00:02:34 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
00:09:02 <int-e> I'm trying to get to 2^20 trials, then pick those parameters and paste :)
00:09:35 <shachaf> How much state do your prisoners have?
00:10:59 <int-e> one is unbounded, but it shouldn't hurt much if it gets truncated at some constant multiple of N anyway.
00:14:22 <rntz> two counters? interesting
00:14:43 -!- adu has joined.
00:16:23 <int-e> and we have a winner, http://sprunge.us/JgAh
00:16:47 -!- zzo38 has joined.
00:18:27 <int-e> (there's further room for tweaking but it'll become messy)
00:19:15 -!- oerjan has joined.
00:21:45 <rntz> what is the purpose of `dark'?
00:22:11 <int-e> it started out as a counter that counted how often the room was dark since the last light...
00:22:58 <int-e> it's still related to that, but basically it acts as a weight that determines how much the prisoner acts as a count.
00:23:14 <int-e> I have no good intuition for how this works!
00:23:17 <rntz> do you have a proof that this never gets stuck?
00:23:32 <rntz> (given that it completed 1048576 runs w/o getting stuck, it seems like it must be very unlikely)
00:24:05 <int-e> as long as the light bulb is off, all 'dark' counters increase, so eventually one of the tokens > 1+dark tests must fail.
00:24:15 <int-e> so tokens keep moving and it'll terminate with probability 1.
00:24:50 <int-e> handwaving, but that's one aspect I do understand well enough to be confident about it
00:24:51 -!- Deewiant has quit (Ping timeout: 244 seconds).
00:25:09 -!- Elronnd has quit (Ping timeout: 260 seconds).
00:26:58 <rntz> I don't do your dark-tracking-to-determine-whether-I-count stuff; we share two optimizations; and I have one optimization you don't
00:27:00 -!- ais523 has quit.
00:27:12 <int-e> actually there's a second aspect to it that's more subtle: as long as all dark counters are big, the same token will be passed on and on until one of the counters becomes small enough to be kept.
00:28:08 <int-e> so not all counters can become large.
00:29:30 <rntz> oh, hm, oh, I totally misunderstood your strategy
00:29:48 <rntz> so depending on your dark-counter you either give try to tokens away or to accept them?
00:30:30 -!- Deewiant has joined.
00:30:32 -!- shikhin has quit (Quit: Alas.).
00:30:32 -!- hydraz has quit (Quit: Bai.).
00:31:31 -!- feliks has joined.
00:31:32 <rntz> hm. how do you know this can't somehow endlessly slosh back and forth between two prisoners?
00:32:18 -!- Phantom_Hoover has quit (Quit: Leaving).
00:32:22 <rntz> hm, I suppose the randomness of the warden prevents that
00:32:28 <int-e> oh it definitely can do that. but it's also possible that some prisoner will never be called to the room
00:32:49 <int-e> so I thought probability 1 assuming a random warden is good enough.
00:32:49 <rntz> yeah, it's an unavoidable problem :P
00:33:01 <rntz> probability 1 assuming random warden was my intended interpretation
00:33:05 -!- Elronnd has joined.
00:38:50 -!- adu has quit (Quit: adu).
00:44:10 -!- zzo38 has quit (Ping timeout: 250 seconds).
00:48:26 <int-e> right, there's another thing... good for shaving off about 600 rounds
00:50:28 -!- hppavilion[1] has joined.
00:55:39 -!- Hooloo42 has quit (Ping timeout: 264 seconds).
00:58:45 <int-e> rntz: so final version for tonight: http://sprunge.us/OWPH
01:01:09 -!- Hoolootwo has joined.
01:04:32 -!- shikhin has joined.
01:04:51 -!- devhydraz has joined.
01:04:52 -!- devhydraz has quit (Changing host).
01:04:52 -!- devhydraz has joined.
01:05:38 -!- devhydraz has changed nick to hydraz.
01:07:01 <rntz> int-e: aha, you found my optimization, I think
01:07:10 <rntz> (the dropped/accepting code)
01:07:52 <int-e> or, perhaps, remembered...
01:08:10 <shachaf> int-e: Did you publish a paper about this or something?
01:08:35 <shachaf> You're referenced in https://www.math.washington.edu/~morrow/336_11/papers/yisong.pdf
01:10:00 <int-e> It's not a paper; the reference is from the Dehaye/Ford/Segerman paper, "Newsgroup rec.puzzles, available through http://groups.google.com, July 2002"
01:11:11 <shachaf> I suppose it's https://groups.google.com/d/msg/rec.puzzles/zkzEEfzqtis/w3pR8g1TJmEJ
01:11:47 <FireFly> Huh. That sounds like an interesting problem
01:12:13 <int-e> rntz: anyway I'm fairly happy now... it's getting quite close to the 10418 number that you get by assigning a dedicated count.
01:12:24 -!- xkapastel has joined.
01:13:10 <int-e> rntz: also I'm interested in your strategy now
01:16:48 -!- zzo38 has joined.
01:17:02 <int-e> shachaf: that looks right
01:17:38 <zzo38> What is wrong with it?
01:18:35 <olsner> <ais523> I like 1040 ← yes, it is very truthy
01:19:00 -!- izabera has joined.
01:20:09 -!- zgrep has joined.
01:21:55 <shachaf> olsner: the deadline for 1040 was apr 15 hth
01:24:02 <int-e> oh well, why shouldn't we all know about US tax forms...
01:26:27 -!- wibblewobble has joined.
01:26:33 -!- wibblewobble has quit (Remote host closed the connection).
01:34:03 -!- ybden has quit (Quit: Lost terminal).
01:38:22 -!- wibblewobble has joined.
01:39:00 -!- wibblewobble has quit (Remote host closed the connection).
01:43:11 -!- wibblewobble has joined.
01:43:19 -!- wibblewobble has quit (Remote host closed the connection).
01:50:54 -!- zzo38 has quit (Ping timeout: 250 seconds).
02:01:55 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
02:11:11 -!- Akaibu has quit (Quit: Connection closed for inactivity).
02:19:32 -!- wibblewobble has joined.
02:19:42 -!- wibblewobble has quit (Remote host closed the connection).
02:20:26 <oerjan> <copumpkin> you were way ahead of your time <-- if it's the story i think, this is quite appropriate hth
02:21:03 -!- wibblewobble has joined.
02:21:21 -!- wibblewobble has quit (Remote host closed the connection).
02:21:32 -!- ChanServ has set channel mode: +o oerjan.
02:22:03 <shachaf> oerjan: are you thinking of _Story of Your Life_
02:22:26 -!- oerjan has set channel mode: +b *!~wibblewob@yakamo.org$##fixyourconnection.
02:22:44 <shachaf> looks like someone wibble-wobbled a little too much
02:22:53 -!- oerjan has set channel mode: -o oerjan.
02:24:08 <oerjan> shachaf: i was assuming that one story would be included in the book.
02:35:17 <shachaf> oerjan: Have you read his other stories?
02:57:45 <shachaf> oerjan: well maybe i should send you a copy of the book instead hth
03:00:39 -!- adu has joined.
03:04:16 <shachaf> copumpkin: Oh, "Soon to be a major motion picture". That's why there's a new edition.
03:04:48 <shachaf> copumpkin: so you'd better hurry up and read it if you want to stay ahead of the hipster curve hth
03:15:48 <shachaf> _Stories of Your Life and Others_ by Ted Chiang
03:19:11 * oerjan hits shachaf with the saucepan ===\__/
03:19:32 <shachaf> too bad you use your saucepan for hitting rather than for sauce
03:22:07 <shachaf> You know, I always figured the saucepan was a skillet.
03:22:15 <shachaf> But I guess it's actually a saucepan.
03:22:43 <shachaf> I guess I thought that because of your skill at hitting people with it.
03:40:10 -!- idris-bot has quit (Ping timeout: 244 seconds).
03:40:55 -!- Melvar has quit (Ping timeout: 252 seconds).
03:43:29 -!- Melvar has joined.
03:44:11 -!- adu has quit (Quit: adu).
03:47:09 -!- zzo38 has joined.
03:49:34 <zzo38> Hello, somehow there seem to be problem with this internet today and restarting the router fixed it, hopefully it is OK by now.
03:50:05 <zzo38> The ping timeout did display on my computer though before it stopped; I did not get a connection error.
03:54:05 -!- adu has joined.
04:07:24 -!- primo has joined.
04:17:15 -!- hppavilion[1] has joined.
04:29:01 -!- adu has quit (Quit: adu).
05:34:52 <zzo38> I am looking at the EM computer architecture it looks good so far
05:38:36 <zzo38> I don't know what the letters stand for.
05:39:23 <zzo38> It is described in a report titled "DESCRIPTION OF A MACHINE ARCHITECTURE FOR USE WITH BLOCK STRUCTURED LANGUAGES".
05:46:52 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
05:49:08 <zzo38> It is a virtual machine code and does not require any particular floating point representation.
06:14:33 -!- incomprehensibly has quit (Ping timeout: 240 seconds).
06:15:21 -!- incomprehensibly has joined.
06:15:33 -!- sebbu has quit (Ping timeout: 240 seconds).
06:18:27 <oerjan> (apparently making up finnish-sounding words works retroactively, who knew)
06:19:19 -!- primo has left ("Leaving").
06:21:12 -!- copumpkin has quit (Ping timeout: 276 seconds).
06:24:09 -!- copumpkin has joined.
06:59:12 -!- oerjan has quit (Quit: Nite).
07:15:14 -!- ocharles_ has quit (Ping timeout: 260 seconds).
07:18:03 -!- ocharles_ has joined.
07:24:21 -!- hppavilion[1] has joined.
07:26:21 -!- yorick has quit (Quit: No Ping reply in 180 seconds.).
07:27:36 -!- yorick has joined.
07:27:36 -!- yorick has quit (Changing host).
07:27:36 -!- yorick has joined.
07:32:50 -!- Kaynato has quit (Ping timeout: 260 seconds).
07:57:45 <HackEgo> Confound \ 1U \ Instant \ Counter target spell that targets one or more creatures. \ Draw a card. \ PS-C
07:57:45 <HackEgo> Intervene \ U \ Instant \ Counter target spell that targets a creature. \ UL-C
08:06:14 -!- lambda-11235 has quit (Quit: Bye).
08:12:01 -!- sebbu has joined.
08:15:39 -!- xkapastel has quit (Quit: Connection closed for inactivity).
08:39:42 -!- zadock has joined.
08:40:37 -!- zadock has quit (Remote host closed the connection).
08:48:55 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
09:01:38 -!- infinitymaster has joined.
09:02:16 -!- infinitymaster has quit (Client Quit).
09:13:24 -!- caradow has changed nick to carado.
09:19:24 <int-e> hah, I missed a key optimization yesterday... now I get fewer than 11200 rounds for the prisoners.
09:19:59 <int-e> (bragging was urgent... details later :-P)
10:10:09 -!- AnotherTest has joined.
10:18:31 -!- AnotherTest has quit (Ping timeout: 244 seconds).
10:19:33 -!- hppavilion[1] has joined.
10:42:14 <int-e> rntz, shachaf: http://sprunge.us/UETC ... I also added comments.
11:02:01 -!- hppavilion[1] has quit (Ping timeout: 252 seconds).
11:05:12 -!- boily has joined.
11:14:50 <rntz> ah, if somebody's already been to the room, you can become a donor
11:26:54 -!- ybden has joined.
11:56:33 -!- atriq has quit (Disconnected by services).
11:56:42 -!- Taneb has joined.
12:14:13 <boily> Tanelle. how many exams today?
12:15:38 -!- xkapastel has joined.
12:19:11 <boily> `relcome xkapastel
12:19:29 <HackEgo> xkapastel: 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.)
12:23:31 -!- boily has quit (Quit: TRIANGULAR CHICKEN).
13:46:43 -!- Deewiant has quit (Ping timeout: 244 seconds).
13:47:58 -!- Deewiant has joined.
14:02:49 -!- Akaibu has joined.
14:15:41 -!- Sgeo has quit (Ping timeout: 240 seconds).
14:42:54 <b_jonas> zzo38: in M:tG, do you happen to know why Intervene and Confound are templated differently? As far as I can tell, the first abilities are functionally identical and there's no reason for them to have different wording.
14:51:10 -!- AnotherTest has joined.
14:58:46 -!- `^_^v has joined.
14:59:28 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: xkcdwhatiflist: not found
15:01:38 -!- AnotherTest has quit (Ping timeout: 244 seconds).
15:32:31 <ybden> `? magic: the gathering
15:32:33 <HackEgo> magic: the gathering? ¯\(°_o)/¯
15:32:43 <ybden> Is this bot case-sensitive?
15:36:24 <fizzie> I don't think bin/? is.
15:37:20 <HackEgo> The magic was in you all along.
15:37:23 <fizzie> Looks case-insensitive to me.
15:42:06 -!- gamemanj has joined.
15:42:37 -!- izabera has quit (*.net *.split).
15:42:37 -!- hydraz has quit (*.net *.split).
15:42:37 -!- myndzi has quit (*.net *.split).
15:42:37 -!- mtve has quit (*.net *.split).
15:53:32 -!- spiette has joined.
16:22:48 -!- lambda-11235 has joined.
16:40:18 -!- izabera has joined.
16:40:18 -!- hydraz has joined.
16:40:18 -!- myndzi has joined.
16:40:18 -!- mtve has joined.
16:53:33 -!- f10d4 has quit (Ping timeout: 240 seconds).
16:57:07 <lambda-11235> I must realy like brainfuck, as I've written at least 3 compilers and 2 interpreters for it over 5 different PLs.
16:57:49 <gamemanj> brainfuck is like othello: it's simple to compile it, but to master compiling it is difficult...
16:59:18 <izabera> but can you actually write programs in it?
17:04:53 <lambda-11235> izabera: Someone wrote a compiler in it, so it's possible. https://github.com/matslina/awib
17:10:58 <lambda-11235> izabera: Depends on how complicated the program is.
17:11:36 <izabera> a program that doubles a number i entered in base 10
17:15:34 <lambda-11235> izabera: Only if I add a compiler extension, but then it wouldn't follow the standard.
17:16:24 -!- f10d4 has joined.
17:17:53 <gamemanj> just pipe the output of the program to /dev/coffeemachine
17:18:40 <lambda-11235> gamemanj: But then you'd be programming in some derivative of sh.
17:19:13 <gamemanj> Well, sh would simply connect the programs together. Pipes themselves are language-independant
17:19:44 <gamemanj> so all sh would be doing is initializing the environment
17:20:00 <gamemanj> and if initializing the environment counts,
17:20:10 <gamemanj> then as the console is also given to the program by sh...
17:20:53 <gamemanj> In which case the environment is "initialized" by... the linux kernel.
17:21:33 <gamemanj> Or the OpenBSD kernel, or the FreeBSD kernel, or... you get the point.
17:22:11 <gamemanj> You can't even win with bare-metal assembly - the environment is initialized by the processor and/or BIOS.
17:22:50 <gamemanj> FPGAs? The programmer & circuit board. A bunch of transistors? The arrangement counts as an initialized environment...
17:23:31 <gamemanj> lambda-11235: What you're programming in is relative.
17:29:36 <lambda-11235> gamemanj: Unfortunately I don't have /dev/coffee installed, or a coffee machine for that matter.
17:30:30 <gamemanj> In which case, the nature of pipes means you can simply swap out /dev/coffee for the NetHack /dev/coffee Implementation Patch.
17:30:43 <gamemanj> (Built using CUSE, of course.)
17:34:35 <lambda-11235> Hmm, I wonder how serious github is about letting me have an unlimited number of repos.
17:43:45 <shachaf> oerjan: The magic was in you all along?
17:46:26 <gamemanj> b_jonas: nice work on the zzo38 computer spec, BTW :)
17:55:39 -!- gamemanj has quit (Ping timeout: 264 seconds).
18:09:15 -!- gamemanj has joined.
18:58:43 <zzo38> b_jonas: It looks like identical to me too and I don't know. Possibly due to historical reasons and the text hasn't been updated because it does not need to be.
18:59:16 <gamemanj> how is it I keep only hearing half of conversations?
19:05:21 <zzo38> gamemanj: Possibly the messages are far apart and you did not read it
19:05:33 <zzo38> (It happens to me too of course)
19:06:26 <gamemanj> (I checked logs during my a-few-minutes downtime
19:07:02 <gamemanj> Since 15:42:07, there have been 0 messages from b_jonas... so I should probably read more logs.
19:40:22 -!- Phantom_Hoover has joined.
19:43:55 <int-e> rntz: huh, the "5 is experimentally best" claim is wrong... it's 6 that gives the average quoted at the top... must have mised up my experiments.
19:47:51 -!- ybden has quit (Ping timeout: 264 seconds).
19:48:31 -!- ybden has joined.
19:56:33 -!- augur has quit (Quit: Leaving...).
19:57:07 -!- Kaynato has joined.
20:14:42 -!- ybden has quit (Ping timeout: 272 seconds).
20:15:20 -!- ybden has joined.
20:45:18 -!- augur has joined.
21:52:23 -!- gamemanj has quit (Ping timeout: 244 seconds).
22:08:19 -!- AnotherTest has joined.
22:14:40 -!- `^_^v has quit (Quit: This computer has gone to sleep).
22:27:33 -!- AnotherTest has quit (Ping timeout: 240 seconds).
22:38:44 -!- moon_ has joined.
22:41:35 <HackEgo> The U are a very mad people.
22:41:43 <int-e> `culprits wisdom/u
22:41:52 <HackEgo> oerjan FreeFull shachaf shachaf nitia
22:42:04 <HackEgo> changeset: 0:e037173e0012 \ user: HackBot \ date: Thu Feb 16 19:42:32 2012 +0000 \ summary: Initial import.
22:43:27 -!- centrinia has joined.
22:44:43 <moon_> I had the idea of making a esolang where the only form of data manipulation is by 'folding' and 'unfolding' the data tape to increment and decrement values
22:45:29 <moon_> each fold goes along the tape until it reaches a 'seam' which it stops at
22:45:49 <moon_> The direction of a fold depends on the direction of the instruction pointer
22:46:00 <moon_> which can be either left or right
22:46:32 <int-e> Origami Tape, hmm.
22:47:37 <moon_> im thinking about adding velocitys (and overshooting the seam) to the mix
22:47:41 <moon_> but i dont think i should
22:47:49 -!- spiette has quit (Quit: :qa!).
22:49:59 <moon_> what do you think int-e?
22:52:09 <int-e> that question is too philosophical for me right now.
22:52:38 <moon_> name is simple too :P 'fold'
22:53:29 <moon_> actually, even better: make it 3d
22:54:03 <moon_> Actually, ill stick to 1d for ease of implentation, i might try 2d later
23:15:18 -!- hppavilion[1] has joined.
23:15:34 <moon_> the wiki wont let me publish
23:15:53 <moon_> can somene publish this for me? http://pastebin.com/zsknMxvN
23:18:48 <moon_> can you publish that for me?
23:20:24 <moon_> yea, the wiki is refusing to publish
23:20:44 <fizzie> Could you unpack that a little.
23:20:58 <fizzie> As in, what is it doing instead.
23:21:03 -!- Sgeo has joined.
23:21:36 <moon_> showing me the preview, and i am a hundred precent certain i have hit the publish button and not the preview one, for 20 times in a row now
23:21:55 <fizzie> Does it have any sort of an error message anywhere on the page?
23:23:56 <moon_> apparently my cookies are off and i never, ever turn them off
23:24:01 <moon_> something is up here
23:24:36 <fizzie> It did say that to me as well.
23:24:55 <fizzie> I vaguely recall it doing something similar before. Possibly something to do with the session storage.
23:25:36 <moon_> anyways.. i came up with the idea for a language that uses 'folds' in the tape to represent data
23:25:46 <fizzie> memcached has been killed by the OOM killer.
23:26:03 <fizzie> total used free shared buffers cached
23:26:03 <fizzie> Mem: 2010 1910 99 43 29 120
23:26:26 <moon_> and we dont have gregor to fix it, he owns the wiki, right
23:26:48 <fizzie> I mean, it's his machine, I do have administrative rights on it.
23:27:05 <moon_> So you can fix it? maybe?
23:27:17 <fizzie> I wonder if it's all those python zombies HackEgo keeps making up these days.
23:27:46 <moon_> you should write a python antizombie *jokes yet not jokes*
23:28:13 <moon_> wait.. python zombies?
23:28:28 <moon_> that means i might be able to access the external system using something like that..
23:30:17 <moon_> the site stopped rendering right on my end
23:30:25 <fizzie> I restarted a couple of things.
23:30:51 <fizzie> Logging in works for me now.
23:31:24 <HackEgo> [wiki] [[Fold]] N https://esolangs.org/w/index.php?oldid=47059 * Moon * (+969) Created
23:33:11 <moon_> also, because you have admin rights, dont you have access to hackego?
23:33:22 <moon_> *noting you could install GHC for hackego again*
23:34:37 <fizzie> Technically, but I try to avoid touching it as much as I can, since it's not really mine.
23:35:40 <fizzie> The zombie thing is strange, I used to think it was recent, but looking at it, it doesn't look like multibot ever waited for its children (or daemonized either).
23:36:14 <moon_> i could cause out of sync processes that way
23:36:26 <fizzie> That doesn't really follow.
23:36:47 <fizzie> I mean, sure, it could be a denial-of-service-ish thing, clogging up the process table.
23:37:05 <moon_> Arn't the processes being ran?
23:37:29 <fizzie> A zombie is a process that's died but its corpse just hasn't been collected by the parent.
23:38:07 <moon_> Well would it be possible to make a process that lives beyond the time of the main one?
23:38:42 <moon_> Does hackego collect threads?
23:40:12 <int-e> it collects the whole kernel
23:40:35 <HackEgo> Linux umlbox 3.13.0-umlbox #1 Wed Jan 29 12:56:45 UTC 2014 x86_64 GNU/Linux
23:40:36 <fizzie> There's a lot of sandboxing and timeouting in the actual HackEgo stuff, and as far as I know they work pretty reliably, always terminating in a reasonable time. It's just that, AFAICT, the multibot core doesn't clean up the trash.
23:40:48 <fizzie> (Or dissassociate itself from it.)
23:41:31 <moon_> Int-e wouldnt it be nice if hackego had GHC?
23:41:33 <fizzie> Of course that's the part written in C, so I'm even less inclined to start patching it. Since I have no idea whether the binary running there is actually from the public source or not.
23:42:07 <moon_> Its not, as far as i know, i have not seen the core in the bitbucket repository
23:42:19 <fizzie> Oh, that's there -- it's just a different repo.
23:42:23 <fizzie> https://bitbucket.org/GregorR/multibot/src
23:42:49 <moon_> I'd happily help patch it then :P
23:42:52 <fizzie> I just don't know if the version HackEgo's running has been customized.
23:44:21 <moon_> Oh wait, i dont run linux
23:45:29 <moon_> the cmd execution of the multibot inthe bitbucket lacks sandboxing
23:45:50 <fizzie> No, that's yet another repository.
23:46:08 <moon_> No, it doesnt use the umlbox command for executing commands
23:46:32 <moon_> it just 'exec' s it
23:46:39 <fizzie> Yes, because it's not *multibot* that does it.
23:46:56 <fizzie> That's done by the *hackbot* code.
23:46:57 <int-e> fizzie: you may want http://codu.org/projects/egobot/hg/
23:47:16 <int-e> hmm, wait, that's also the wrong one...
23:47:39 <fizzie> https://bitbucket.org/GregorR/hackbot
23:47:43 <int-e> yeah, https://bitbucket.org/GregorR/hackbot
23:48:07 <shachaf> What do you think of the "everything in one repository" approach?
23:49:10 <fizzie> It scales up to a Google.
23:49:32 <int-e> define "one repository
23:50:23 <shachaf> As in one hg or Perforce or whatever repository.
23:50:44 <moon_> Im thinking about writing a basic text editor for hackego
23:51:19 <shachaf> Is there a way to keep state between lines of HackEgo without committing to hg?
23:52:08 <int-e> I may have picked too broad a context for "everything".
23:52:25 <shachaf> I mean all your code and so on.
23:52:28 <fizzie> shachaf: Not that I know of.
23:52:46 <int-e> Well I don't want to have a tangled mess of all my projects.
23:52:59 <shachaf> Well, you put each one in its own directory. So it's not a mess.
23:53:12 <int-e> Then I can have a repository for each.
23:53:43 <moon_> fizzie, what should i set the max file size input to? (aka how much memory it devotes to the file)
23:54:25 <izabera> what if you need something that scales up to two googles
23:54:37 <moon_> I dont think hackego has that much ram
23:54:58 <int-e> (There is a limitation of the RCSs that I'm aware of... they don't offer nice views ala "what's the history of this subdirectory")
23:55:17 <int-e> "clone this subdirectory" "merge from the previously cloned subdirectory"
23:55:25 <fizzie> moon_: I think you shouldn't try to do the thing at all. You can't run anything persistently, so your editor would need to write all state to file (and make a commit) for every operation.
23:55:41 <moon_> no, it is more of a file operator
23:55:47 <int-e> it's far easier to use something like submodules to compose a view of a larger project from smaller parts, right now.
23:55:53 <moon_> it takes a command, and preforms it
23:56:25 <int-e> So as things stand, I don't think I want to use anything resembling a "repository for everything."
23:58:42 <fizzie> Even so, I think we generally would like to discourage interactive editing on the bot, since it involves so much history.
23:59:01 <fizzie> (sed already exists, anyway.)
23:59:34 <moon_> i feel like doing it, plus it will test my shellscripting, ill do it offscreen
23:59:49 <int-e> and regarding the ghc question... I've not seriously missed it.