←2015-11-08 2015-11-09 2015-11-10→ ↑2015 ↑all
16:11:32 -!- esowiki has joined.
16:11:33 -!- glogbot has joined.
16:11:34 -!- EgoBot has joined.
16:11:36 -!- esowiki has joined.
16:11:37 -!- esowiki has joined.
16:13:05 <oerjan> on the bright side, HackEgo has been very responsible
16:13:29 <Gregor> I'm honestly not sure why glogbot didn't pop back up on its own... the system was up...
16:16:01 <oerjan> i was getting into this strange ritual of copying tunes logs into vim to reformat them
16:16:57 <Gregor> I'm sorry it's all my fault :'(
16:17:09 <oerjan> it's all forgiven
16:17:14 <izabera> https://www.codeeval.com/open_challenges/185/ i'm having problems with the repeated parts
16:19:26 -!- MoALTz has joined.
16:20:21 <izabera> any idea? :\
16:20:38 <oerjan> well it says in point 1 that you have _all_ shifts
16:21:15 -!- AnotherTest has quit (Ping timeout: 250 seconds).
16:21:21 <oerjan> that probably is enough to make it solvable
16:23:58 <oerjan> as long as the n-1 first letters and the n-1 last letters aren't identical, then it's obviously not
16:24:22 <oerjan> (because then you get a loop that can start anywhere)
16:25:24 <oerjan> otherwise, here's how to find the first n-1 letters: it's the unique combination which occurs once more as the first n-1 chars in a piece than it does as the last n-1 chars.
16:25:42 <oerjan> similarly for last n-1.
16:25:54 <izabera> yes that's all good
16:26:00 <izabera> but how do you deal with the repeated parts?
16:26:36 <oerjan> hm
16:27:03 -!- AnotherTest has joined.
16:27:57 <oerjan> if there are any that are repeated 3 or more times, you cannot.
16:28:17 <oerjan> unless the part in between is also identical, i guess.
16:28:54 <int-e> but do you have to... I mean they promise you that the solution will be unique
16:28:57 <oerjan> e.g. 123abc456abc789abc0 must become ambiguous
16:29:13 <oerjan> (with length 3)
16:30:07 <oerjan> actually isn't this basically a hamiltonian path problem.
16:30:33 <int-e> yes.
16:30:41 <oerjan> each piece is a node, each possible next piece gives an edge
16:31:10 <int-e> but I guess we can hope for a very sparse parse with lots of nodes having degree 2
16:31:11 <oerjan> or wait
16:31:19 <int-e> ...sparse graph...
16:31:20 <oerjan> each piece is an _edge_
16:31:46 <int-e> well, hmm, it's a directed hamiltonian path problem
16:31:50 <oerjan> ...and it suddenly becomes a eulerian path problem
16:31:57 <oerjan> oh wait, directed
16:31:57 <int-e> (at least that's how I would view it)
16:32:46 <oerjan> directed eulerian path sounds easier than hamiltonian, anyway
16:33:44 <fizzie> What are the nodes in your graph where each piece is an edge?
16:33:44 <oerjan> oh and the graph can have multiedges
16:34:13 <oerjan> fizzie: n-1 length substrings
16:34:31 <int-e> fizzie: in my graph, the pieces are the nodes, and there's an edge between pieces that can be successive shifts.
16:34:41 <fizzie> int-e: Yes, that was easier.
16:34:46 -!- sacadouche has quit (Quit: Leaving...).
16:35:05 <fizzie> But I guess both make sense.
16:35:25 <int-e> (oerjan wrote that differently above... hmm)
16:35:42 <int-e> but it's an obvious encoding anyway
16:35:45 <oerjan> i'm sure i meant the same thing
16:36:09 <oerjan> well my point is that eulerian path problem is polynomial, while hamiltonian is NP-complete
16:36:15 <int-e> oerjan: I meant you phrased it differently, not that it was different.
16:36:32 <oerjan> but that's undirected, not sure with directed
16:36:45 <int-e> I don't see the eularian path
16:37:19 <oerjan> hm? when the pieces are edges, and you visit each piece once
16:37:43 <int-e> oh.
16:38:21 * int-e missed a step there
16:39:59 <oerjan> many nodes with only two edges does help, you can collapse them
16:42:39 <oerjan> and then only the repeating n-1 strings will be left
16:42:49 <int-e> okay, this looks plausible: there is a directed eularian circuit if a graph is strongly connected, and the indegree and outdegree of each node are equal. (without the strong connectedness, you can decompose it into circles, and by the connectedness you can glue them together into a single one)
16:43:12 <int-e> (eularian paths follow)
16:44:02 <oerjan> int-e: um we cannot have circuits in this problem
16:44:13 <oerjan> they obviously give non-unique solutions.
16:44:25 <int-e> oerjan: (eularian paths follow) ... we identify the two odd nodes and connect them by an artificial edge
16:44:26 <oerjan> although - we could add an edge from the beginning to the end
16:44:40 <oerjan> are we doing that thing again.
16:44:45 <int-e> no
16:45:05 <oerjan> er end to beginning
16:45:18 <oerjan> which is that artificial edge
16:45:31 <int-e> it's just that the eularian circuit case is nicer to characterize and reason about
16:45:40 <oerjan> quite so
16:46:00 <oerjan> now we need a uniqueness requirement.
16:46:40 <int-e> constraints: "There is only one correct answer for each test case."
16:46:43 <oerjan> also his name was Euler not Eular hth
16:47:08 <izabera> that was his last name
16:47:15 <int-e> I should know.
16:48:01 <int-e> I guess the "rian" is just too difficult to get right
16:48:35 * int-e is rationalizing... at least that part of the brain still works :P
16:48:46 <oerjan> ok the step i'm unsure of is how to decompose into circles
16:48:54 <int-e> (and that's good; it's the most important part for staying sane)
16:48:55 <HackEgo> [wiki] [[Pointer-ng]] http://esolangs.org/w/index.php?diff=45226&oldid=45188 * 93.198.137.24 * (+35) Updated download to official interpreter
16:49:23 <int-e> oerjan: you just pick an edge and follow a path until you end up at the starting point... remove that circle, repeat
16:49:58 * izabera was doing that
16:50:01 <oerjan> hm you're right, because you still have ingoing = outgoing
16:50:42 <int-e> In fact, amazingly, it's even easier than the usual parity argument for undirected graphs.
16:50:50 <oerjan> heh
16:51:25 <oerjan> ok so then it might be the gluing together...
16:51:43 <oerjan> that, too, should be simple
16:51:44 <int-e> (trivial exercise: prove that in any eulerian graph, one can orient the edges such that indegree = outdegree for all vertices)
16:52:01 -!- mauris has joined.
16:52:24 <oerjan> izabera: ok so once you have decomposed it into circles, just join them in an arbitrary way
16:52:37 <oerjan> and you'll end up with a large one
16:52:59 <oerjan> and by the problem statement, it must be unique once you identify the starting/ending
16:53:08 -!- zadock has joined.
16:53:23 <oerjan> well unique in what string it gives, anyway
16:56:12 -!- mauris has quit (Ping timeout: 244 seconds).
16:57:16 <oerjan> step 1: make a dictionary of all n-1 size subpieces and their in/out pieces.
16:57:40 <oerjan> find the two unique ones that don't have the same number of in/out.
16:57:56 <oerjan> (that was step 2)
16:58:22 <oerjan> step 3: find a path, just by following edges, from the initial to the final subpiece.
16:58:37 <oerjan> (adjust the dictionary to remove the edges used)
16:58:45 -!- J_Arcane has quit (Ping timeout: 246 seconds).
16:59:22 <izabera> another problem
16:59:29 <oerjan> step 4: while there are edges remaining in the dictionary, pick a subpiece in the dictionary that is also in the string constructed so far.
16:59:35 <izabera> how do i identify the edge?
16:59:42 <oerjan> expand the string by a new path from that subpiece to itself.
16:59:57 <izabera> let me explain
16:59:58 <oerjan> izabera: an edge is the same as a piece, here
17:00:04 <izabera> no i mean
17:00:07 <izabera> yes i know that
17:00:19 <oerjan> oh and they must be counted with multiplicity.
17:00:20 <izabera> i mean the start/end of the cicle
17:00:32 <izabera> edge in that sense
17:00:42 <oerjan> izabera: i told you how to identify the start/end n-1 subpieces
17:00:55 <izabera> yes but
17:00:59 <izabera> in that example
17:01:01 <oerjan> the edges will just fall into place when you complete the algorithm, i think
17:01:25 <izabera> "The evil plan to hack CodeEval ranking." this is the first one i remove
17:01:57 <izabera> after that, i'm left with "il-ev" "l-evi" "evil-" "-evil" "vil-e"
17:02:09 <izabera> so i thought i could start from il-ev
17:02:35 <oerjan> izabera: no, you must choose a subpiece that is in the string so far. only vil fits, i think.
17:02:35 <izabera> and if i try to build a new cycle from that, i can join all those pieces to il-evil-ev
17:02:48 <izabera> i see
17:02:50 <izabera> thanks
17:02:54 <oerjan> er, evil
17:08:55 -!- mauris has joined.
17:09:06 -!- AnotherTest has quit (Ping timeout: 260 seconds).
17:24:27 <HackEgo> [wiki] [[RegexPL]] http://esolangs.org/w/index.php?diff=45227&oldid=45223 * Oerjan * (-13) Fix template call (only named parameters may contain =)
17:25:03 -!- AnotherTest has joined.
17:30:49 <izabera> can i end up with a piece that doesn't fit in the original cycle?
17:31:12 <HackEgo> [wiki] [[Template:Wayback]] http://esolangs.org/w/index.php?diff=45228&oldid=30812 * Oerjan * (+188) Explain how to use =
17:31:16 <izabera> i mean i have the first cycle, then i find a new one, will it extend the first one?
17:32:28 <izabera> no
17:32:49 <izabera> but i can always find a cycle that extends the first one
17:32:53 <izabera> this sounds correct
17:33:56 <oerjan> yeah because the graph is connected
17:34:24 -!- J_Arcane has joined.
17:39:07 <HackEgo> [wiki] [[Pointer-ng]] http://esolangs.org/w/index.php?diff=45229&oldid=45226 * 93.198.137.24 * (+209) Added one more example
17:39:26 <HackEgo> [wiki] [[Pointer-ng]] http://esolangs.org/w/index.php?diff=45230&oldid=45229 * 93.198.137.24 * (+6) /* Example Code */
17:41:46 -!- gniourf has quit (Quit: Leaving).
17:42:04 <HackEgo> [wiki] [[Special:Log/newusers]] create * Weloxux * New user account
17:44:13 -!- gniourf has joined.
17:49:09 -!- Welo has joined.
17:53:25 -!- Guest78999 has joined.
17:53:30 <Guest78999> Hey.
17:53:38 <Guest78999> Name's Alek.
17:53:52 <Guest78999> If ANY OF YOU ARE ALIVE, Please Tell Me.
17:53:57 <izabera> i'm mostly alive
17:54:02 <Guest78999> Good.
17:54:39 <Taneb> I'm not
17:55:04 <scoofy> i'm a zombie
17:55:06 <scoofy> does that count?
17:55:29 <izabera> i guess it depends on which kind of spell was used to summon you
17:55:34 -!- drdanmaku has quit (Quit: .).
17:55:36 <Guest78999> Okay...
17:55:54 <Guest78999> Still
17:56:15 <izabera> fungot: u alive?
17:56:16 <fungot> izabera: " sometimes known as crew". really? i've never seen a such bad situation in his screenplay but also included a scene from the " s", they were lower in social position than the arabs and berbers, and were fnord it around! odd! fnord 15:09, 12 aug 2004 ( utc)
17:56:57 <Guest78999> IRC SPAM?!
17:57:08 <int-e> unbelievable
17:57:25 <Guest78999> That. Was Too Long.
17:57:48 <izabera> i want to dye my hair pink
17:57:58 <Guest78999> ...
17:57:58 <izabera> not pink like the girl from lazy town
17:57:59 <int-e> izabera: sounds good
17:58:03 <Guest78999> Oh Well.
17:58:04 <Guest78999> Bye.
17:58:05 <izabera> more like pastel pink
17:58:06 -!- Guest78999 has left.
17:58:35 <int-e> fungot: please explain Guest78999
17:58:35 <fungot> int-e: ' ' the new york times mainly uses kiev. the times style guide http://www.timesonline.co.uk/ tol/ fnord/ fnord fnord
17:58:50 <int-e> thanks, that made more sense than the actual person
17:58:54 <izabera> :D
17:59:17 <izabera> does it always quote fnord?
17:59:24 <izabera> is that a fnord mode or something?
17:59:48 <int-e> if I remember correctly, "fnord" is a placeholder for rare words that didn't make it into the language model's dictionary
18:00:19 <int-e> (not sure how fizzie deals with actual occurrences of "fnord")
18:01:36 <izabera> http://arin.ga/L249tM/raw ok i submitted this and it scores a whopping 5%
18:01:44 <izabera> guess i'm doing something wrong
18:01:48 <izabera> not sure what <.<
18:02:32 <izabera> it solves their example and all the ones i came up with >.>
18:03:35 <oerjan> with 20 tests, 5% is just what you'd expect if it solves _only_ that example, right :P
18:04:01 <izabera> can you show an example that fails?
18:04:02 <int-e> what is the input format there...
18:04:11 <izabera> it's |piece|piece|piece|
18:04:57 <int-e> |a|a| doesn't work, for example
18:05:44 <izabera> i don't think that's valid input <.<
18:06:14 <izabera> ok it is but still
18:06:33 <oerjan> that's pretty degenerate
18:06:35 <int-e> hmm, "The minimum length of a piece of text is 8"
18:06:35 <int-e> sigh
18:06:40 <izabera> \o/
18:06:53 <oerjan> izabera: are you really making a cycle initially?
18:06:57 <izabera> yes?
18:07:03 <izabera> i hope so?
18:07:14 <oerjan> how do you deal with the initial and final pieces
18:07:53 <izabera> i start with the first piece in that list, try to extend it as much as i can, and call what i get a cycle
18:08:35 <izabera> initial and final pieces should be part of that first cycle
18:08:50 <oerjan> i don't see how you can get the initial piece
18:09:00 <izabera> why not?
18:09:13 <oerjan> or, hm, are you going in both directions
18:09:16 <izabera> yes
18:09:23 <izabera> in the first cycle i am
18:09:26 <oerjan> ah.
18:09:36 <oerjan> well, it's a "path", not a cycle, anyway :P
18:09:46 <izabera> ok
18:09:54 <izabera> anyway it bails out at the first error
18:10:04 <izabera> so it solves the first one, then gets an error
18:10:06 <oerjan> a hunch: what happens if your test includes strange chars that bash might have trouble with?
18:10:10 <oerjan> (e.g. ')
18:10:11 <izabera> so it could potentially solve more than 1
18:10:20 <izabera> no problem with any character
18:10:31 <oerjan> hmph
18:11:07 <izabera> i'll change the exit to a break and resubmit, let me see how bad it is
18:11:34 <oerjan> izabera: um how are you reading the input, it's supposed to come from a file given on the command line
18:11:56 -!- mroman has quit (Quit: Lost terminal).
18:11:57 <izabera> while loop in line 41
18:12:16 <oerjan> oh i didn't notice the < "$1" at the end
18:12:56 <izabera> cool, now it timeouts
18:17:13 <oerjan> fancy
18:17:17 -!- oerjan has quit (Quit: Later).
18:25:19 <int-e> if you start with substrings of length 2 of a^40b^40...y^40, then it does take a while to reconstruct the string
18:25:59 <int-e> and that's still true if you change the piece length to 8
18:26:25 <int-e> or 28
18:27:09 <int-e> ("a while" meaning several seconds: about 5 here)
18:29:53 <int-e> (concrete input: http://int-e.eu/~bf3/tmp/something)
18:30:13 <izabera> oh god int-e wtf is that
18:30:45 <int-e> izabera: well, try your bash script on it... burn a bit of CPU.
18:32:39 <int-e> (I think this is within the parameters: pieces of length 28, and there's 973 of them)
18:34:15 <int-e> Hmm, the file also produces unintended but interesting visual effects at various line break lengths, fun...
18:37:17 <int-e> e.g. http://int-e.eu/~bf3/tmp/something.png
18:41:19 <int-e> (I have not tried to figure out why it's slow; note that the order that the pieces are given in matters... so this may be far from the worst case)
18:42:05 -!- hppavilion[1] has joined.
18:44:35 -!- zadock has quit (Ping timeout: 264 seconds).
18:52:35 -!- Welo has quit (Quit: Leaving).
18:54:49 <izabera> so it just times out even if my code is fine
18:54:51 <izabera> great
18:55:38 -!- hppavilion[2] has joined.
18:56:22 <izabera> actual input i managed to get from the site, + solution produced by my code http://arin.ga/3eZ8kV/raw
18:56:56 -!- zadock has joined.
18:57:10 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
18:58:38 -!- atrapado has joined.
19:16:51 <quintopia> why isn't mroman here
19:22:49 <fizzie> int-e: The png looks FAST.
19:23:44 <fizzie> Just, you know, æsthetically.
19:25:03 -!- Phantom_Hoover has joined.
19:26:23 <quintopia> pbbbt
19:27:13 <quintopia> @tell boily Give me an upboat: http://codegolf.stackexchange.com/questions/62732/implement-a-truth-machine/63402#63402
19:27:13 <lambdabot> Consider it noted.
19:27:34 <Taneb> fizzie, oh wow it does
19:28:19 <quintopia> lightning!
19:30:49 <quintopia> where in england is "not" pronounced like "naught"? is it like the whole country or something?
19:31:03 <Taneb> I don't
19:31:09 <Taneb> So it's not the whole country
19:31:16 <quintopia> oh okay
19:31:18 <int-e> hmm, english pronunciation i a knotty business
19:31:27 <quintopia> and a naughty one too
19:31:55 <int-e> naught / not / knot was indeed my inspiration
19:32:38 <quintopia> but i know it's said that way in brighton... the kooks guy does it
19:33:21 <quintopia> in the U.S., e.g. Cali, the vowel shift moves in the other direction
19:34:00 <quintopia> (i.e. the "cot" side of the cot/caught merger)
19:34:40 <quintopia> i'm pretty sure england is an esolang
19:37:03 <Phantom_Hoover> i like how 'compiler language with no pronounceable acronym' has a perfectly pronouncable acryonym in welsh
19:39:13 <Taneb> Kinda like cloonpa, right?
19:39:28 <Taneb> No, that would by CLYNPA
19:39:32 <Taneb> Maybe?
19:39:37 <Taneb> I don't know Welsh phonetics
19:43:33 <b_jonas> oh man. it was a strange day today.
19:43:38 <Taneb> Looking it up, I think it'd be cloonpa
19:43:45 <Taneb> b_jonas, what happened?
19:44:13 <b_jonas> Taneb: work stuff. complicated meeting where we actually talked about lots of relevant stuff, even if we didn't quite see eye to eye.
19:44:27 <b_jonas> we definitely couldn't solve everything, but at least know more about what each of us thinks should be solved.
19:44:57 <b_jonas> So basically, I waited for the meeting because some people were busy so it had to be delayed and then delayed again,
19:45:10 <b_jonas> then participated on the complicated meeting and asked stupid questions and took notes,
19:45:23 <b_jonas> then wrote a memo which was difficult because it was a long meeting with lots of things said,
19:45:38 <b_jonas> but still had to be done so that later people can't claim they don't remember what happened on the meeting,
19:45:55 <b_jonas> and in the middle of this I did lots of other small things that had to be done.
19:46:55 <b_jonas> Like, with changes in how some procedures are done at the job, which are for the better in the long term, but cause a little hitch when we change to it, and people don't always know what's happening and say contradictory things.
19:49:43 <Taneb> Today I wrote a FizzBuzz that calls C (indirectly) from Agda
19:50:40 <b_jonas> Taneb: great!
19:56:13 <HackEgo> [wiki] [[DStack]] http://esolangs.org/w/index.php?diff=45231&oldid=45168 * -Dark-Phantom- * (-219) New instructions, string literals and improved examples
19:56:47 <Taneb> b_jonas, https://github.com/Taneb/AgdaCBuzz
19:57:22 <Taneb> Build instructions are non-existent, but it's cabal install ==> compile the agda file
20:05:39 <hppavilion[2]> I'm happy with my command prompt design :)
20:05:44 -!- hppavilion[2] has changed nick to hppavilion1.
20:05:48 <HackEgo> [wiki] [[Hello world program in esoteric languages]] http://esolangs.org/w/index.php?diff=45232&oldid=45159 * -Dark-Phantom- * (-60) /* DStack */
20:06:17 -!- hppavilion1 has changed nick to hppavilion[1].
20:06:23 <HackEgo> [wiki] [[List of quines]] http://esolangs.org/w/index.php?diff=45233&oldid=45169 * -Dark-Phantom- * (-439) /* DStack */
20:06:24 <hppavilion[1]> There we go
20:06:55 <HackEgo> [wiki] [[Truth-machine]] http://esolangs.org/w/index.php?diff=45234&oldid=45160 * -Dark-Phantom- * (-10) /* DStack */
20:07:09 -!- Patashu has joined.
20:18:57 -!- zadock has quit (Quit: Leaving).
20:29:00 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
20:33:58 -!- _leaf_ has joined.
20:34:35 -!- hppavilion[1] has joined.
20:37:12 -!- hppavilion[2] has joined.
20:39:24 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
20:45:54 -!- Patashu has quit (Ping timeout: 250 seconds).
20:57:03 -!- Frooxius has quit (Quit: *bubbles away*).
20:57:06 -!- ^v has joined.
21:24:45 <b_jonas> fungot, ais523 still hasn't returned?
21:24:45 <fungot> b_jonas: fnord/ court/ state/ fnord. -user:splashsplashsmallsupuser talk:splashtalk/sup/small 00:24, 26 january 2007 ( utc))
21:30:01 -!- TieSoul has quit (Read error: Connection reset by peer).
21:36:13 <HackEgo> [wiki] [[SPLEMIT21]] N http://esolangs.org/w/index.php?oldid=45235 * 93.198.137.24 * (+1090) Created page with "SPLEMIT21 ''(short for Shortest Programming Language Ever Made In The 21st [Century])'' is a joke language developed by [http://twitter.com/fucketh1cs Maximilian Krause aka fu..."
21:37:54 <HackEgo> [wiki] [[List of quines]] http://esolangs.org/w/index.php?diff=45236&oldid=45233 * 93.198.137.24 * (+24) /* Cheating Quines */ SPLEMIT21
21:40:19 -!- hppavilion[2] has quit (Ping timeout: 240 seconds).
21:50:54 <izabera> do you say 'the omega+1-st step' ?
21:51:33 <izabera> or 'the omega+1-th step' ?
21:52:34 <b_jonas> izabera: "the step of index omega plus one", or "on the day omega plus one"
21:52:55 <b_jonas> izabera: oh, you mean which of "st" or "th" suffix you use? probably "st" because English is crazy
21:53:04 <izabera> yeah that
21:55:51 <izabera> i guess + changes the normal rules
21:57:32 -!- hppavilion[2] has joined.
22:03:02 <hppavilion[2]> I now have an editor made solely for manipulating the default filesystem for my OS
22:04:28 <izabera> a fs editor?
22:04:46 <izabera> what does it do?
22:05:49 <hppavilion[2]> izabera: It allows me to manipulate files outside of the OS for convenience
22:06:09 <hppavilion[2]> No way to create a new file though, at least not yet
22:07:58 <b_jonas> hppavilion[2]: can it grow the size of an existing file?
22:08:27 <hppavilion[2]> b_jonas: I believe so; I mean, no reason it wouldn't be able to
22:08:58 <hppavilion[2]> I'm using python for this (so it's not a "real" OS), meaning that I don't have arbitrary limitations like I would in C or the like
22:10:11 <b_jonas> hppavilion[2]: finding where the bytes of the file are located by reading the metadata is easy. you have to know more about the fs than that if you also want to modify the metadata to allocate space for growing the file and associating that space with the file.
22:10:58 <b_jonas> if you know where the bytes are stored, you can modify them directly, that much is easy.
22:11:18 <b_jonas> this may depend on the details of file system, sure, but still
22:12:19 <hppavilion[2]> b_jonas: I'm storing the FS in JSON xD
22:12:40 <hppavilion[2]> It's saved to disk as a ".fsys" file (windows) that is really just a subset of JSON
22:13:22 <b_jonas> that sounds strange for a filesystem, but ok
22:13:42 <hppavilion[2]> b_jonas: I mean that that's how it's stored when the OS is closed, so that you can go back to it later
22:14:26 <b_jonas> does your fs allow storing arbitrary byte sequences as file content? if so, how do you encode that to json?
22:15:31 -!- atrapado has quit (Quit: Leaving).
22:15:35 <hppavilion[2]> b_jonas: Yes it does; the file data is just a string.
22:15:56 <hppavilion[2]> ALL files are like that, if you're willing to abuse ASCII
22:16:54 <izabera> you can't seek if you have to decode json
22:17:24 <hppavilion[2]> izabera: ?
22:17:31 <hppavilion[2]> Can't seek as in...?
22:17:41 <izabera> as in "go to byte 12345 in this file"
22:18:14 <b_jonas> hppavilion[2]: the input of JSON encoding is a deep structure that contains character strings, and the output is a character string (often encoded as UTF-8, which is recommended for JSON).
22:18:49 <b_jonas> hppavilion[2]: if you want to store binary strings, you have to somehow encode them to character strings, or else you won't know what backslash escapes to use for characters that you can't represent literally.
22:19:00 <hppavilion[2]> izabera: Ah. I can do that, because I decode it to Python Dicts on loading. I think that works.
22:19:20 <hppavilion[2]> b_jonas: I used the json module, which seems to handle everything like that for me xD
22:19:30 <hppavilion[2]> I'm doing this the cheaty way
22:20:06 <b_jonas> ah, so basically you don't know what your code does. ok.
22:20:24 <b_jonas> that could work.
22:20:45 <hppavilion[2]> b_jonas: That's how programming works xD
22:21:16 <hppavilion[2]> I don't so much "Not know what my code does" as "Using external services and tools to do the heavy lifting, which I need not understand because they're complicated"
22:21:29 <hppavilion[2]> "But they work, so it's OK"
22:21:38 <hppavilion[2]> "And they're open-source, so it's doubly-OK"
22:21:38 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
22:22:03 -!- Phantom_Hoover has joined.
22:36:02 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
22:36:52 -!- sacadouche has joined.
22:38:11 <sacadouche> list
22:40:09 -!- bb010g has quit (Quit: Connection closed for inactivity).
22:42:26 <izabera> `welcome
22:42:32 -!- _leaf_ has quit (Ping timeout: 272 seconds).
22:42:34 <izabera> `welcome sacadouche
22:42:38 <izabera> aww come on
22:42:39 <HackEgo> sacadouche: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/>. (For the other kind of esoterica, try #esoteric on EFnet or DALnet.)
22:42:40 <HackEgo> Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/>. (For the other kind of esoterica, try #esoteric on EFnet or DALnet.)
22:42:50 <izabera> did i lag or HackEgo did?
22:43:22 -!- hppavilion[2] has quit (Ping timeout: 272 seconds).
22:43:23 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
22:43:43 <Taneb> I think you did
22:44:09 -!- hppavilion[2] has joined.
22:44:11 <izabera> :(
22:46:14 <HackEgo> [wiki] [[Talk:Esoteric Operating System]] http://esolangs.org/w/index.php?diff=45237&oldid=43607 * Hppavilion1 * (+129) /* Making one */ new section
22:46:24 -!- hppavilion[2] has changed nick to hppavilion[1].
22:48:41 <hppavilion[1]> I would like someone to define a command-line shell inspired by BF (because it's like the crowning esolang, despite its flaws (namely, the derivatives)). I can implement it.
22:48:48 <zzo38> SQLite now includes an extension to read/write JSON (although I also have my own extension to read JSON data with SQLite; my own doesn't write though)
22:51:03 <zzo38> hppavilion[1]: I wouldn't know how to define such a thing (as I cannot think of it right now), although if I do I will tell you; but probably you or someone else to figure it out.
22:53:26 <hppavilion[1]> I'm making the shape of the FileSys a digraph.
22:54:26 <zzo38> If you want to make it a graph then you might use RDF though, rather than JSON
22:56:17 <hppavilion[1]> zzo38: Well it's a digraph, which is fairly easy to represent with JSON
22:57:20 <hppavilion[1]> Perhaps a HYPERgraph would be more Eso...
22:57:49 <hppavilion[1]> zzo38: What do you think about a hypergraph filesys?
22:59:09 <hppavilion[1]> Maybe later. Don't know yet how to make that work
22:59:26 <zzo38> I don't know either, but if you do know how to make it work, go ahead and try please!
23:01:12 <zzo38> You certainly can represent a digraph with JSON, although I think RDF would work better for this purpose. For a hypergraph you will need to do something more (it can also be represented with RDF easily enough, I expect)
23:01:59 -!- boily has joined.
23:02:34 <izabera> i thought you used json because objects form a tree-like structure
23:02:47 -!- Phantom_Hoover has joined.
23:03:39 <izabera> i don't see how a generic graph is easy to represent with json
23:03:46 <zzo38> JSON is good for tree-like structure yes
23:06:35 <boily> what happens when a YAML parser parses a mutually recursive structure?
23:06:51 <zzo38> <> :edges [:v _:v1, _:v2, _:v3], [:v _:v2, _:v3], [:v _:v3, _:v5, _:v6], [:v _:v4]. # It is one possible way you may represent a hypergraph in RDF; you could even have more than one that shares nodes in this way
23:07:09 -!- oerjan has joined.
23:07:47 <boily> @massages-loud
23:07:48 <lambdabot> quintopia said 3h 40m 34s ago: Give me an upboat: http://codegolf.stackexchange.com/questions/62732/implement-a-truth-machine/63402#63402
23:08:52 <boily> @tell quintopia the upboat was gladly given! :D
23:08:52 <lambdabot> Consider it noted.
23:09:07 <shachaf> implement a pooch machine
23:09:29 <boily> hellochaf. pooch?
23:09:51 <boily> @tell quintopia heh, nice usage of "-ai".
23:09:51 <lambdabot> Consider it noted.
23:10:31 <shachaf> boily: pooch
23:11:09 <shachaf> boily: apparently it means cabot hth
23:12:06 <boily> ah, un cabot!
23:12:28 <boily> tdh.
23:12:54 <quintopia> boily: its the fastest way to subtract large nums
23:13:08 <shachaf> boily: http://i.imgur.com/9n3q5pV.gifv
23:14:30 <boily> PURE UNDILUTED WIZZARDRY!
23:14:57 <izabera> what is diluted wizardry like?
23:15:18 <oerjan> <-- i guess you need some kind of faster dictionary that allows you to quickly find the neighboring nodes and edges
23:15:21 <oerjan> oops
23:15:26 <oerjan> <izabera> so it just times out even if my code is fine <-- i guess you need some kind of faster dictionary that allows you to quickly find the neighboring nodes and edges
23:15:52 <oerjan> don't know if bash has anything
23:15:56 <izabera> it does
23:16:00 <izabera> associative arrays
23:16:15 <izabera> i was trying to improve it but then i started playing online <.<
23:16:19 <zzo38> The JSON1 extension of SQLite does not include aggregate functions.
23:17:11 <boily> izabera: Fentimans.
23:17:32 <quintopia> boily: you didnt vote :(
23:18:21 <boily> quintopia: yes I did, but because I just created an account it's not publicly visible :(
23:18:41 <boily> (I need 15 reputations for it to become public.)
23:18:48 <oerjan> hm you can vote immediately?
23:19:29 <boily> I can, and I get a nice small informative blue popup telling me I need to build more reputations.
23:19:50 <oerjan> well i didn't know it actually _saved_ your vote?
23:19:57 <oerjan> as opposed to ignoring it
23:20:46 <boily> Thanks for the feedback! Once you earn a total of 15 reputation, your votes will change the publicly displayed post score.
23:20:58 <oerjan> but then i didn't try voting in that short timespan
23:21:12 <oerjan> i probably didn't even register properly
23:21:18 <boily> :/
23:21:44 <shachaf> uh oh
23:21:49 <shachaf> boily is too disreputable to vote?
23:22:06 <oerjan> i recall i waited about a year before i turned my loginless account into a proper one, and only because the cookie expired so it was the only way to get it back
23:22:49 <oerjan> boily: btw if you have enough rep on another SE account you can get 100 free by linking them
23:23:02 <boily> yeeeeesh...
23:29:05 -!- hppavilion[2] has joined.
23:29:20 <oerjan> `? yeeeeesh
23:29:21 <HackEgo> yeeeeesh? ¯\(°​_o)/¯
23:29:47 <oerjan> the only place i know yeeeeesh from is the Mutts comic
23:29:53 <oerjan> well, yesh
23:31:27 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
23:36:10 -!- dtscode has changed nick to nchambers.
23:52:01 -!- boily has quit (Quit: KNIT CHICKEN).
23:55:50 -!- hppavilion[2] has quit (Ping timeout: 240 seconds).
←2015-11-08 2015-11-09 2015-11-10→ ↑2015 ↑all