneb.
01:16:33 <HackEgo> perl -pe 's/nooodl/"n@{[o x(3+rand 7)]}dl"/ge'
01:17:14 -!- tromp has joined.
01:17:19 <shachaf> `run echo >bin/rnooodl 'perl -pe '\''s/[Nn]ooodl/"\1@{[o x(3+rand 7)]}dl"/ge'\'
01:17:43 <shachaf> `run echo 'nooodl nooodle' | rnooodl
01:17:46 <nooodl> i think you need ([Nn])
01:17:54 <nooodl> or is it \([Nn]\) who knows
01:17:59 <shachaf> in my head i wrote the parentheses
01:18:12 <shachaf> `run echo >bin/rnooodl 'perl -pe '\''s/([Nn])ooodl/"\1@{[o x(3+rand 7)]}dl"/ge'\'
01:18:13 <nooodl> U+1F04A MENTAL LEFT PARENTHESIS
01:18:18 <shachaf> `run echo 'nooodl nooodle' | rnooodl
01:18:24 -!- kmc has left.
01:18:51 <oerjan> did we break HackEgo again
01:18:56 <HackEgo> perl -pe 's/[Nn]ooodl/"\1@{[o x(3+rand 7)]}dl"/ge'
01:19:12 <shachaf> `run echo 'nooodl nooodle' | rnooodl
01:19:13 <HackEgo> .oooooooodl .ooooooooodle
01:19:21 <shachaf> `run echo >bin/rnooodl 'perl -pe '\''s/([Nn])ooodl/"\1@{[o x(3+rand 7)]}dl"/ge'\'
01:19:24 <shachaf> `run echo 'nooodl nooodle' | rnooodl
01:19:32 <HackEgo> perl -pe 's/([Nn])ooodl/"\1@{[o x(3+rand 7)]}dl"/ge'
01:20:03 <nooga> this gets almost as annoying as this haskell bot i remember from a while ago
01:20:46 <oerjan> `run echo >bin/rnooodl 'perl -pe '\''s/([Nn])ooodl/"$1@{[o x(3+rand 7)]}dl"/ge'\'
01:20:59 <oerjan> `run echo 'nooodl nooodle' | rnooodl
01:21:16 <nooodl> nooga: imagine being me. imagine the pings
01:22:01 <HackEgo> Noooooodles are the invention of the Chinese. They were brought to Europe by Marco Polo, a distant ancestor of Taneb.
01:22:43 <nooga> `run cat bdsmreclist
01:22:45 <HackEgo> <oerjan> YOU are out of order.
01:22:46 <oerjan> also, boily will probably kill us
01:25:14 <oerjan> nooga: when he has to update the pdf
01:25:41 <nooga> oh, he does that by hand?
01:26:01 <nooga> anyway, where's elliott?
01:32:21 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
01:35:16 -!- glogbackup has joined.
01:37:19 <oerjan> glogbackup: WHY ARE YOU HERE
01:45:08 -!- yorick has quit (Remote host closed the connection).
01:45:28 <shachaf> oerjan: you should have written your own stackoverflow answer rather than editing the other one
01:45:38 <shachaf> that way you'd have gotten karma instead of rejected
01:45:51 <oerjan> maybe. i see someone else did.
01:46:49 <oerjan> i just hadn't understood edits were not meant to be used that way, i'm a wiki man you know
01:47:14 <shachaf> i have no idea how edits are meant to be used
01:48:24 <oerjan> well my edits were rejected by 3 people, all of which were shown as usually accepting edits, so i must have done _something_ wrong.
01:48:33 <oerjan> (and 1 person accepted.)
01:50:00 <oerjan> i don't know if you are allowed to open this link but http://stackoverflow.com/review/suggested-edits/3792784
02:08:55 -!- Sellyme has quit (Excess Flood).
02:11:04 -!- Sellyme has joined.
02:12:55 -!- kmc has joined.
02:16:06 <kmc> more projects from the person who made the rule 110 scarf: http://www.ravelry.com/projects/fbz
02:16:20 <kmc> and more info on how she programmed the knitting machine http://www.ravelry.com/projects/fbz/wolfram-1d-cellular-automata-01001001-scarf http://fabienne.us/tag/knittingmachine/
02:16:30 <kmc> now i want a knitting machine, knitting machines are the new 3D printers
02:20:57 -!- nooga has quit (Ping timeout: 245 seconds).
02:31:36 -!- Sellyme has quit (Excess Flood).
02:32:03 -!- Sellyme has joined.
02:35:43 -!- tromp has quit (Remote host closed the connection).
02:43:18 -!- tromp has joined.
02:49:34 <oerjan> <int-e> we all know that it takes a steel foundry to really destroy malware. <-- now i'm wondering if it is possible to make a terminator^Wcomputer out of materials that can survive that
02:55:32 <pikhq> Knee pain makes me feel like an old man. :(
03:01:11 <oerjan> oh the knees. that's two parts of my body that usually _don't_ hurt.
03:01:49 <oerjan> well, not more than the legs in general, anyway.
03:02:12 <pikhq> Yeah, but you are older than time.
03:02:36 <pikhq> Actually, no, you aren't. Time started 1970, right?
03:02:56 <oerjan> well i was probably _conceived_ before time.
03:09:53 -!- Sellyme has quit (Excess Flood).
03:10:03 -!- Sellyme has joined.
03:10:59 <quintopia> oerjan: lucky you. can i trade knees w/ you?
03:11:17 <pikhq> No, they're mine now! HAHAHA!
03:11:24 <oerjan> only if you take the rest of my body too
03:15:39 -!- nys has quit (Quit: sleep).
03:28:53 -!- Sorella has quit (Quit: It is tiem!).
03:30:46 -!- nooodl has quit (Quit: Leaving).
03:34:23 <quintopia> oerjan: including your brain? that can only work out well for me i'm sure
03:34:39 * quintopia wonders what can be done with an extra brain
03:50:30 -!- zzo38 has joined.
04:02:52 <Sgeo> I was in a Windows store today
04:02:54 <Sgeo> They had a MakerBot
04:03:09 <Sgeo> Showing off that Windows 8 comes with 3d printer drivers, apparently
04:06:26 <oerjan> @tell ais523 <ais523> if you do that, you have a "primitive recursive" function, equivalence is decidable for those <-- i'm _very_ skeptical. it seems to me that you can encode something like "are there odd perfect numbers" as "is this primitive function which checks whether an odd number is perfect equivalent to const False".
04:06:56 <oerjan> @tell ais523 *primitive recursive
04:09:02 <oerjan> @tell ais523 actually more clearly, "does this TM halt" is would be decidable as "is 'does this TM halt in n steps' equivalent to const False"
04:12:49 -!- mauke has quit (Disconnected by services).
04:12:59 -!- mauke has joined.
04:15:14 -!- preflex has quit (Ping timeout: 272 seconds).
04:15:25 -!- preflex_ has joined.
04:15:50 -!- preflex_ has changed nick to preflex.
04:18:21 <zzo38> How do I determine my ANI? (MY-ANI-IS doesn't work anymore.)
04:18:29 <oklopol> my knees only hurt after reeeally long periods of sitting without movement, but they do make these clicky sounds on occasion
04:19:18 <oerjan> @tell Bike <Bike> does that converge slow? <Bike> like the rational approximations do <-- the best rational approximations to φ _are_ fib(n+1)/fib(n).
04:19:45 <oerjan> oklopol: you're probably a robot
04:21:20 -!- drlemon has joined.
04:26:07 -!- prooftechnique has quit.
04:27:58 <Bike> oerjan: so yes
04:30:31 <shachaf> time for a 2GB dist-upgrade
04:31:01 -!- kmc has left.
04:32:12 <oerjan> i think 2GB was about how much hard disk my last linux machine back at the university had in total.
04:32:48 <shachaf> it took over 12 minutes to download :'(
04:35:14 <shachaf> if only my internet connection at home was that fast
04:49:47 <zzo38> There are services to access other people's ANI, but I wish to access my own ANI; I have no need for other people's.
04:54:24 <zzo38> (For example in order to know the line class and telephone number of a line dialed from, including my own )
04:59:57 -!- zzo38 has quit (Remote host closed the connection).
05:08:00 -!- oerjan has quit (Quit: leaving).
05:23:58 -!- tromp has quit (Remote host closed the connection).
05:24:30 -!- tromp has joined.
05:28:45 -!- tromp has quit (Ping timeout: 245 seconds).
05:44:16 -!- madbr has joined.
05:44:49 <madbr> why is sin() in radians
05:45:02 <Bike> radians are the true unit.
05:45:28 <madbr> radians are cool for calculus
05:46:40 <madbr> and internally of course it does a taylor series which is easier to express in radians
05:46:42 <madbr> but for like applied math it should be in turns really
05:50:53 <madbr> I use sin(stuff * 2 * 3.141592653589793) way more than sin(stuff)
05:51:08 <quintopia> madbr: sin() is in radians for many reasons. but one good reason is so that lim x->0 sin(x)/x = 1
05:51:48 <madbr> yeah that's calculus related
05:52:05 <oklopol> also very much geometry related
05:52:24 <madbr> I get why in calculus stuff you want to use radians
05:52:41 <quintopia> measuring angles as arcs of the unit circle subtended makes a lot of things nice
05:52:42 <oklopol> sin(x) means you travel along the circle for distance x (upward, starting from the rightmost point), and check how high you got
05:52:57 <madbr> but every time I use the C++ sin() functions it's never in that kind of application
05:53:01 <oklopol> "lim x->0 sin(x)/x = 1" states that you will initially go up
05:53:03 <quintopia> another good reason is that it makes it really simple to state the definition of sine in terms of exp and log
05:53:20 <madbr> I use sine for generating sine wave in sound, where radians are totally useless
05:53:31 <oklopol> if you used turns instead, then angle x would not correspond to that much movement along the circle
05:53:37 <madbr> and fft factors (radians are also useless there)
05:53:37 <quintopia> which makes it easy to write the sine program
05:53:55 <madbr> and various video game applications (where radians are also useless)
05:54:12 <quintopia> madbr: if you need a different sine, just define a sineturn macro.
05:54:16 <oklopol> okay calculus related in that sense, then what i said was sort of useless
05:54:55 <madbr> doesn't the e^i kind of stuff rotate every 1 or 2 units?
05:54:58 <oklopol> for stuff other than math, it doesn't really matter too much what you use
05:56:03 <oklopol> i don't understand the question; however, e^(2pi * a) is a complex number that has length 1 and points in direction a as a vector, where a is in turns
05:56:32 <oklopol> that's basically what you said i guess
05:56:46 <oklopol> just no explicit variable in there
05:56:51 <madbr> quintopia : ah, right
05:57:22 <madbr> the one that rotates every 2 is i^x
05:57:24 <oklopol> but yeah that's another good reason for using radians
05:57:39 <quintopia> madbr: if you actually cared about the distance traveled around the circle, you'd want radians. i suppose you never do.
05:58:07 <madbr> yeah I don't commonly use arcs
05:58:16 <madbr> or calculate the perimeter of stuff
05:58:30 <quintopia> too bad. computational geometry is awesome
05:58:48 <madbr> the computational geometry I do is 3d graphics :D
05:58:58 <oklopol> quintopia: that's a bit of a sweeping statement
05:59:19 * oklopol wonders if that pun is gettable
05:59:22 <quintopia> oklopol: well my floor is kind of dirty, so i needed one
06:00:12 <madbr> kinda wonder how common e^stuff or ln() is in applied code too
06:00:13 <quintopia> oklopol: not very good pun no. it's nowhere near randall's "Kepler the janitor" joke
06:00:49 <oklopol> except i totes didn't get it (i do now)
06:01:37 <quintopia> madbr: you mentioned fourier transforms. if you want a real one (not a fft) you'd need exp and log.
06:01:48 <quintopia> but you don't need radian cos for DCT?
06:02:18 <madbr> quintopia: doesn't it use sin() and that's it?
06:04:47 <quintopia> madbr: it is defined in terms of complex exp. which means you could use sin and cos also, but you'd want the complex versions, which would be internally defined by complex exp and log
06:05:10 <madbr> except no cpu supports complex numbers :D
06:10:43 <madbr> and it's not just done "in software"? (ie by compositing regular floating point operations)
06:12:06 <quintopia> of course it is. your libraries would normally do it that way
06:13:45 <madbr> then it's just a fancy way of writing sin/cos no?
06:14:50 <madbr> dunno why you'd want a non-fft fourier transform anyways :D
06:16:35 <oklopol> i thought there would be hardware support for matrix multiplication, in which case also complex multiplication would've been plausible
06:16:55 <oklopol> but i can't find any evidence of such support, so maybe it's just my imagination
06:17:19 <madbr> I'm not sure putting that many FPU's together makes sense
06:18:00 <oklopol> well, everyone who does anything with graphics will be multiplying matrices
06:18:13 -!- conehead has quit (Quit: Computer has gone to sleep.).
06:18:18 <oklopol> the question is if there's _anything_ the gpu designers can do to make that faster
06:18:44 <madbr> yeah but you have to multiply a hell of a lot of matrixes to make a matrix multiply opcode have any gain
06:18:57 <madbr> not to mention there's already a dot product opcode in SSE
06:19:17 <oklopol> but people _do_ multiply a hell of a lot of matrices
06:19:40 <oklopol> at least if you do any sort of 3d stuff
06:19:52 <madbr> well, you do a lot of vector*matrix
06:20:24 <madbr> which is what I think sse was designed for
06:20:38 <oklopol> usually, what you have when rendering is three matrices, model, view and projection
06:21:08 <madbr> yeah and you have to multiply them together, but that's like once per model
06:21:15 <oklopol> well okay i guess what i'm about to say can be done faster without matrix multiplication
06:22:03 <oklopol> okay i agree if we discount matrix * vector, then there may be no true gain
06:22:22 <oklopol> (i counted that as matrix * matrix before you mentioned it)
06:22:54 <madbr> simd instruction sets are designed with matrix*vector in mind actually
06:22:55 -!- Sprocklem has quit (Quit: Leaving).
06:23:07 <quintopia> matrix*matrix is just a series of matrix*vector ops
06:23:13 <madbr> except maybe the original mmx which was designed for video decoding I think
06:23:56 <madbr> the problem is that matrix*vector is still a whole bunch of fpu operations
06:24:11 <oklopol> quintopia: i don't think that is enough to call it "hardware support" though (ofc i'm not sure we should care so much about what is)
06:24:38 <madbr> it's like 16 multiplies and 12 additions
06:24:42 <quintopia> oklopol: in fact, i find the whole question rather tedious
06:24:52 <Bike> i need hardware support for /sparse/ matrix multiplication. by friday, thanks
06:24:57 <quintopia> indeed, i could give a rats ass how math is applied most of the time
06:25:21 <oklopol> sure, that's a good position
06:25:30 <madbr> a single opcode for doing 16 muls and 12 adds is hard to make worthwhile
06:25:58 <Bike> confession i don't know any sparse matrix algorithms :(
06:26:19 <quintopia> madbr: i dunno seems like it'd be great for pipelining purposes :D
06:26:29 <oklopol> Bike: they are a trivial special case of non-sparse matrix algorithms, by conjugating with sparsification hth
06:26:30 <shachaf> edwardk was telling me nifty things about sparse matrices but i forgot all of them :'(
06:26:37 <shachaf> maybe they had to do with morton ordering
06:26:43 <madbr> quintopia: it's hard to pipe in enough data to keep that many adders and multipliers busy
06:27:00 <Bike> implementation left as an exercise for the reader
06:27:13 <quintopia> oklopol: that sounded very wise you should tell hackego to remember that
06:27:26 <madbr> quintopia: and also the whole thing has like a latency of * plus 4-in-1 summing
06:27:32 <oklopol> let's see i've only seen that done 10000 times
06:27:52 <madbr> quintopia: so it's going to stay a pretty long time in a pipeline which will make the cpu very hard to design
06:27:52 <quintopia> madbr: oops i stopped caring again. sorry. hardware tends to do that to me.
06:28:15 <Bike> how about hardware support for linear operators
06:28:18 <Bike> get my differentiation on
06:28:28 <oklopol> `learn Sparse matrix algorithms are a trivial special case of non-sparse matrix algorithms, by conjugating with the sparsification operation.
06:28:46 <madbr> what's a linear operator
06:29:03 <HackEgo> Sparse matrix algorithms? ¯\(°_o)/¯
06:29:13 <oklopol> except that you define them by their algebraic properties
06:29:14 <shachaf> https://www.fpcomplete.com/user/edwardk/revisiting-matrix-multiplication
06:29:15 <Bike> an operator L is linear iff x * L(y) = L(x*y)
06:29:19 <oklopol> and they just happen to correspond to matrices
06:29:29 <oklopol> where x is a scalar and y is a vector
06:29:37 <madbr> oklopol : the day when ffpu multiply can be done at 1 cycle latency at 4ghz maybe :D
06:29:43 <oklopol> and also you want L(y + y') = L(y) + L(y')
06:29:57 <Bike> but, cauchy monsters...
06:30:27 <madbr> ARM has opcodes for 1 float * vector of 4 floats
06:30:34 <oklopol> quintopia: ? (don't wanna leave you out)
06:30:53 <Bike> oklopol: https://en.wikipedia.org/wiki/Cauchy's_functional_equation#Properties_of_other_solutions
06:30:58 <quintopia> oklopol: i don't know where that wisdom landed
06:31:10 <HackEgo> Sparse matrix algorithms are a trivial special case of non-sparse matrix algorithms, by conjugating with the sparsification operation.
06:32:11 <quintopia> someone should make that learn program split the wisdom name on a be verb
06:32:38 <madbr> oklopol : afaik you're supposed to do it by doing vector multiplications
06:32:48 <madbr> oklopol : and stuff like pairwise addition
06:33:50 <oklopol> to clarify, a linear operator is a matrix multiplication of the form matrix times column vector
06:34:34 <Bike> and the other kind of matrix multiplication is composition.
06:35:41 <oklopol> Bike: those are called additive functions on the wp page
06:36:06 <madbr> oklopol : so stuff like matrix applied on 3d vertexes... SSE was designed for that
06:36:32 <oklopol> the kind of functions considered on that page
06:36:47 <Bike> well, yes, but additive functions include linear functions.
06:36:56 <oklopol> yes, so do general functions
06:37:07 <Bike> i mean22:29 < quintopia> :\
06:37:15 <Bike> i mean, and you defined linear operators as additive functcions.
06:37:37 <oklopol> by _also_, i meant, as a second axiom.
06:37:58 <Bike> oh is it not implied by the scalar bit
06:38:07 <oklopol> it's in the case of reals i guess
06:38:23 <Bike> i've never thought about it.
06:41:00 <oklopol> usually you say L(au + bv) = aL(u) + bL(v)
06:41:36 <Bike> i guess that about covers it
06:44:26 <madbr> looked up intel's asm optim guide
06:44:42 <madbr> you can do 4x4 matrix times 4 element vector in 7 ops
06:48:07 <oklopol> ah yes, on the linear algebra side, that corresponds to Petergren's Lemma
06:48:21 <oklopol> it can be used as an alternative axiomatization
06:48:31 <oklopol> by a basic compactness argument
06:48:38 <oklopol> gotta take a shower though
06:49:07 <madbr> looking up more docs
06:49:23 <madbr> if you try to use more complex opcodes that do more stuff
06:49:41 <madbr> they get divided into simpler microops that take just as much time to execute
06:49:46 <madbr> you're not winning anything :/
06:52:32 <madbr> so that's why you can't have hardware support for matrix multiply
06:53:31 <madbr> the best way to do it is still to use regular multiply and brute force SIMD
06:57:38 <oklopol> well i'm sure you can optimize it, but yeah probably it's not useful
06:58:59 <oklopol> or otherwise the fastest way to do matrix multiplication in hardware is to implement whatever inter ams happened to implement and simulate a program that does it, which to me sounds pretty ridiculous given that that's very restricting
06:59:13 <oklopol> quintopia: in math or in general?
07:00:08 <madbr> GPUs afaik have some hard wired matrix multiplies so it probably has an edge over cpus there
07:00:21 <madbr> but the matrix size is fixed
07:00:32 -!- clog has quit (Ping timeout: 272 seconds).
07:00:37 <madbr> and it's probably in hardware that does nothing else
07:00:46 <Bike> just won a fields medal #yolo #swag
07:01:43 <oklopol> madbr: i'm very much talking about 4x4 matrices, for nxn ones i can imagine that the most sensible hardware solution is to use software.
07:03:02 <madbr> for NxN you'd probably have to cover the thing in multiplexers which means you'd probably end up with no gain over software yes
07:03:10 <oklopol> also i have only 3 publications this year :(
07:03:26 <oklopol> (also some articles in review and one accepted, but still)
07:04:31 <Bike> i have zero HA
07:04:36 <Bike> i win paper golf
07:04:37 <oklopol> moving from conferences to journals apparently means that you get one year of nothing :'(
07:06:18 <Bike> i hear if a mathematician hasn't done their best work by the time they've published negative five papers they never will
07:08:50 <Bike> thankfully, i am not a mathematician, i just hear rumors of their existence
07:14:33 <Bike> that most horrible of monstrosities, an undergrad
07:16:29 -!- peapodamus has quit (Read error: Connection reset by peer).
07:18:03 -!- peapodamus has joined.
07:18:11 <Bike> that wasn't colorful enough, i apologize
07:23:26 -!- FreeFull has quit.
07:28:53 <oklopol> gotta go learn to drive a car
07:29:57 <Bike> don't thit the poeople
07:33:42 -!- clog has joined.
07:35:54 -!- Sellyme has quit (Excess Flood).
07:36:33 -!- Sellyme has joined.
07:40:23 <fizzie> UPS, master of the waffle. "Scheduled Delivery: Thursday" "Scheduled Delivery Updated To: Friday" "Scheduled Delivery Updated To: Thursday" "Scheduled Delivery Updated To: Friday" "Scheduled Delivery Updated To: Thursday"
07:41:38 <fizzie> Not the tasty kind of.
07:41:52 <shachaf> p. sure that's the kind i like
07:42:57 <fizzie> We had some waffles in Belgium (as one does) last summer, and they were somehow p. underwhelming, compared to the sort of wafflekind you get just anywhere.
07:44:52 <fizzie> (Regarding the UPS thing, it seems almost as if they update the delivery estimate to Friday whenever the package has an "Arrival Scan" somewhere, and then revise it to Friday after the following "Departure Scan".)
08:51:39 -!- MindlessDrone has joined.
08:54:17 -!- MindlessDrone has quit (Client Quit).
08:59:34 -!- Sgeo_ has joined.
09:01:55 -!- madbr has quit (Quit: Rouringu de hajikunda!).
09:01:57 -!- Sgeo has quit (Ping timeout: 248 seconds).
09:28:36 -!- MindlessDrone has joined.
09:53:28 -!- stuntaneous has joined.
10:10:51 -!- Frooxius has quit (Read error: Connection reset by peer).
10:11:08 -!- Frooxius has joined.
11:08:14 -!- nooga has joined.
11:20:15 -!- impomatic has quit (Ping timeout: 252 seconds).
11:38:56 -!- atriq has joined.
11:52:09 -!- impomatic has quit (Ping timeout: 252 seconds).
11:58:54 -!- Sorella has joined.
12:29:04 -!- ski has joined.
13:01:32 -!- ais523 has joined.
13:04:14 -!- boily has joined.
13:04:25 -!- metasepia has joined.
13:07:04 -!- Sgeo_ has quit (Read error: Connection reset by peer).
13:13:28 <boily> good last day of the Negative Cow Club morning!
13:31:12 -!- yorick has joined.
13:31:36 -!- atriq has quit (Quit: Leaving).
13:42:46 <fizzie> Good confusing greeting afternoon.
13:56:36 <boily> hah! I defiantly used a semicolon in my Python code!
13:56:55 <boily> (granted, it's in a docstring describing a special case, but still!)
14:09:48 <FireFly> ...I've used semicolons in Python outside of strings
14:09:55 <FireFly> I'm probably a horrible python programmer though
14:10:16 <nooga> noboty cares about python
14:10:56 * boily forcefully mapoles nooga
14:13:51 <FireFly> If you accidentally mapole the wrong person, do you mapologise afterwards?
14:14:37 <boily> only if the mapolee didn't deserve it.
14:15:45 <boily> I don't hold any mapoling urges against you, and ducks taste good, so you don't have to hide.
14:16:20 * FireFly steps away from the duck again
14:26:30 -!- mrhmouse has joined.
14:32:48 <mroman> ah. semicolons are fine in python
14:40:53 -!- conehead has joined.
14:44:23 <mauke> they form a semicolony
14:44:31 <mauke> but what is a lon?
14:45:41 -!- Ngevd has joined.
14:47:54 <fizzie> lo [-n, -ar, -arna]: a lynx.
15:06:41 -!- Ngevd has quit (Remote host closed the connection).
15:06:48 -!- Ngevd has joined.
15:21:05 -!- Ngevd has quit (Remote host closed the connection).
15:21:11 -!- Ngevd has joined.
15:21:15 <boily> Taneb: identity troubles?
15:23:39 -!- Ngevd has quit (Client Quit).
15:24:17 <boily> you are the brother of your brother, but neither of you are Roujo, nor me. you are also a Hamite, consequence of your ex-Hexamite status.
15:24:42 <boily> (unless you still reside there, in which case you are a exexhexhamite.)
15:28:03 <boily> and, I think according to your tanebventions and Yoneda's Lemma, we can entirely define you.
15:44:54 <fizzie> Was this "Malbolge in popular culture" thing already mentioned here? http://moviecode.tumblr.com/post/72635628609/heres-a-fun-one-using-the-obscure-language
15:47:00 -!- Phantom_Hoover has joined.
15:49:08 <boily> `pastelogs moviecode
15:49:52 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.3459
16:08:41 -!- FreeFull has joined.
16:09:15 <fizzie> Seems to have been a different thing.
16:55:02 -!- nooodl has joined.
16:57:30 -!- Bike has quit (Ping timeout: 245 seconds).
17:02:48 -!- peapodamus has quit (Ping timeout: 265 seconds).
17:04:20 -!- Bike has joined.
17:04:36 -!- peapodamus has joined.
17:37:12 -!- Bike has quit (Ping timeout: 245 seconds).
17:47:17 <olsner> http://drj11.wordpress.com/2013/09/01/on-compiling-34-year-old-c-code/#comment-5093 <-- interesting tidbit about structs in ancient C
17:51:55 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
17:52:38 -!- atrapado has joined.
18:03:08 -!- Bike has joined.
18:07:09 -!- FreeFull has quit (Ping timeout: 272 seconds).
18:12:53 -!- peapodamus has quit (Read error: Connection reset by peer).
18:14:59 -!- peapodamus has joined.
18:16:11 -!- nooodl has quit (Read error: Connection reset by peer).
18:16:37 -!- nooodl has joined.
18:32:07 -!- Edwardz has joined.
18:33:50 <HackEgo> Edwardz: 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, try #esoteric on irc.dal.net.)
18:37:55 <mrhmouse> Does anybody have any experience using Husk Scheme? (http://justinethier.github.io/husk-scheme/)
18:38:51 <Edwardz> wish I could say yes. hahaha, but no.
18:40:00 <elliott> nooga: I hear you wanted me
18:51:13 -!- Edwardz has quit (Ping timeout: 272 seconds).
18:58:48 -!- Bike has quit (Ping timeout: 265 seconds).
19:07:02 -!- Bike has joined.
19:09:02 <Taneb> The first NYHC meeting is tomorrow. NYHC stands for "(New) York Haskell Compiler" because there are a lot of Haskell programmers here and we felt like making a Haskell compiler and YHC is a thing that exists and sort of died and "New York Haskell Compiler" is delightfully ambiguously named
19:09:27 <ais523> Taneb: along the lines of "New Glasgow Haskell Compiler"
19:10:32 <ais523> @message oerjan but "does this TM halt within n steps" is decidable, so there's no problem; a primitive recursive function doesn't gain the magic ability to iterate over all n
19:10:33 <lambdabot> Maybe you meant: messages messages-loud messages?
19:10:49 <ais523> @tell oerjan but "does this TM halt within n steps" is decidable, so there's no problem; a primitive recursive function doesn't gain the magic ability to iterate over all n
19:11:08 <Taneb> Unless some people in New Glasgow decide to make a Haskell compiler
19:11:49 <Bike> huh, that's a real place is it
19:11:58 <shachaf> YHC stands for York Haskell Compiler
19:12:09 <shachaf> it's not fair to take someone else's compiler and add "new" to the beginning
19:12:16 <shachaf> well, maybe it's fair if it's dead
19:12:19 <Bike> what if it was a compilation system instead.
19:12:21 <shachaf> http://yhc06.blogspot.com/2011/04/yhc-is-dead.html
19:12:31 <shachaf> Bike: is that like a factorization system
19:12:43 <shachaf> i don't think i understand factorization systems
19:12:46 <shachaf> https://en.wikipedia.org/wiki/Factorization_systema
19:13:26 <Bike> what's the typo
19:14:14 <ais523> @tell oerjan or to put it another way, you can compare "Does this TM halt within n steps" and "const False" given any specific value for n, and if you don't have a value for n, you don't have a primitive recursive program, you have a function from integers to primitive recursive programs
19:18:26 <ais523> <Slashdot> <timothy> <ananyo> "At present, an ampere is defined as the amount of charge flowing per second through two infinitely long wires one meter apart, such that the wires attract each other with a force of 2×10^-7 newtons per meter of length."
19:18:57 -!- FreeFull has joined.
19:19:30 <Bike> yeah, they've wanted to change that for a while
19:19:33 -!- prooftechnique has joined.
19:20:09 <quintopia> really? it's not currently defined in terms of electron flux?
19:21:04 <ais523> apparently they're considering defining it in terms of the charge of an electron
19:22:13 <Bike> it's intuitively weird to me that amperes are fundamental rather than coulombs
19:22:20 <Bike> amperes are more useful, i guess
19:23:47 <ais523> btw, I was thinking about noit o'mnain worb
19:24:18 <ais523> I think it has an entropy issue that prevents you creating amplifiers, but haven't proved it
19:24:24 <quintopia> oh? i think that language could use a few more features. :D
19:24:28 <ais523> also the wire-crossing problem there is really obnoxious, but might be solvable
19:24:42 <ais523> what I'd like would be a wire-cross, and an amplifier
19:24:50 <ais523> together with an infinitely repeating program for infinite memory
19:25:09 <quintopia> the ability to cause particles to stick to each other, or attract each other, or repel each other. that'd be cool.
19:25:13 <ais523> the amplifier has three ports, it'll convert a bobule at one specific port to one at the other two ports, or vice versa
19:25:37 <ais523> (this works fine with both p-type and n-type communication)
19:25:43 -!- FreeFull has quit (Ping timeout: 252 seconds).
19:29:09 <mauke> php -r 'var_dump(in_array(2000, array(md5("dbox"))));'
19:29:25 <boily> ais523: an amplifier, as in that triangular thingie you draw in circuitous diagrams, and that produces smoke when used in real life?
19:29:42 <ais523> boily: pretty similar to that, yeah
19:29:47 <ais523> except they don't always produce smoke
19:29:58 <ais523> just usually, especially when exposed to schoolchildren or undergraduates
19:30:44 * boily has fond memories of obliterated electronic components :D
19:32:06 <quintopia> ais523: also the division program doesn't work like it should.
19:33:19 <ais523> anyway, I really want that language to work
19:33:21 <ais523> I just don't think it does
19:34:28 <quintopia> perhaps if you added the ability for a bobule to duplicate in certain situations, like wireworld?
19:35:06 <ais523> quintopia: that's why I suggested amplifiers
19:35:29 <quintopia> ais523: oh you were suggesting that as a new language feature
19:35:44 <quintopia> ais523: i thought you were looking to build something that did that
19:36:27 <ais523> I'd like to, but suspect it's impossible
19:36:30 <ais523> (haven't proven it though)
19:36:33 <boily> couldn't you just emulate wireworld in nomw?
19:37:00 <quintopia> seems straightforward enough to add the ability to connect sinks to sources
19:37:24 <quintopia> "any time this sink consumes a bobule, those sources produce one"
19:37:38 <quintopia> perhaps by just giving them the same name
19:38:19 <quintopia> that would accomplish both amplification and wire crossing and other things
19:38:50 <ais523> yeah, and you can still get original-style sinks and sources
19:38:56 <ais523> also it accomplishes diodes too
19:39:07 <ais523> or, well, they don't act anything like electronic diodes
19:39:13 <ais523> but those one-directional things
19:40:50 <quintopia> now i kind of want to implement that
19:42:46 <ais523> so the spec, basically is "you can create arbitrary connections between sources and sinks; every step, a connected group of sinks can (and will, if possible) destroy a bobule adjacent to each sink, creating a bobule in a blank space adjacent to every connected source"
19:43:04 -!- prooftechnique has quit.
19:44:56 <quintopia> sounds close to what i was thinking, though i might have required that every sink actually have a bobule on top of it before they fire. so that it functions something like a synapse.
19:45:18 <quintopia> maybe even allow the sinks to lock a bobule in place
19:46:14 <ais523> quintopia: I meant to require every sink to have a bobule, every source to have a free adjacent space
19:46:26 <ais523> the "on top" variation is probably sipler, though
19:49:03 <quintopia> ais523: also in that variant, perhaps sources could act as walls, so that they are always free to produce a bobule on top of themselves?
19:49:30 <ais523> quintopia: yeah, sinks lock bobules, sources lock holes
19:49:46 <ais523> I think it's important for the elegance of the language that bobules and holes are entirely symmetrical in every way
19:49:56 -!- monotone has quit (Quit: "I can hear myself... I think I'm a bit afraid.").
19:49:57 <ais523> (and a solid wall blocks both bobules and hole from moving)
19:50:23 -!- successus has joined.
19:50:30 <ais523> my 's' key is unreliable, in case you haven't noticed already
19:50:36 <HackEgo> successus: 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, try #esoteric on irc.dal.net.)
19:51:20 <quintopia> ais523: ah yes, then it's time reversible, in a sense, since a step is just a bobule switching with a hole. there is a correlate here of CPT-invariance.
19:51:24 -!- monotone has joined.
19:51:29 -!- Bike has quit (Ping timeout: 265 seconds).
19:51:50 <ais523> quintopia: except that reversing it acts identically to not reversing it
19:52:04 <ais523> reversing it swaps sources and sinks
19:52:16 <ais523> and that's why you can get any "voltage" into the circuit at all
19:52:24 <ais523> I'm not sure if the electrics analogy is a good one or not
19:52:31 <ais523> it acts similarly, but the details area all different
19:52:42 <ais523> perhaps we have to use a phrase like "bobule density gradient"
19:52:51 <quintopia> ais523: i was thinking that the relevant symmetry is swapping the direction of time and swapping bobules with holes. holes moving backwards in time is the same as bobules moving forwards
19:52:56 <ais523> anyway, I came to one really neat realization
19:53:02 <ais523> AFAICT, there's no way to create an inverter
19:53:21 <ais523> have two wires for everything, which always have bobule densities of opposite polarities
19:53:27 <ais523> you can do that because of the symmetry of the system
19:53:43 <ais523> then an inverter is simply swapping the two wires
19:54:04 <ais523> and I've already worked out how to construct AND and OR out of amplifiers and one-way walls
19:54:21 <quintopia> ais523: gotta work now. maybe i'll implement that thing this weekend?
19:55:20 <ais523> the hard part is coming up with a reasonable syntax
19:55:39 <ais523> allow for infinite repeating programs, if you can
19:55:55 <ais523> (might be difficult due to infinite repeating patterns not interacting well with nondeterminism)
19:56:02 -!- successus has quit (Quit: Saliendo).
19:56:07 <ais523> (although you could still get away with it via exploiting the speed of light)
19:59:25 -!- quintopia has quit (Remote host closed the connection).
19:59:35 -!- Bike has joined.
20:17:04 <Bike> http://moviecode.tumblr.com/post/73179804685/ please don't do this
20:18:26 <boily> historical accuracy is hard. let's go ride hoverbikes in the renaissance!
20:19:56 <Bike> by this i mean being so pedantic it's hard to believe you can eat you're so busy "criticizing", and also i mean using "epic fail" in any context
20:20:25 <Bike> that wasn't a very good thingie
20:20:35 <ais523> that entire miniblog is about being excessively pedantic, though, because that's what the readers want
20:21:00 <ais523> also, a parody of that sort of thing would be great
20:21:03 <boily> I must criticize you here. I am not pedantic, just arrogant.
20:21:33 <ais523> like, a film set in 1990 doing "for (int i; i < 0; i++)", and getting a compiler message "error: initial declarations in for-loops have not been invented yet"
20:21:37 <Bike> by 'pedantic' i mean this weird-ass criticism for using a book five years too new, not just hunting down sources, which is maybe obsessive
20:21:38 <ais523> although maybe C++ accepted them back then?
20:22:06 <Bike> didn't some C implementations accept it?
20:22:28 <ais523> if they did, it's probably because implementing C++ makes that feature easy to implement too
20:23:45 <Vorpal> fizzie, I now have 3x 10% off Toki Tori 2 on steam...
20:25:22 <fizzie> Sadly, you probably can't combine them for a 30% off.
20:26:12 <fizzie> Rumour has it my ISP is in the progress of pushing out a firmware upgrade that disables entirely the telnet admin interface.
20:26:33 <Bike> 27.1% off, surely
20:27:34 <fizzie> On one hand, it's nice, because you can log in as "root:public" or "ztedebug:ztedebug" for root shell access, making any locally set "passwords" and such a joke; on the other, it's less nice, because the webterface is quite sucky, and anyway sometimes you might just want a shell.
20:27:55 <fizzie> (The joys of ISP-manageable hardware.)
20:28:52 -!- rodgort has quit (Ping timeout: 245 seconds).
20:32:40 <ais523> fizzie: what is this an interface to? your router?
20:34:08 <fizzie> ais523: I guess you could technically call it a router, since it has some features of that kind, but I'm using it as a plain VDSL2 modem.
20:35:26 <fizzie> (One of its Ethernet ports is configured in "bridged" mode, and that's the only thing that's connected to anything.)
20:36:16 <ais523> I take it the admin interface isn't accessible to the public Internet?
20:36:40 <ais523> at least, I hope it isn't
20:37:06 <fizzie> As far as I know, it's not. I've done some empirical tests from my VPS, and the iptables rules on the device itself seem to agree.
20:37:44 <fizzie> Still, there's all that talk about malware that gets to your computer via vector X, then takes over your router and sets up some fake DNS and so on.
20:37:52 <fizzie> Perhaps they're worried about that sort of stuff.
20:38:44 <fizzie> (I'd be more worried about the telnet interface if I were trying to use the thing to run a web cafe or something.)
20:39:04 <ais523> the only benefit to malware for hiding on a router is that it could persist after wiping the computer, and might be a little hard to discover if the antivirus doesn't know about it
20:39:15 <ais523> otherwise it may as well just compromise the machine it's running on, that's easier
20:40:22 <fizzie> I suppose it's easier to spread to other machines in the local network when you can rewrite DNS for google.com. (Or do other stuff w.r.t. yourbank.tld.)
20:42:09 -!- gdjfgh has joined.
20:45:13 <fizzie> Huh. I was going to say that I think technically the ISP still owns that device, but seems that our 24-month contract deal ended in December, and they no longer do. Perhaps I should go and disable the remote control functionality.
20:49:08 -!- glogbackup has quit (Remote host closed the connection).
20:49:42 <HackEgo> olist 938: shachaf oerjan Sgeo FireFly
20:52:47 -!- rodgort has joined.
20:53:15 <boily> `run cat bin/olist
20:53:17 <HackEgo> echo -n "$(basename "$0")${@:+ }$@: "; tail -n+2 "$0" | xargs; exit \ shachaf \ oerjan \ Sgeo \ FireFly
20:53:49 <boily> uhm, could someone please add me to the olist? past experience with quoting didn't end well...
20:54:05 <shachaf> boily: don't cat olist, it pings everyone :'(
20:54:11 <shachaf> `run echo boily >> bin/olist
20:55:42 <boily> shachaf: sorry >_>'...
20:55:57 <FireFly> shachaf: clearly he should've s///'d everyones' nicks out
20:56:30 * boily hides under a cardboard box with みかん written on it
20:56:31 <HackEgo> rpub -a "$(onfranzr "$0")${@:+ }$@: "; gnvy -a+2 "$0" | knetf; rkvg \ funpuns \ brewna \ Ftrb \ SverSyl \ obvyl
20:57:11 <HackEgo> [U+307F HIRAGANA LETTER MI]
20:57:43 <fizzie> "brewna" is very catchy name.
20:59:00 <boily> `run echo "Mon aéroglisseur est plein d'anguilles" | r13
20:59:01 <HackEgo> Zba néebtyvffrhe rfg cyrva q'nathvyyrf
20:59:15 <boily> bleh. should've outputted ŕ.
20:59:46 <fizzie> Other people have so rot13able names. :/
21:00:40 -!- tromp has joined.
21:01:00 -!- Slereah_ has joined.
21:01:30 <FireFly> that's byfare the most rot13able name
21:01:40 -!- Bike has quit (Ping timeout: 245 seconds).
21:02:00 <ion> What fun puns
21:02:22 <FireFly> > map sort ["boily", "fizzie", "shachaf", "olsner"]
21:02:23 <lambdabot> ["biloy","efiizz","aacfhhs","elnors"]
21:02:45 <boily> olsner is El Norseman!
21:02:54 <ion> `r13elcome
21:02:55 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: r13elcome: not found
21:03:21 <fizzie> I guess "efiizz" could be some sort of a title. Like "effendi".
21:03:28 <HackEgo> 802) <Gregor> !rot13 Fluttershy Rainbow Dash Rarity Applejack Twilight Sparkle Pinkie Pie <EgoBot> Syhggreful Envaobj Qnfu Enevgl Nccyrwnpx Gjvyvtug Fcnexyr Cvaxvr Cvr <olsner> oh, they're all named after rot13'd welsh words
21:03:29 -!- Slereah has quit (Ping timeout: 265 seconds).
21:03:54 <FireFly> `run echo -e '#!/bin/sh\nrelcome | r13' >bin/r13elcome; chmod +x bin/r13elcome
21:04:05 <HackEgo> Jrypbzr gb gur vagreangvbany uho sbe rfbgrevp cebtenzzvat ynathntr qrfvta naq qrcyblzrag! Sbe zber vasbezngvba, purpx bhg bhe jvxv: <uggc://rfbynatf.bet/jvxv/Znva_Cntr>. (Sbe gur bgure xvaq bs rfbgrevpn, gel #rfbgrevp ba vep.qny.arg.)
21:04:05 <olsner> (from last time we did this rot13 everything thing)
21:04:20 <HackEgo> Jrypbzr gb gur vagreangvbany uho sbe rfbgrevp cebtenzzvat ynathntr qrfvta naq qrcyblzrag! Sbe zber vasbezngvba, purpx bhg bhe jvxv: <uggc://rfbynatf.bet/jvxv/Znva_Cntr>. (Sbe gur bgure xvaq bs rfbgrevpn, gel #rfbgrevp ba vep.qny.arg.)
21:04:42 <nooodl> i guess you need relcome $@ or what is it
21:05:04 <nooodl> `run sed -i 's/relcome/relcome $@/' bin/r13elcome
21:05:05 <FireFly> `run sed -i s/relcome/& "\$@"/ bin/r13elcome # will this work?
21:05:09 <HackEgo> bash: $@/: No such file or directory \ sed: -e expression #1, char 10: unterminated `s' command
21:05:25 <FireFly> um i think your maybe worked
21:05:27 <HackEgo> abbbqy: Jrypbzr gb gur vagreangvbany uho sbe rfbgrevp cebtenzzvat ynathntr qrfvta naq qrcyblzrag! Sbe zber vasbezngvba, purpx bhg bhe jvxv: <uggc://rfbynatf.bet/jvxv/Znva_Cntr>. (Sbe gur bgure xvaq bs rfbgrevpn, gel #rfbgrevp ba vep.qny.arg.)
21:06:09 <olsner> FireFly: insufficient quoting of & and "
21:06:22 <FireFly> I should've put it all in single-quotes
21:06:38 <olsner> (and the space between & and ")
21:06:48 -!- tromp has quit (Ping timeout: 276 seconds).
21:06:48 -!- atrapado has quit (Quit: Leaving).
21:13:27 -!- conehead has quit (Quit: Computer has gone to sleep.).
21:17:58 <ais523> is `r13elcome rainbow in addition to being rot13?
21:18:36 <ais523> I filter out colors in this client, and `relcome does not really give me much reason to stop that
21:18:47 -!- Bike has joined.
21:20:05 -!- MindlessDrone has quit (Quit: MindlessDrone).
21:20:38 <fizzie> ais523: It's colorful, yes.
21:20:50 -!- ais523 has quit.
21:32:35 -!- conehead has joined.
21:36:24 -!- atrapado has joined.
21:39:05 -!- monotone has quit (Quit: "I can hear myself... I think I'm a bit afraid.").
21:44:17 -!- monotone has joined.
21:45:37 -!- glogbackup has joined.
21:45:37 -!- glogbackup has quit (Remote host closed the connection).
21:47:06 -!- Bike has quit (Ping timeout: 276 seconds).
21:48:28 -!- Bike has joined.
21:49:51 -!- monotone has quit (Quit: "I can hear myself... I think I'm a bit afraid.").
21:51:35 -!- boily has quit (Quit: HEARD CHICKEN).
21:51:40 -!- metasepia has quit (Remote host closed the connection).
21:54:38 -!- monotone has joined.
22:00:23 -!- glogbackup has joined.
22:01:38 -!- glogbackup has quit (Remote host closed the connection).
22:23:07 <mrhmouse> I can't actually read `relcome anyways - bright colors on a bright background and all. I just assume it's a friendly message.
22:25:09 -!- oerjan has joined.
22:25:22 <monotone> mrhmouse: It actually insults you, but only if you can't read the message.
22:26:16 <lambdabot> ais523 said 3h 15m 26s ago: but "does this TM halt within n steps" is decidable, so there's no problem; a primitive recursive function doesn't gain the magic ability to iterate over all n
22:26:16 <lambdabot> ais523 said 3h 12m 1s ago: or to put it another way, you can compare "Does this TM halt within n steps" and "const False" given any specific value for n, and if you don't have a value for n, you don't have a primitive recursive program, you have a function from integers to primitive recursive programs
22:28:40 <oerjan> @tell ais523 No, the function from n to "Does this TM halt within n steps" _is_ primitive recursive. almost trivially so. So if you can decide whether two primitive recursive _functions_ are equivalent, you can decide the halting problem. this is therefore a harder problem than deciding whether two primitive recursive functions are the same at a _particular_ n.
22:36:58 <oerjan> @tell ais523 or put differently, n is the _input_ to your primitive recursive program, not a part of it.
22:38:59 -!- nchambers has joined.
22:46:41 -!- FreeFull has joined.
22:49:01 <oerjan> `run mv wisdom/sparse{, matrix algorithm}
22:49:03 <HackEgo> mv: target `algorithm}' is not a directory
22:49:18 <oerjan> `run mv wisdom/sparse{," matrix algorithm"}
22:49:33 <oerjan> `? sparse matrix algorithms
22:49:36 <HackEgo> Sparse matrix algorithms are a trivial special case of non-sparse matrix algorithms, by conjugating with the sparsification operation.
22:49:58 <oerjan> oklopol: DO I HAVE TO CLEAN UP AFTER EVERYONE
22:50:26 <HackEgo> #!/bin/sh \ relcome $@ | r13
22:51:06 -!- nchambers has changed nick to dtscode.
22:51:25 <oerjan> `run sed -i '2crelcome "$@" | r13' bin/r13elcome
22:51:42 <HackEgo> qgfpbqr: Jrypbzr gb gur vagreangvbany uho sbe rfbgrevp cebtenzzvat ynathntr qrfvta naq qrcyblzrag! Sbe zber vasbezngvba, purpx bhg bhe jvxv: <uggc://rfbynatf.bet/jvxv/Znva_Cntr>. (Sbe gur bgure xvaq bs rfbgrevpn, gel #rfbgrevp ba vep.qny.arg.)
22:51:58 <dtscode> do you do irp programming here?
22:52:07 <oerjan> nooga: I REPEAT, DO I HAVE TO CLEAN UP AFTER EVERYONE
22:52:13 <oerjan> dtscode: we split off #irp for that.
22:54:33 -!- mrhmouse has quit (Quit: Leaving.).
23:01:50 -!- nooodl has quit (Quit: Ik ga weg).
23:01:52 -!- dtscode has quit (Quit: Leaving).
23:06:37 <fizzie> fungot: Is HackEgo a bot?
23:06:37 <fungot> fizzie: there is so much simpler. it's easier to do in c" sounds like an interesting project, but i
23:06:41 <fizzie> (Takes one to know one.)
23:07:33 -!- doesthiswork has joined.
23:08:16 -!- FreeFull has quit (Ping timeout: 272 seconds).
23:09:46 -!- FreeFull has joined.
23:10:57 -!- ^v has joined.
23:35:29 <oerjan> <fizzie> We had some waffles in Belgium (as one does) last summer, and they were somehow p. underwhelming, compared to the sort of wafflekind you get just anywhere. <-- wait, i thought belgian waffles were the overwhelming kind. maybe norwegian ones are just even worse.
23:39:03 <oerjan> <fizzie> lo [-n, -ar, -arna]: a lynx. <-- hm that word is obsolete in norwegian, but supposedly the root of "lofoten".
23:39:30 <oerjan> (we call the animal "gaupe" these days.)
23:40:26 <oerjan> (also no:lo still has the meaning of en:lint)
23:40:50 <Bike> the "temp" web drive in my department has 928 GB free out of 5.85 TB. what the fuck
23:42:52 <Bike> it seems to be mostly in the folder "AC/3/Small Animal"
23:43:08 <oerjan> <fizzie> Was this "Malbolge in popular culture" thing already mentioned here? http://moviecode.tumblr.com/post/72635628609/heres-a-fun-one-using-the-obscure-language <-- at least twice, also i edited some of the wikipedia reference
23:43:50 <Bike> i have no idea
23:44:00 <oerjan> hm you _are_ a biology department, so the folder name isn't _that_ weird.
23:44:08 <Bike> it seems to be mostly books
23:44:27 <Bike> black's vet dictionary and shit like that
23:44:55 <Bike> anyone want a copy of "A Colour Handbook of Skin diseases of the dog and cat"?
23:45:14 <oerjan> no. absolutely nobody wants that.
23:45:26 <oerjan> (i kid, there's probably _some_ crazy person who does)
23:46:03 -!- impomatic has quit (Quit: http://corewar.co.uk).
23:49:20 <oerjan> fizzie: that is, the "elementary" reference. the moviecode link itself i don't recall.
23:50:55 * oerjan suspects fizzie of idling
23:52:05 <olsner> IDLING!? such heinous acts
23:52:19 <oerjan> clearly a bannable offense.
23:53:50 <olsner> Bike: probably someone hid 4TB of porn in some deeply nested subfolder
23:53:51 <oerjan> idling is bannable as unreasonably indistinguishable from "intent to ping out, never reading what you said".
23:53:58 <olsner> ... or maybe it's just sciency stuff
23:54:21 <Bike> a lot of the pdfs look like terrible scans, which means one book is half a gig
23:54:27 <Bike> ...well, that's still a thousand books
23:55:25 <oerjan> how long until you can fit scans of all the world's books on a disk
23:56:22 <Bike> olsner: i found Dentistry/AdultK9Roots *waggles eyebrows*
23:56:50 <oerjan> shouldn't you be wagging your tail instead