00:02:31 -!- chaosagent has joined.
00:03:09 <quintopia> dammit. how am i going to overwrite my OS with this one if ubi-partman always crashes :(
00:05:19 -!- callforjudgement has joined.
00:05:32 -!- ais523 has quit (Disconnected by services).
00:05:34 -!- callforjudgement has changed nick to ais523.
00:10:14 -!- vanila has quit (Quit: Leaving).
00:14:28 <boily> what's an ubi-partman?
00:15:15 <quintopia> now i'm having trouble even getting the graphics to boot
00:15:18 <ais523> thought experiment: someone new joins the channel, claiming to be really interested in an obscure bad esolang
00:15:30 <ais523> how bad does it have to be before you assume they are / they are in league with its creator?
00:15:44 <boily> quintopia: don't worry. partitioning is either the simplest thing, or the hardest thing.
00:16:08 <quintopia> partitioning will be easy if i can even GET to that point
00:16:20 <boily> (one of my friends has a desktop. I don't want to touch that desktop anymore. the boot process on that thing greatly surpasses my necromancy skills.)
00:16:22 <quintopia> since i'm just going to wipe every thing and start over
00:16:49 <quintopia> right now all i'm getting is a screen that keeps flashing
00:16:56 <FireFly> ais523: I think that might describe my first visit to the channel
00:16:59 <boily> quintopia: if you haven't done it yet, burn an image of grml on a disc or stick, and boot that. this distro is magic.
00:17:05 <ais523> FireFly: which language was it?
00:17:18 <boily> quintopia: it should most verily probably solve almost all your problems.
00:17:25 <boily> (almost used in the mathematical sense.)
00:18:02 <quintopia> that implies that the number of possible problems is infinite
00:18:02 <boily> quintopia: it has heavy hardware autodetection, with manual options in case your machine is rebarbative and resists your dominance.
00:18:13 <quintopia> i think the number of possible problems is really a large finite number
00:18:18 <boily> I never said that the set of possible problems is finite.
00:18:31 <boily> large finite numbers are good approximations of infinity.
00:18:34 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
00:18:56 <quintopia> every finite number i know is much less than infinity
00:19:56 <boily> large numbers either hit NaN or Infty. therefore, they are good approximations.
00:20:24 <boily> but apart from that, you really should get a copy of grml.
00:20:45 * boily doesn't have any sentimental attachment towards grml, I swear.
00:21:28 <quintopia> i have a sentimental attachment to the obsolete system i am replacing
00:22:34 <Koen_> quintopia: sequences of large finite numbers are a good approximation of infinity
00:22:38 -!- ais523 has quit (Read error: Connection reset by peer).
00:22:53 -!- ais523 has joined.
00:23:23 <quintopia> Koen_: only infinite sequences! but that's beggaring the question!
00:25:15 <Koen_> quintopia: are you trying to tell ihatehex he's a value virgin?
00:26:03 <quintopia> mostly i'm just complaining about people hinting at problems they are trying to solve without actually saying what problem it is
00:29:52 <quintopia> i feel like i've broken this usb drive by trying to boot from it >.>
00:30:57 <oren> that can happen
00:33:21 <quintopia> it worked once, but now it only ever goes into an infinite loop of turning the display off and on (sleep/wake). i will deal with this another day.
00:57:44 -!- hjulle has quit (Ping timeout: 245 seconds).
01:01:28 -!- Koen_ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
01:03:27 -!- Patashu_ has joined.
01:03:36 -!- Patashu has quit (Disconnected by services).
01:25:36 <ais523> irrelevant URL for the content, but for anyone interested in esoprogramming in Magic: the Gathering, this page trying to find the largest non-infinite combo in Magic is pretty amusing: http://www.rpgdl.com/metroidcomposite/phpconversion.php
01:36:57 -!- ais523 has quit (Read error: Connection reset by peer).
01:36:58 -!- callforjudgement has joined.
01:37:26 -!- callforjudgement has changed nick to ais523.
01:57:24 -!- koo7 has quit (Quit: Leaving).
01:59:11 -!- koo7 has joined.
02:16:56 -!- callforjudgement has joined.
02:16:57 -!- ais523 has quit (Read error: Connection reset by peer).
02:22:26 -!- scarf has joined.
02:24:25 -!- callforjudgement has quit (Ping timeout: 264 seconds).
02:24:40 -!- scarf has changed nick to ais523.
02:57:44 <oren> this just in-- woman asks what color dress is, internet goes down in flaming ruins
02:58:47 <ais523> I assume this is some sort of huge meme
02:58:57 <ais523> because I've seen it be brought up unprompted in like 4 different places
03:00:23 <oren> basically people can't all agree on whether the dress is white and gold (me) or blue and black
03:00:51 <ais523> based on what? seeing it in person? photos?
03:01:02 <ais523> with photos it could literally be white and gold in one, and blue and black in another, depending on lighting
03:01:12 <oren> http://swiked.tumblr.com/post/112073818575/guys-please-help-me-is-this-dress-white-and
03:01:20 <oren> this photo specifically
03:02:19 <ais523> oh, that depends a lot on gamma
03:02:33 <ais523> if you're viewing it on a laptop screen, try changing your viewing angle
03:02:38 <oren> https://twitter.com/emmyrossum/status/571128558325608448
03:02:55 <ais523> it's black and blue when viewed from below, white and gold when viewed from above
03:03:14 <ais523> so different people could genuinely be seeing different colours
03:03:28 <oren> HOLY SHIT how did that happen
03:04:06 <ais523> so I guess what's happened is that the Internet hasn't discovered the existence of gamma yet
03:04:14 <ais523> (also, fwiw, I see it as blue and gold in an in-between gamma)
03:08:40 <oren> so if the screen angle is causing it, maybe different people naturally hold their tablets at different angles
03:09:18 <Jafet> Tablets that use IPS shouldn't be affected by viewing angle
03:09:23 <ais523> inconsistent gamma based on viewing angle is one of the major disadvantages of LCDs
03:13:00 <int-e> I'm so tempted to make a screenshot to prove that it's blue and black :P
03:15:06 <ais523> we could use the xkcd colour classification chart, perhaps?
03:15:14 <ais523> although I'm not sure the whole three-color-dimensional thing is public
03:17:50 <int-e> there are lengthy (though not very substantial) articles about this... http://motherboard.vice.com/en_ca/read/there-is-the-dress-and-only-the-dress
03:20:09 <quintopia> i cant see white and gold at any angle
03:23:13 <ais523> anyway, one way to think about it
03:23:17 <int-e> I'd like to know whether this was accidential (an actual photograph that was hopelessly overexposed) or deliberate.
03:23:20 <ais523> is that white and blue are adjacent colors (in between, you just have light blues)
03:23:25 <ais523> and likewise, black and gold are adjacent
03:23:36 <ais523> so there have to be some colors that land right on the boundary
03:29:32 <oren> Haha! I jury rigged an LCD monitor SIDEWAYS and it doesn't have the problem. (instead the colrs shift with left-right viewing angle)
03:31:55 <ais523> I'm not sure that counts :-)
03:32:08 <int-e> oren: that sounds awful
03:33:04 <oren> I taped the mount on the bottom to the wall with a lot of duck tape
03:36:14 <oren> I think I might keep it like this
03:36:57 -!- arjanb has quit (Quit: .).
03:41:46 <oren> http://ctrlv.in/509633
03:45:30 <oren> actually nahhh it does'nt work with fullscreen
03:47:10 <oren> Hmmm, what If I took two and taped them side-by side
03:48:22 <oren> if each one is 16:9 then the result would be 18:16
03:50:10 <oren> but I would need to knock together a wooden frame to hold them, too much work
03:52:20 <Jafet> Multiple screens h ave other problems.
04:02:59 <oren> hmm, I can also tape the monitor over my head
04:03:29 <oren> yeah, that workd
04:04:08 -!- boily has quit (Quit: TAPERED CHICKEN).
04:06:41 <oren> ideally it would be even more overhead, but I think i would need to find a stud to nail the monitor up
04:07:22 <oren> http://ctrlv.in/509648
04:08:03 -!- MDude has changed nick to MDream.
04:10:28 <oren> hold on, maybe I can suppor the monitor overhead with a wooden frame
04:12:15 <quintopia> why the heck do you want it overhead?
04:12:56 <oren> So I can be even more lazy and not even lift my head to loom at my screen
04:13:06 <HackEgo> [wiki] [[Deviating Percolator]] http://esolangs.org/w/index.php?diff=42064&oldid=42000 * AJF * (+56) /* Expressions */ Munroe algorithm
04:18:56 <FireFly> So I turned my font into a proper one; here's what IRC looks like: http://xen.firefly.nu/up/2015-02-28_051818.png
04:21:16 <zzo38> The font is not bad
04:21:50 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
04:34:13 -!- PinealGlandOptic has joined.
04:34:31 <zzo38> When trying to send something to sprunge now I get a HTML document back instead of the proper response; the HTML document says "503 Over Quota" on it.
04:35:25 <ais523> zzo38: presumably you've uploaded too much to sprunge, and it isn't accepting any more right now
04:35:38 <ais523> although that should be a 4xx rather than 5xx code
04:35:53 <ais523> whatever the user error code is
04:37:06 <zzo38> When am I supposed to try again?
04:37:37 <zzo38> It says "Please try again later." but I don't know what time
04:39:00 <zzo38> Even such a simple HTML document is full of worthless junk
04:40:03 -!- chaosagent has quit (Remote host closed the connection).
04:40:22 -!- chaosagent has joined.
04:57:52 <Jafet> It's called markup because it's so expensive to parse
04:59:09 <MDream> "Please try again harder."
05:00:12 -!- Lymia has quit (Ping timeout: 256 seconds).
05:11:48 -!- ProofTechnique has joined.
05:24:22 <zzo38> Is there any Magic: the Puzzling that involves conceding and/or subgames?
05:25:16 <ais523> I doubt it; conceding doesn't fit into the normal 1v1 setups, and in order to work out what would happen in a subgame, you'd need to know what's in the opponent's deck
05:26:16 <zzo38> Yes, although some puzzles tell you what's in the opponent's deck; there are also things you can use to affect the opponent's library before the subgame starts.
05:26:56 <zzo38> The possibility to concede also might be important if there are subgames
05:27:22 <zzo38> Or if you are making a puzzle involving Team vs Team instead of 1v1
05:31:56 <zzo38> Maybe it is even necessary to concede a subgame you have started, and to somehow stop your opponent from conceding before you have completed what you need to do, perhaps by making opponent have only 1 life point in the main game, and then if Shahrazad is used they will lose half, rounded up
05:38:36 <oren> Shahrazad? How many other cards refer to reality like that?
05:39:03 <oren> Well, real literature anyway
05:39:54 <zzo38> Some very old flavor texts do refer to real literature
05:40:43 <oren> Holy shit therea re cards alled "jihad" and "army of allah" how did they get away with that?!?!?
05:41:00 <zzo38> Yes, those are very old
05:41:10 <zzo38> Also there is a card "Aladdin"
05:41:41 <ais523> I think the origin arabian nights stories are out of copyright by now
05:42:10 <zzo38> But even outside of the Arabian Nights set, some old cards referenced older literature including Shakespeare and the Bible and a few others
05:44:38 <oren> oh, they got away with it because they came out the year I was born... the whole war hadn't started yet
05:46:05 <oren> I like how "king suleiman" is a card.
05:46:32 <oren> since he was a completely real person, but from CENTURIES later
05:49:34 <oren> He appears as a enemy ai character in aOE3
05:50:36 -!- Froox has quit (Quit: *bubbles away*).
05:53:06 <oren> Interesting to scroll chronologically through the versions of a card. Like apparently before I was playing, there was a period where "destroy, it can't be regenerated" was shortened to "bury"
05:54:26 <oren> But before that, it was originally "destroy, it can't regenerate"
05:55:15 <zzo38> The original "Flight" said "Target creature is now a flying creature."
05:55:42 -!- ProofTechnique has quit (Ping timeout: 245 seconds).
05:57:39 <zzo38> The original text of "Black Lotus" didn't list an activation cost; it said "Adds 3 mana of any single color of your choice to your mana pool, then is discarded. Tapping this artifact can be played as an interrupt."
05:57:59 <ais523> zzo38: that's because it had "mono artifact" on its type line
05:58:15 <ais523> which means that anything that looks like a tap ability in its text box is a tap ability
05:58:39 <zzo38> Yes, I know, "mono artifact" means it can be activated by tapping in additional to any other activation costs if any.
05:59:06 <zzo38> And then there is also "poly artifct" which is similar but doesn't require tapping; however I think it still can't be used if it is already tapped
06:00:32 <zzo38> However, there is another thing too: Before the Alpha cards were even released, they originally had the activation costs printed next to the mana cost, and there was no "mono artifact" and "poly artifact" either; any card that did not require tapping to activate explicitly said so.
06:01:06 <zzo38> Another thing about the pre-Alpha cards is that the Plains had pictures of airplanes on them.
06:01:23 <zzo38> Did you know any of these things before?
06:05:44 <oren> no... how the hell does "Copper Tablet"'s effect make any sense?
06:06:56 <oren> actually better question: why does it say "does X damage" instead of "deals X damage" on old cards?
06:08:09 <oren> "does damage" sounds childish to me somehow
06:13:52 <oren> Ok weirdest thing: one particular version of the card "wrath of god", instead of saying "destroy all creatures, they can't be regenerated", instead says ""
06:13:55 <oren> Put all creatures into their owners' discard piles. (This includes your creatures.)
06:15:10 <oren> I don't even know if that's equivalent
06:16:49 <oren> well, assuming that regeneration refers to "destroying" a creature, then it might be
06:18:28 -!- password2 has joined.
06:18:47 -!- password2 has quit (Max SendQ exceeded).
06:19:23 -!- password2 has joined.
06:19:50 -!- password2 has quit (Max SendQ exceeded).
06:20:31 -!- password2 has joined.
06:23:21 <ais523> hmm, repeated joining + Max SendQ exceeded is rarely a good sign
06:29:02 -!- password2 has quit (Read error: Connection reset by peer).
06:38:21 -!- Lymia has joined.
06:45:28 -!- password2 has joined.
06:46:30 -!- password2 has quit (Read error: Connection reset by peer).
06:57:07 -!- Frooxius has joined.
07:02:06 -!- password2 has joined.
07:11:08 -!- password2 has quit (Ping timeout: 245 seconds).
07:33:54 -!- ais523 has quit (Read error: Connection reset by peer).
07:34:10 -!- ais523 has joined.
07:38:10 -!- ais523 has quit (Read error: Connection reset by peer).
07:38:25 -!- ais523 has joined.
08:23:05 -!- oerjan has joined.
08:29:02 -!- chaosagent has quit (Ping timeout: 245 seconds).
09:10:05 -!- bb010g has quit (Quit: Connection closed for inactivity).
09:29:03 -!- atehwa has quit (Remote host closed the connection).
09:29:03 -!- villasukka has quit (Remote host closed the connection).
09:36:56 -!- ais523 has quit.
09:55:38 -!- TieSleep has changed nick to TieSoul.
10:19:58 -!- AnotherTest has joined.
10:27:28 -!- Zuu has joined.
10:39:24 -!- Fleur has joined.
11:00:05 -!- AnotherTest has quit (Ping timeout: 246 seconds).
11:08:05 -!- oerjan has quit (Quit: leaving).
11:20:00 -!- Koen_ has joined.
11:26:07 -!- Patashu_ has quit (Ping timeout: 252 seconds).
11:26:26 -!- arjanb has joined.
11:28:33 -!- Zuu has quit (Remote host closed the connection).
11:28:48 -!- Zuu has joined.
12:12:33 -!- FreeFull has quit (Ping timeout: 264 seconds).
12:13:38 -!- Tritonio has joined.
12:33:00 -!- Tritonio has quit (Ping timeout: 252 seconds).
12:59:01 -!- vanila has joined.
13:15:30 -!- ProofTechnique has joined.
13:16:53 <vanila> https://t.co/etIIQZzvFQ
13:16:56 -!- SopaXorzTaker has joined.
13:17:00 <vanila> https://www.youtube.com/watch?v=unCQHAbGsAA
13:17:09 <vanila> what's it like to write a nethack bot
13:20:18 <vanila> im curious how the sokoban bits are done...
13:20:26 <vanila> those are very hard to solve with a program
13:22:00 <int-e> you can always hardcode some bits
13:23:25 <vanila> yeah i guess the sokobans are hard coded, so the solver doesn't need to be general
13:24:10 <int-e> generating sokoban puzzles is even harder (except for very small sizes, but then solving is still easy...)
13:25:30 <int-e> (There are some collections of automatically generated puzzles, based on the principle that one looks for long shortest solutions. They're somewhat interesting for humans to solve because the taxicab metric is generally useless for them.)
13:26:16 <int-e> Not sure if there are more interesting puzzles, I'm not really up-to-date on this topic; perhaps ais523 knows more.
13:26:33 <int-e> ... more interesting puzzle generators, that is.
13:34:45 -!- boily has joined.
13:38:13 -!- MoALTz_ has joined.
13:41:47 -!- MoALTz has quit (Ping timeout: 252 seconds).
13:45:41 -!- AnotherTest has joined.
13:49:28 -!- ProofTechnique has quit (Ping timeout: 265 seconds).
14:06:46 <Jafet> It's PSPACE-complete, so puzzles can contain an unlimited variety of interesting devices that are difficult to invent automatically.
14:14:11 -!- FreeFull has joined.
14:31:47 -!- Koen__ has joined.
14:33:59 -!- Koen_ has quit (Ping timeout: 246 seconds).
14:40:14 -!- GeekDude has joined.
14:41:57 -!- gde33 has joined.
14:53:36 <Jafet> Also beating nethack with a pudding farming dvalk... not exactly impressive
15:04:14 -!- villasukka has joined.
15:16:06 <boily> `relcome villasukka
15:16:07 <HackEgo> villasukka: 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 irc.dal.net.)
15:21:34 -!- AnotherTest has quit (Ping timeout: 264 seconds).
15:24:39 -!- ProofTechnique has joined.
15:36:34 -!- MDream has changed nick to MDude.
15:37:04 <vanila> hows it going? any news
16:00:12 -!- Fleur has quit (Remote host closed the connection).
16:07:04 -!- ^v has joined.
16:09:07 <vanila> https://www.youtube.com/watch?v=vEVq295zKps
16:10:39 <MDude> Well I still ought to make a compiler.
16:10:55 <MDude> Yestersay I tried to figure out some stuff about how IBNIZ works.
16:11:55 <MDude> For an esoteric language I want to make.
16:12:04 <vanila> i like learning about compilers
16:14:22 <vanila> i want to make something
16:14:30 <MDude> I plan to have it compile to IBNIZ code, actually, which is one reason I was trying to start using IBNIZ's memory system.
16:14:44 <MDude> As oppossed to just using stack commands.
16:18:10 -!- Lymia has quit (Ping timeout: 250 seconds).
16:20:54 <MDude> OK, this is annyoing.
16:21:13 <MDude> I try something out and it seems to work differently than when i tired the exact same thing last night.
16:21:23 -!- nortti has changed nick to lawspeaker.
16:21:30 -!- lawspeaker has changed nick to nortti.
16:22:25 -!- Lymia has joined.
16:23:32 <MDude> What I'm trying to do is take the time value, and store that in the memory cell equal to itself modulo the number of display pixels.
16:24:41 <MDude> Then take the X and Y values and turn them into a single number, each cooresponding to one of the same numbers.
16:25:12 -!- Tritonio has joined.
16:25:28 <MDude> With the desired result being that the screen just gets slowly scanned over and filled.
16:37:37 -!- aretecode has quit (Quit: Toodaloo).
16:49:52 -!- ProofTechnique has quit (Ping timeout: 244 seconds).
16:56:30 <HackEgo> 1206) <ais523> wait, is Pluso basically a more limited version of Deadfish? <ais523> Deadfish minimization is not an area of esolang development I had really considered
17:02:00 <HackEgo> 1124) <Taneb> kmc, I was trying to go to a sci-fi and fantasy society social, and I went to the wrong bar <Taneb> Wound up at my university's fetish society <Taneb> Didn't realise for an hour and a half
17:02:18 <boily> ah! my favourite Tanebquote :)
17:02:29 <Taneb> Yes, it was quite an evening
17:02:39 <Taneb> Over a year ago now, wow
17:07:01 <Taneb> Apparently someone made the opposite mistake this year
17:08:27 -!- Frooxius has quit (Quit: *bubbles away*).
17:19:53 -!- Koen__ has quit (Ping timeout: 250 seconds).
17:22:19 -!- oren has quit (Quit: Lost terminal).
17:24:32 -!- bb010g has joined.
17:24:39 <boily> Taneb: you ought to merge the two societies together to prevent unnecessary confusion hth
17:34:49 <olsner> you could send out fake invites inviting the sci-fi/fantasy people to a social at the same place and time whenever the fetish society does something, and vice versa
17:57:13 -!- herbary has joined.
17:57:23 -!- herbary has left.
18:12:07 -!- vanila has quit (Quit: Leaving).
18:13:01 -!- CADD has quit (Quit: Lost terminal).
18:13:21 -!- CADD has joined.
18:14:49 -!- v4s has joined.
18:21:30 -!- boily has quit (Quit: SPREAD CHICKEN).
18:27:15 -!- SopaXorzTaker has quit (Read error: Connection reset by peer).
18:45:46 -!- Frooxius has joined.
18:46:23 <MDude> I just realized the Ibniz instruction manual mentions a "movesp" command that it does not actually describe.
18:46:37 -!- Froox has joined.
18:50:08 -!- Frooxius has quit (Ping timeout: 246 seconds).
18:59:00 -!- Fleur has joined.
19:04:43 -!- gamemanj has joined.
19:46:43 <mroman> Couldn't match expected type `()' with actual type `Maybe t1'
19:46:44 <mroman> In the pattern: Nothing
19:47:02 <mroman> so why the hell does it expect ()?
19:47:34 <int-e> mroman: such complaints are useless without context.
19:47:47 <int-e> > case () of Nothing -> ()
19:47:48 <lambdabot> with actual type ‘Data.Maybe.Maybe t0’
19:50:54 <gamemanj> ...Just read about lambdabot. Interesting. So if I were to do...
19:51:36 <int-e> mroman: fwiw, I treat "Couldn't match expected type 'a' with actual type 'b'" symmetrically; it's hard to predict which of two mismatching types will be 'a' and which will be 'b' for a given type-incorrect program.
19:52:57 <int-e> gamemanj: sorry, #esoteric becomes a secondary #haskell from time to time.
19:53:37 <gamemanj> int-e: ...That's fine. Lambdabot does seem fun for people to play with, and judging by that error certainly esoteric :)
19:55:03 <gamemanj> ...Course, maybe that's just a special case. What would "case () of Nothing" be useful for, exactly?
19:55:57 <int-e> gamemanj: nothing. it was meant to illustrate, by a simple example, how the type error mroman was complaining about could arise in practice
19:56:27 -!- Koen_ has joined.
19:57:48 <int-e> > let f Nothing = 1; f () = 2 in undefined -- testing
19:57:49 <lambdabot> Couldn't match expected type ‘Data.Maybe.Maybe t1’
19:58:08 <int-e> anyway, plenty of possibilities.
20:23:42 -!- J_Arcane has quit (Quit: ChatZilla 0.9.91-rdmsoft [XULRunner 32.0.3/20140923175406]).
20:34:28 -!- ais523 has joined.
20:34:42 <gamemanj> I'm messing around with making a "light" esolang. One that won't hurt people's brains too much, but will compile to another esolang so they can study the results :)
20:37:20 -!- J_Arcane has joined.
20:43:49 -!- ais523 has quit (Ping timeout: 252 seconds).
20:47:29 -!- oerjan has joined.
20:49:03 <int-e> oerjan: did you see Dewcup resurface?
20:49:57 <int-e> (yafgc has such a funny update style; rather than having a buffer, the updates tend to happen in batches...)
20:51:19 <int-e> (Oh and I'm now totally confused about the relative timing of the various stories.)
20:51:34 <oerjan> int-e: there's a forum post with a chronology
20:51:55 <oerjan> unfortunately it's by chapter name, not comic numbers, so still a bit confusing
20:52:51 <oerjan> or rather, not confusing, but it makes it impossible to see where the actual jumps are, when you don't remember the _posting_ chronology
20:52:53 <Koen_> "yet another freaking golf compiler"?
20:53:05 <oerjan> fantasy gamer comic hth
20:53:37 <oerjan> it's sort of D&D based story like oots
20:54:02 <int-e> unlike oots, it doesn't take miles of text to get to a punchline
20:54:08 <oerjan> better drawing (but usually sketches, not properly inked) and frequently somewhat nsfw
20:54:14 <int-e> (sorry, I know I'm repeating myself there)
20:55:35 <oerjan> int-e: the updates are whenever he has free time to draw, i take. and also the recent server move has taken a lot off that time.
20:55:41 <int-e> The very first one is an indicator of how unsafe for work it gets, though maybe the Drow's dungeon (very SM) is worse. http://yafgc.net/comic/bob-meets-gren/
20:56:01 <oerjan> mostly because he's individually reuploading and reindexing every comic
20:56:45 <oerjan> (i could say something about obvious lack of scripting skills, but otoh the new indexing system is pretty nice)
20:57:03 <oerjan> also he isn't finished yet, so some comics are actually missing from the new archive
20:58:51 <oerjan> int-e: hm was i the one who pointed you to yafgc? that makes at least two of your comics my fault, then :P
20:59:13 <int-e> no, I've been reading yafgc like forever.
20:59:34 <int-e> "like" -- I really need to work on my English skills.
20:59:38 <oerjan> maybe it was the other way around, then, i don't actually remember where i learned of its existence
20:59:56 <oerjan> int-e: i think that's pretty normal english in certain sociolects
21:00:19 <oerjan> possibly you'd want it between commas
21:00:20 <int-e> oerjan: It might have been an independent discovery. The comic isn't exactly hard to find.
21:00:48 -!- AnotherTest has joined.
21:01:04 <oerjan> maybe. it was pretty old when i found it. i think it was in the middle of the vampire story.
21:01:28 <oerjan> the one with the penanggalan
21:01:54 <oerjan> (there were some vampires earlier who moved in when lewie left)
21:03:03 <oerjan> it's a bit confusing now with _both_ a bbs-like forum and comments on individual comics
21:03:46 <oerjan> the latter having the annoying property that afaict you have to check every comic to see them.
21:04:53 <oerjan> dewcup resurfacing near lewie was like pretty predictable
21:05:12 <oerjan> now i'm waiting to see how the meeting with ata goes
21:05:34 -!- chaosagent has joined.
21:05:41 <int-e> I didn't make the connection(sic!) between that one underground cave and the other.
21:06:06 <oerjan> um i think that was pretty ad hoc
21:06:22 <oerjan> they're meeting because the plot requires lewie to meet ata
21:06:39 <int-e> oerjan: and of course we still have to see Dewcup's other half to turn up. That will be fun. I expect them both to be evil, but it's hard to predict.
21:07:03 <oerjan> maybe the other half is still in the tomb city
21:07:31 <int-e> Yeah, I've also lost track of the plot, because of all the stories inbetween. :)
21:07:37 <oerjan> or maybe, as some predicted on the forum, dewcup is so unrestrained that she simply doesn't have any repressed sides for a succubus to form with
21:08:27 <oerjan> i think some years passed between dewcup getting captured and finding her way out
21:08:48 <oerjan> although i cannot say how many
21:09:30 <int-e> And what does the Ata priestess have to do with Kila, if anything?
21:10:56 <int-e> Hah, knowing Dewcup's natural charm I bet the priestess didn't last a week before giving her the run of the place (in order to get her out of sight.)
21:10:57 <oerjan> int-e: i suggest clicking on the link to Ata in the index for the comic
21:14:40 <oerjan> maybe i should find it, since it's not on the _current_ comic
21:15:40 <oerjan> um does that mean you found it
21:16:00 <oerjan> http://yafgc.net/character/princess-ata/ anyway
21:16:59 <int-e> Yes, it meant that I found it.
21:17:26 -!- ProofTechnique has joined.
21:19:30 <oerjan> *sigh* the ventilation system has started making that annoying sound again.
21:19:31 -!- Patashu has joined.
21:21:27 <mroman> can you do incremental parsing with parsec?
21:21:47 <oerjan> what kind of incremental
21:22:12 <mroman> Let's say what I want to parse comes from a stream
21:22:24 <oerjan> it's not very designed for it
21:22:26 <mroman> and I don't know if I have enough data present for parsing the next entry
21:22:41 <mroman> also there might be one entry in the buffer + half of the next entry
21:22:50 <mroman> parsec can't really deal with that
21:22:54 <oerjan> but i _think_ you can do it by parsing with a monad transformer
21:23:25 <oerjan> not sure how that interacts with backtracking though
21:23:30 <mroman> can parsec return me as a string what it couldn't parse?
21:24:50 <mroman> You can always wrap everything in a try I guess
21:25:08 <mroman> but that wouldn't tell me how far it parsed
21:26:02 <oerjan> "attoparsec supports incremental parsing directly"
21:26:38 <oerjan> http://stackoverflow.com/questions/15752243/incremental-parsing-from-handle-in-haskell
21:27:05 <oerjan> "Parsec, alas, doesn't."
21:29:33 <mroman> you can use try lookAhead I guess
21:29:52 <mroman> and call getPosition to determine how far it parsed
21:30:00 <mroman> but I think you *could* still do it
21:30:01 <int-e> Hmm, is that true? http://hackage.haskell.org/package/parsec-3.1.8/docs/Text-Parsec.html#t:Stream looks like this is now possible, though I'm not sure it's possible to not store everything that was already read for potential backtracking...
21:30:17 <mroman> "Returns the full parser state as a State record."
21:31:30 <oerjan> int-e: the problem isn't _reading_ incrementally, but parsing incrementally if you need lookahead.
21:32:23 <oerjan> parsec returns just one result, in bulk, and if that result used lookahead at the end your Stream will have read it, i think.
21:32:35 <oerjan> so restarting won't do the right thing.
21:34:12 <int-e> oerjan: well, it's a bit of both
21:34:30 <int-e> reading on demand and returning partial results
21:35:21 <mroman> oh well you can totaly do that
21:35:25 <int-e> Note also that that Stream class is a recent addition to Parsec, added in September last year, in version 3.1.6.
21:35:36 <oerjan> mroman: that's not likely to be lazy, though.
21:35:37 <mroman> *Main> runParserWithString parseIt "(9 9)(8"
21:35:38 <mroman> ("(8",RLON_Object (RLON_Integer 9) [RLON_Integer 9])
21:36:01 <mroman> state <- getInput; return $ (state, rlon)
21:36:10 <int-e> lambdabot is still on parsec-3.1.5
21:36:12 <mroman> you can call the parser
21:36:31 <mroman> then use getInput at the and of your parse function and return a tupel of the parsed stuff and a string of what's left to parse
21:36:32 <oerjan> mroman: i'm saying that you won't get lazy parsing this way
21:36:50 <mroman> why would I need lazy parsing?
21:37:11 <oerjan> well if you are reading from a file...
21:37:46 <oerjan> hm if you are reading from a list, will getInput just return it
21:38:04 <int-e> Parsec isn't the only parsing library out there. I wonder if anybody has done a comprehensive survey...
21:39:05 <mroman> I think try isn't that lazy anyway
21:39:10 <mroman> it requires backtracking after all
21:39:30 <mroman> so having a huge try will cause parsec anyway to keep the whole stuff in memory all the time
21:40:06 -!- bb010g has quit (Quit: Connection closed for inactivity).
21:40:44 <mroman> but you need a dummy thing you can return when parsing failed
21:40:50 <mroman> (i.e. when nothing could be completely parsed)
21:40:58 <mroman> http://codepad.org/LfQ80IF7
21:41:05 <mroman> wrap the outermost parse function in a try
21:41:08 <oerjan> mroman: i am thinking that you can parse a list incrementally in parsec by using ParsecT ... (Writer [...])
21:41:13 <mroman> then do (try outer) <|> (return DummyShit)
21:41:40 -!- sdegutis has joined.
21:41:43 -!- sdegutis has left ("Leaving...").
21:41:43 <oerjan> and _instead_ of return a result, you just lift $ tell ... each list item once you're sure of it
21:42:29 <oerjan> ...i guess this will ruin the normal error reporting, though.
21:42:41 <mroman> you can't tell if a parse error occured
21:42:45 <mroman> or not enough input is there
21:43:32 <mroman> so if I ever need incremental parsing
21:43:54 <mroman> (I don't actually need it right now, I was just curious)
21:44:01 <int-e> oerjan: lazy, incremental parsing is conceptually weird. you have to accept that you process partial results before you know that there won't be a syntax error later on.
21:44:14 <mroman> (I might do some network stuff some day)
21:44:50 <mroman> can you parse integer <> double without try btw?
21:44:56 <int-e> oerjan: And it gets weird if the result is in any way tree-structured, rather than a stream.
21:45:19 <oerjan> mroman: definitely not without making a custom combinator
21:45:39 <mroman> I usually use (try parseDouble) <|> parseInt
21:45:45 <mroman> and parseDouble expects a "."
21:45:49 <oerjan> 12 and 12.0 cannot be distinguished without reading input
21:45:54 <mroman> so when no "." is present it fails and parseInt will parse it
21:46:01 <mroman> but that requires a try
21:46:53 <mroman> it looks ahead on symbol
21:47:10 <oerjan> it's LL(whateveryouneed)
21:47:35 <mroman> having a char 'a'; char 'b'; and a char 'a'; char 'c' wont work
21:47:53 <mroman> because once it sees an 'a' it will expect 'b' ALWAYS
21:47:59 -!- Patashu has quit (Remote host closed the connection).
21:48:18 <int-e> that's actually the main innovation of Parsec
21:48:20 -!- Patashu has joined.
21:49:24 <int-e> (Bounding backtracking so that input can be discarded even after encountering an alternative.)
21:49:44 <oerjan> mroman: you cannot combine an integer and a double parser without try, but you can certainly refactor integer <|> double into something that doesn't use try
21:50:10 <oerjan> it's a regular sublanguage, after all
21:55:06 -!- Tritonio has quit (Ping timeout: 252 seconds).
21:57:17 <fizzie> I tried our home internet out, and it was not there yet. You will be fungotless for two more weeks.
21:58:16 <oerjan> or well, lookahead might still be a problem. e.g. how do you parse 12e
21:58:30 <fizzie> (In retrospect, I should have probably just moved it on a different server, but I didn't expect a delivery time this long.)
22:04:07 -!- CADD has quit (Remote host closed the connection).
22:05:51 <fizzie> The state of not having a fungot.
22:06:16 <fizzie> (See, it's been so long everyone has already forgotten.)
22:06:41 <oerjan> i am not sure gamemanj was here when it started hth
22:06:58 <J_Arcane> http://www.theguardian.com/tv-and-radio/2015/feb/27/finnish-punk-band-take-punt-eurovision-title
22:08:29 <gamemanj> I don't appear much on IRC, so I wouldn't know what a fungot is...
22:09:01 <gamemanj> Judging by the channel topic, it likes the internet.
22:09:13 <int-e> fungot is a treasure trove of wisdom
22:09:27 <HackEgo> 10) <fungot> GregorR-L: i bet only you can prevent forest fires. basically, you know. \ 13) <fizzie after embedding some of his department research into fungot> Finally I have found some actually useful purpose for it. \ 14) <fungot> oerjan: are you a man, if there weren't evil in this kingdom to you! you shall find bekkler! executing program. plea
22:09:43 <HackEgo> 14) <fungot> oerjan: are you a man, if there weren't evil in this kingdom to you! you shall find bekkler! executing program. please let me go... put me out! he's really a tricycle! pass him!
22:10:11 <int-e> a bot, trained in linguistics. or something.
22:10:48 <int-e> (It can generate sentences using some Markov-ish language models.)
22:11:30 <gamemanj> ...Sounds fun....hang on, why can't fizzie just download fungot and connect from the same place he's IRCing from?
22:11:31 <int-e> The more impressive part is that it's written in Befunge.
22:11:51 * int-e is hazy on the fungoids.
22:11:54 <gamemanj> Or is fungot's code currently--WHAT.
22:12:22 <gamemanj> I know of the SOCK fingerprint, but still...
22:13:12 <fizzie> gamemanj: It felt like too much work to move everything (a gigabyte of data, plus having to build a patched cfunge) for such a should-have-been-short while.
22:13:13 <int-e> https://github.com/fis/fungot lists the fingerprints. (The code is there as well. Fizzie might give you a tour if you ask nicely, or not. I've never tried.)
22:13:29 <gamemanj> ...wait, really? A gigabyte of data? That's a bit of a large Befunge program.
22:13:44 <oerjan> i think that's the markov styles
22:13:53 <int-e> fizzie: patched how?
22:13:54 <fizzie> https://github.com/fis/fungot/blob/master/fungot.b98
22:13:57 <oerjan> the rest shouldn't be very large.
22:14:17 <oerjan> in theory there could be some huge ^def'ed commands, i guess
22:15:12 <int-e> (I didn't patch cfunge when I tried--unsuccessfully--to get fungot to run. Did I miss anything?)
22:15:31 <fizzie> int-e: There's a built-in chroot-and-drop-privs thing I wrote. Could go without or use some other (stronger) sandboxing, but haven't had the inclination.
22:16:14 <gamemanj> Well, what is the sandboxing needed for?
22:16:15 <fizzie> It should work just fine with the regular one. Though I may have not tried the most recent versions.
22:16:53 <fizzie> At any rate, I have no reasonable means of moving the data files out from here.
22:16:56 <oerjan> to protect against evil befunge hackers, of course
22:18:13 <gamemanj> How on earth did Fungot get so big in data?
22:18:31 <fizzie> It's the babbling styles.
22:19:03 <fizzie> Some of them were trained from relatively large source materials.
22:19:27 <fizzie> Like europarl, or even the irclog one.
22:19:52 <int-e> gamemanj: "Nonsense generation based on variable-length ngram models." -- there's a link to the variKN toolkit that fizzie used to generate the models (I think).
22:19:59 <gamemanj> ...And how are those loaded in?
22:20:18 <int-e> seek and destroy ... err, open, seek and read
22:20:24 <fizzie> Also the old (pre-varikn) training script did fixed-length unpruned models.
22:21:06 <fizzie> Yes, they're not loaded to fungespace, I seek/read with FILE on-demand.
22:21:30 <fizzie> Very inefficiently. Though I assume they do end up in caches.
22:21:30 <gamemanj> Sensible if they're a gigabyte.
22:23:17 <fizzie> `run quote 'department research' # is about the VariKN thing
22:23:18 <HackEgo> 13) <fizzie after embedding some of his department research into fungot> Finally I have found some actually useful purpose for it.
22:24:12 <fizzie> The biggest individual styles are (IIRC) 300 megs or so, but these days there are so many of them.
22:26:44 <fizzie> (Although many are real - too - small.)
22:28:34 <int-e> fizzie: can you get the fungot.dat file out?
22:30:33 <fizzie> Circuitously. But yes, in theory. I have something to do first, but I'll try to remember to put a link to it here within the next two hours.
22:30:39 -!- ProofTechnique has quit (Ping timeout: 245 seconds).
22:31:05 * FireFly isn't sure if "it" is the department research or fungot.
22:31:15 -!- bb010g has joined.
22:31:36 <fizzie> (Things would be easier if I'd just tether this mobile thing, but it's so explicitly against the terms of service.)
22:32:10 <int-e> FireFly: I read it as referring to the research
22:32:22 <gamemanj> I've made a "temporary" model so Fungot can be booted, although he only says "hello" with this model. And it's not much of a model...
22:32:32 <gamemanj> ...and I'm not sure if it'll work.
22:33:25 <fizzie> It should boot with no babbling models too, it just requires a fungot.dat miscellaneous data file.
22:34:34 <fizzie> int-e: I'm not sure if it wants a '\0'-terminated 'styles.list' file too, by the way.
22:34:52 <gamemanj> Oh. I just went ahead and read your file format documentation :)
22:35:02 <int-e> fizzie: I have that. I even have a style.
22:36:52 <gamemanj> ...Does styles.list have newlines or not? It's written as having "lines" which ALSO have \0 in them.
22:39:28 -!- fnordbot has joined.
22:39:38 <fnordbot> int-e: hitch the length, he rather suspected this. trillian had turned aside in disgust they all went, " yuch." it's a lot. " so," said ford, " you reckon, what's the problem? a sudden silence afterwards." you can tell. at the moment. a slight hiss built into a deafening roar the engines. a galactic civilization, leapt to the controls, " i didn't pick up hitch the length, he rather suspected this. trillian had turned aside in d
22:39:47 -!- ProofTechnique has joined.
22:39:49 <int-e> fizzie: ok, it was the cfunge version
22:40:13 <int-e> (and the language model doesn't know when to stop)
22:40:18 <oerjan> i see it starts with a loop
22:41:08 <fnordbot> Selected style: Not really what you'd expect
22:41:34 <int-e> gamemanj: so as far as I can see it's actual lines, just as described.
22:41:52 <fizzie> gamemanj: It needs both \0\n at the ends of lines, and a \0 in first column after the entries.
22:42:25 <int-e> fizzie: so I guess I won't need that state file after all. but thanks for the offer.
22:42:32 <fizzie> Er, that is, \0 in first column on the line after the entries.
22:42:44 <gamemanj> Should styles.list and the model files be in data\ ?
22:43:06 <fizzie> int-e: Good, less work for me. Though you are missing the defined commands without it.
22:43:33 <fizzie> gamemanj: I think not. Which is a bit silly.
22:43:59 -!- ProofTechnique has quit (Ping timeout: 246 seconds).
22:44:22 <fizzie> fnordbot: So which IRC channel was that from?
22:44:22 <fnordbot> fizzie: for. " all the doors." " a what? far out in the dordellis wars, and the megafreighters had to be equipped with the most fantastic. in a few seconds," he continued, " you just let the machines began to think that the chances of getting picked. he was a hell of a lot. " so," said ford, " you reckon, what's the problem? all his heirs are now long dead, and this is frankie, " a simple one!" " i just don't say things like th
22:44:25 <int-e> fizzie: right. but mainly I was curious whether I could get it to run at all. I didn't expect cfunge 0.9 to be incompatible with 0.3.3 (I haven't tried 0.4.1)
22:44:59 <int-e> fizzie: none. it's trained from the first volume of hitchhiker
22:45:23 <gamemanj> ...I'm guessing it doesn't work with CCBI?
22:46:11 <fizzie> int-e: Bizarre. I'm pretty sure I've ran it post-0.9. I'd like to debug this out, but the circumstances are not good for it right now. Perhaps when I get back from the Americas.
22:46:53 <fizzie> int-e: Feel free to report a github issue. :p (Might even be the first one.)
22:50:33 -!- ProofTechnique has joined.
22:53:20 <gamemanj> Well, after finding out I messed up the config, CCBI starts fungot! bad news: it doesn't give a nickname.
22:54:12 <int-e> did you edit the fungot-load-whatever.b98 file?
22:54:30 <gamemanj> Yep, that's what I broke(and then fixed, and then it still failed)
22:56:07 <gamemanj> Seems I didn't mess up config, and "-" isn't a banned letter in IRC AFAIK.
22:56:17 <gamemanj> (didn't mess it up twice, that is)
22:59:49 -!- fnordbot has quit (Remote host closed the connection).
23:03:58 <gamemanj> Fizzle, are the bot owner details still valid?
23:04:59 -!- Tritonio has joined.
23:05:36 <int-e> gamemanj: did you align the v above the strings with the < at the end of the lines, and make sure that there's no strage letter inbetween?
23:06:04 <gamemanj> Well, that was what my first mistake was. I've fixed that now, but I'm using CCBI, so I'm trying with the right version of cfunge.
23:06:05 -!- AnotherTest has quit (Ping timeout: 252 seconds).
23:06:49 <gamemanj> I'm probably going to lose connection for the night soon, though.
23:10:24 <gamemanj> ...cmake is starting, I'm compiling cfunge on the target server(which has a mostly-reliable connection)
23:10:46 <gamemanj> (and more importantly can run cfunge)
23:11:36 <fizzie> int-e: Did you test out the bf/ul interpreters while it was online?
23:12:01 <gamemanj> I didn't see any testing(could be done via PM)
23:12:53 <fizzie> (I went and got the Bluetooth keyboard, this seems so involved.)
23:13:06 -!- fnordbot has joined.
23:13:28 <fizzie> As for the owner details, you're supposed to fill them with your own, and then it'll accept the "owner" commands only from that nick!user@host prefix.
23:13:30 <gamemanj> Cfunge built on Raspberry Pi B+, who should own the bot?
23:13:49 <int-e> oh right, underload.
23:14:46 <int-e> ^bf ++++[>++++++++<-]>+.+.+.+.+.+.
23:15:08 <gamemanj> Starting...Still "no nickname given". Re-re-checking config.
23:16:43 -!- oerjan has quit (Quit: ZZZ).
23:18:22 <Koen_> ^befunge <@_.<0"hello"
23:18:26 <int-e> what's that funge testsuite...
23:19:26 <Koen_> ^funge <@_.<0"hello"
23:19:28 <fizzie> Koen: It only does the two.
23:20:02 <Koen_> I guess an inline interpreter for a 2d language wasn't that useful anyway
23:20:07 <fizzie> Although there's an owner-only command to run "raw" Funge-98 code, but it's mostly just useful for hot-patching (more like hot-breaking).
23:20:53 <fizzie> I can't get a hat out of this keyboard.
23:21:27 <gamemanj> fizzie: "fungot-g" is under your control. Have fun with it until you can run real fungot again. Don't abuse that feature I didn't know about until you mentioned it.
23:21:29 <Koen_> I can't get a bakslash out of mine
23:21:46 <fizzie> It has the Finnish keyboard, so the key with all kinds of diacritics is dead. Normally you get the ASCII hat with that + space, but on Android it just gives a ˆ.
23:21:57 <fizzie> Which is some other kind of a hat.
23:22:24 <Koen_> do you know if i can plug my phone to the computer and type with the phone keyboard?
23:23:14 <gamemanj> fizzie:try the brainf*** code: ++++++++++[->+++++++++<]>++++.[-]<
23:23:35 <fizzie> Koen_: I've done something like that by using a SSH client on the phone, and then doing screen -x to attach to a screen on the computer.
23:24:07 -!- Fleur has quit (Read error: Connection reset by peer).
23:24:15 <fizzie> Koen_: Though that didn't involve any plugging in. I would be surprised if it could (at least out-of-the-box) pretend to be a USB keyboard or something, if that's what you mean.
23:24:15 <gamemanj> Koen_:RDP,VNC are capable, but why type with a phone keyboard...?
23:24:29 <Sgeo> https://www.reddit.com/r/netsec/comments/2xgypm/using_rubys_openuri_one_step_away_from_rce/
23:24:53 <Koen_> my right hand is in a cast and has been for over two weks
23:25:21 <Koen_> i've practised typing with three fingers from the left hand but that'sstill slow
23:25:37 <Koen_> and some chars are hard to get
23:25:53 <fizzie> ^ul (Thanks for the copy-hat.)S
23:25:58 <Koen_> the backslash is downright impossible, it's alt+shift+/
23:26:32 <Koen_> and the alt key is only on the left side of the keyboard
23:27:06 <int-e> fizzie: so 0.3.3 is the latest cfunge that works for me.
23:28:00 <fizzie> int-e: And what did you say happened on the newer versions? Infinite loop parsing fungot.dat?
23:28:01 <gamemanj> What happens when there's two fungots on the same channel?
23:28:24 <fizzie> The command character (^) is in the configuration, I think.
23:28:31 <fizzie> You can have two non-conflicting fungots.
23:29:30 <gamemanj> So, if you had to choose a command character(under this scenario) what would it be?
23:30:29 <int-e> fizzie: it's playing ping-pong: http://dpaste.com/26CHQMY
23:31:28 <fizzie> I guess that wasn't the ASCII backtick either.
23:31:36 <fizzie> Man, this keyboard layout sucks.
23:31:37 <int-e> so yes that looks like the code reading fungot.dat (line 15 of fungot.b98)
23:32:20 <gamemanj> ...Actually, let me check...Fizzie, it *is* the ASCII backtick.
23:33:38 <gamemanj> At least, as far as I can tell from this web client, anyway.
23:34:57 <fizzie> int-e: And what was your fungot.dat like again? I keep forgetting.
23:35:01 <HackEgo> [U+02CB MODIFIER LETTER GRAVE ACCENT]
23:35:13 <int-e> fizzie: just 10 newlines
23:37:04 <HackEgo> U+0060 GRAVE ACCENT \ UTF-8: 60 UTF-16BE: 0060 Decimal: ` \ ` \ Category: Sk (Symbol, Modifier) \ Bidi: ON (Other Neutrals)
23:37:20 <FireFly> <gamemanj> So, if you had to choose a command character(under this scenario) what would it be? ← It occurred to me to check the list of botprefixes, but then I realised it's provided by fungot...
23:37:59 <fizzie> int-e: Okay, I have a hypothesis.
23:38:25 <fizzie> int-e: It has to do with the way STRN's substring things got more exact specifications for edge cases.
23:38:42 <gamemanj> FireFly: Well, fnordbot is fungot, but with the model database "modified".
23:38:42 <fnordbot> gamemanj:. " you see," he shouted to the guard, " not really. " we had a look at this," said slartibartfast, " that was one of mine," shouted ford. " i don't want to go to work at," and he turned. " but who the man with the five heads all the tests, learning to distinguish between him pretending to be an out a wild whoop in major thirds, threw ford prefect, " it's dark," he shouted to the guard, " not really. " we had a look at
23:39:17 <FireFly> gamemanj: sure, but it was stored as a defined command, so it's in the fabled fungot.dat file I suppose
23:39:55 <fizzie> int-e: Try with 10 non-empty lines.
23:40:36 -!- Koen_ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
23:42:33 <fizzie> int-e: (I think the problem is that the newline-stripping code no longer works on an empty line. Maybe.)
23:42:39 <int-e> fizzie: that gets past parsing the fungot.dat file, apparently, but loops later on, printing lots of numbers.
23:45:14 <fizzie> Maybe I should still just give you the .dat file.
23:45:58 <int-e> it gets to the point where it connects to the server, at least.
23:46:07 -!- gamemanj has quit (Ping timeout: 246 seconds).
23:46:49 <fizzie> FireFly: "All" the bots provide the list of botprefixes, that's one of the points of it.
23:47:02 <fizzie> FireFly: That's no longer quite true, but I think at least HackEgo's got it.
23:47:18 <fizzie> FireFly: (That's what I was trying to do there, with the backtick.)
23:47:30 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: botnames: not found
23:47:38 <fizzie> It's "prefixes". I think.
23:47:53 <HackEgo> Bot prefixes: fungot ^, HackEgo `, EgoBot !, lambdabot @ or ?, thutubot +, metasepia ~, idris-bot ( , jconn ) , blsqbot !
23:48:06 <EgoBot> Bot prefixes: fungot ^, HackEgo `, EgoBot !, lambdabot @ or ?, thutubot +, metasepia ~, idris-bot ( , jconn ) , blsqbot !
23:48:35 <fizzie> That might be it. I don't think lambdabot is #esoteric-specific enough for that. :)
23:48:39 <int-e> fizzie: http://dpaste.com/3ZG5D0Z ... playing pingpong with the very top and bottom of funge-space.
23:49:03 <FireFly> Rather than having bots ignore other bots, I think other bots should adopt HackEgo's ZWSP prefix, and then ignore incoming lines starting with ZWSP
23:49:36 <int-e> fizzie: but how does a space go from 22/192 to 3/192... that's wild.
23:49:41 <FireFly> (well, I suppose specifically ignoring lines starting with ZWSP is redundant if it only takes action on a certain prefix)
23:50:16 <FireFly> int-e: I think it might skip until the next non-space characer
23:50:31 <fizzie> Yes, spaces are coalesced in Funge-98.
23:51:05 <fizzie> Sounds like something that might be a bonafide fungot bug.
23:51:19 <fizzie> Or maybe just another "unexpected reflect".
23:51:32 <int-e> anyway, these are the last few instructions before it veers off like that: http://dpaste.com/0HYK3CC
23:52:46 <int-e> fizzie: if you give me the .dat file I can put it up with fnordbot.
23:52:47 <fnordbot> int-e: what a day. " no you can't possibly, let out a wild whoop in major thirds, threw ford prefect, " it's dark," he shouted to the guard, " not really. " we had a look at this," said slartibartfast, " that was one of mine," he was saying, " magrathea is a myth, a fairy story, it's what parents tell their kids at school nicknamed him ix, which in the language. you want to try a guess at all, ford?" " well you see," he shouted
23:54:12 <fizzie> int-e: I think I'll do that. I'd like to figure out what's wrong for realsies, but I don't think I'll be doing that on this phone.
23:57:07 <fizzie> Okay, file is on the phone, thanks to adb push.
23:57:15 <fizzie> Now I just need to stuff it somewhere.
23:57:23 <fizzie> I don't have a Dropbox client on this thing.
23:57:33 <fizzie> Maybe you can share a file from Google Drive or something.
23:58:42 * int-e would be lost, and try to look for an ssh/sftp client...
23:59:05 <fizzie> I had one, but it was not a good one.
23:59:24 <fizzie> You could try what happens if you click at https://docs.google.com/file/d/0B4J9OAzXNfZAX2Z6OW15YXZGaWM/edit?usp=docslist_api