←2008-08-29 2008-08-30 2008-08-31→ ↑2008 ↑all
00:15:04 <tusho> dogface: commas are the enemy of this style, mostly
00:15:06 <tusho> (how ironic)
00:15:10 <tusho> i mean commas to seperate two facts
00:15:15 <tusho> "I'll also mention that domestic dogs are a subspecies of the gray wolf, and that there was some Soviet Union project to domesticate wolves."
00:15:37 <dogface> Pressing the comma key was... more attractive than pressing enter.
00:15:40 <tusho> "Perhaps it's because I think about how I'm going to say it before I manage to flip over to the IRC window." <- if you stop thinking that, and stop being afraid to split sentences, you'll get my style :P
00:15:49 <tusho> dogface: but it won't get you anywhere as far as emulating me goes :D
00:16:04 <dogface> I'll have to do some sentence breaking exercises.
00:21:33 -!- tusho has quit.
00:21:56 -!- Slereah_ has quit (Read error: 113 (No route to host)).
00:25:13 -!- Slereah_ has joined.
00:33:20 -!- Slereah_ has quit (Read error: 113 (No route to host)).
00:41:37 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Don't care..
00:48:49 <dogface> optbot!
00:48:49 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | no NE,SW, etc.
00:48:54 <dogface> optbot!
00:48:54 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Uh. . . Don't think so. . ..
00:48:58 <dogface> optbot!
00:48:58 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Nope..
00:49:11 <dogface> Well, they were getting longer for a moment, there.
00:49:14 <dogface> optbot!
00:49:14 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | mandelbrot in 6.77 secs?? impressive.
00:49:21 <dogface> Not bad.
00:53:21 <psygnisfive> dogface!
00:53:32 <dogface> psygnisfive!
00:53:37 <psygnisfive> hey.
02:02:44 -!- fungot has quit (zelazny.freenode.net irc.freenode.net).
02:02:44 -!- funktio has quit (zelazny.freenode.net irc.freenode.net).
02:02:44 -!- oklopol has quit (zelazny.freenode.net irc.freenode.net).
02:02:44 -!- dogface has quit (zelazny.freenode.net irc.freenode.net).
02:02:45 -!- lament has quit (zelazny.freenode.net irc.freenode.net).
02:02:45 -!- cmeme has quit (zelazny.freenode.net irc.freenode.net).
02:02:46 -!- Quendus has quit (zelazny.freenode.net irc.freenode.net).
02:03:14 -!- oklopol has joined.
02:03:14 -!- dogface has joined.
02:03:14 -!- fungot has joined.
02:03:14 -!- funktio has joined.
02:03:14 -!- lament has joined.
02:03:14 -!- cmeme has joined.
02:03:14 -!- Quendus has joined.
02:35:19 -!- dogface has quit ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]").
03:30:54 -!- GreaseMonkey has joined.
04:34:17 -!- puzzlet_ has quit (Read error: 60 (Operation timed out)).
04:34:22 -!- puzzlet has joined.
04:38:59 <psygnisfive> don' be quite!
04:39:08 <psygnisfive> oklopol
04:39:12 <psygnisfive> <3you
05:04:09 -!- Sgeo has joined.
05:29:56 -!- GreaseMonkey has quit ("Jesus loves you").
06:11:50 -!- Sgeo has quit (Remote closed the connection).
06:41:37 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | yeah that's pretty funny.
07:24:38 <oklopol> psygnisfive: i'm not quite
07:24:48 <oklopol> i'm always completely or not at all
07:25:07 <oklopol> anyway i don't have the time to be loved right now, have to sleep a while
07:25:09 <psygnisfive> :)
07:25:12 <psygnisfive> haha
07:26:46 <oklopol> worked all night, so tired
07:26:53 <psygnisfive> where do you work?
07:28:53 <oklopol> we do night watchmanship, supply better term if one exists, with a small group
07:29:01 <oklopol> but just like a few times a year
07:29:13 <psygnisfive> you're a rentacop?
07:29:14 <psygnisfive> lol
07:29:32 <oklopol> i do other work too, but that's more flexible
07:29:42 <psygnisfive> hah. oklopol's a rentacop. :D
07:29:44 <psygnisfive> thats so cute
07:30:14 <oklopol> perhaps, it's not really about being a cop, more about chatting with friends in a car
07:30:35 <oklopol> and call the police if we see someone, which never happens, since this is finland
07:31:28 <oklopol> anyway sleep ->
07:33:47 <psygnisfive> night <3
07:44:22 -!- puzzlet has quit (Read error: 104 (Connection reset by peer)).
07:45:27 -!- puzzlet has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:35:08 -!- Mony has joined.
08:35:51 <Mony> hi
08:47:06 -!- KingOfKarlsruhe has joined.
09:31:45 -!- Mony has quit ("À vaincre sans péril on triomphe sans gloire...").
10:01:28 -!- Corun has joined.
10:10:45 -!- kar8nga has joined.
10:50:22 -!- puzzlet_ has joined.
10:50:43 -!- puzzlet has quit (Remote closed the connection).
10:55:52 -!- olsner has joined.
10:57:03 -!- LinuS has joined.
11:29:29 -!- KingOfKarlsruhe has quit (Remote closed the connection).
11:48:27 -!- Corun has quit ("This computer has gone to sleep").
12:22:22 -!- tusho has joined.
12:29:13 -!- Corun has joined.
12:41:37 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | intuition tells me it's probably a Really Hard problem, mathematically.
12:47:29 <tusho> great topic
12:48:36 <oklopol> :---)
12:59:21 -!- kar8nga has left (?).
13:08:18 -!- Mony has joined.
13:14:23 <tusho> hmm
13:14:31 <tusho> since optbot came here, this channel has been thriving XD
13:14:31 <optbot> tusho: and similar
13:14:40 <tusho> optbot: yes.
13:14:40 <optbot> tusho: Kittens
13:14:43 <tusho> optbot: Yum.
13:14:44 <optbot> tusho: very nice writeup
13:14:48 <tusho> optbot: Thanks.
13:14:48 <optbot> tusho: latex is especially gruesome, {[]}\~ all being AltGr combinations
13:14:55 <tusho> optbot: Poor you.
13:14:55 <optbot> tusho: you can torture me by correcting my obvious mistakes ;p
13:15:03 <tusho> optbot: None so far.
13:15:03 <optbot> tusho: I'm curious, though. . . Would it be possible, or even *sane*, to implement my asm_gdt_load function via just some inline assembly?
13:15:15 <tusho> optbot: Well... assembly _can_ do anything C can...
13:15:16 <optbot> tusho: Principia Mathematica
13:15:19 <tusho> optbot: er, ok
13:15:19 <optbot> tusho: okay
13:57:08 <AnMaster> Deewiant, have you updated mycology for correct SOCK yet?
13:59:01 <Deewiant> yes
13:59:16 <Deewiant> there /is/ a homepage you can check, you know :-P
14:10:48 -!- oklopol has quit (Read error: 113 (No route to host)).
14:16:44 -!- oklopol has joined.
14:18:01 <AnMaster> Deewiant, k
14:23:49 -!- LinuS has quit ("Puzzi. S, parlo proprio con te. Puzzi.").
14:23:53 <fizzie> "Thriving and similar".
14:24:52 <fizzie> Or is it "optbot and similar"?
14:24:52 <optbot> fizzie: retoki
14:25:09 <fizzie> optbot: What does "retoki" mean?
14:25:09 <optbot> fizzie: lists would work better... being mutable.
14:25:26 <fizzie> optbot: Why do you answer tusho so sane-like and talk nonsense to me?
14:25:26 <optbot> fizzie: what position?
14:25:35 <tusho> fizzie: operating hours, dude
14:25:46 <tusho> while he sleeps he puts a random nonsense bot up
14:25:49 <fizzie> Yes, but you hit the operating hours oftener.
14:25:49 <tusho> and he sleeps a LOT
14:25:54 <tusho> well yeah
14:25:56 <tusho> I created him
14:25:57 <tusho> i know when he sleeps
14:26:21 <fizzie> fungot: Have you ever been awake? Based on your output, I'd think you're in a coma or something.
14:26:21 <fungot> fizzie: but i don't care' what more specifically are you in ohio? ( or if you have more of fnord original material than the english ones do. are there ipv6-enabled dns servers?
14:26:47 <fizzie> "I'm in a coma but I don't care", heh.
14:26:47 <tusho> pretty sexy coma
14:26:48 <tusho> wait what
14:27:00 <tusho> fizzie: he does care if you're in ohio though
14:27:09 <tusho> he also wants more of the "fnord original material"
14:27:14 <tusho> original, presumably, meaning not translated into english
14:27:15 <fizzie> fungot: Well, I'm not in Ohio. I've never even been in Ohio.
14:27:15 <fungot> fizzie: scheme48 does have a few minutes? :) imagine if the contents of your structs. :-p so, i suggest that you try to actually compare the results? :) maybe its a bug in paredit last night, then!
14:27:18 <tusho> he's also curious about ipv6
14:28:36 <fizzie> Yes, he sure asks a lot of questions.
14:32:11 <AnMaster> fizzie, what fingerprints do you use now?
14:34:41 <fizzie> AnMaster: NULL, FING, STRN, SOCK and SCKE are loaded in the beginning and are fatal errors if not found. FILE is used for persistent storage and the nonsense generator, but it should work even without FILE.
14:35:22 <fizzie> But I haven't yet fixed the STRN L/R "more characters than are in the string" undefined-behaviour dependency.
14:35:55 <fizzie> Should really do that next, it's just something like N5`| before the L.
14:38:39 <fizzie> I'm not sure how sensible it is that SCKE is a fatal error since it's only used for http:// URLs, which do not even work. And I guess some people stick the SCKE extensions in their SOCK? I could just do "if SCKE fails to load, test for H anyway, and disable the http thing if it doesn't work".
14:46:01 <tusho> ELER hasn't been updated since december 2007.
14:46:02 <tusho> :\
14:46:12 <AnMaster> ELER? another fingerprint?
14:47:31 <tusho> http://geekz.co.uk/
14:47:35 <tusho> Everybody Loves Eric S. Raymond
14:47:49 <tusho> Premise: rms, torvalds and esr live in an apartment together.
14:47:53 <tusho> Hilarity ensues.
14:48:02 <tusho> ... once every 6 months ...
14:50:22 <GregorR> Hahaha
15:52:49 -!- Mony has quit ("À vaincre sans péril on triomphe sans gloire...").
15:57:06 -!- oklopol has quit (Read error: 113 (No route to host)).
16:24:51 <Deewiant> fizzie: what do you use NULL for?
16:28:12 * tusho is working on a nifty little thing that lets you use a google account as an openid
16:28:28 <tusho> (it never sees your password, it uses the app engine login mechanism)
16:28:42 <tusho> the app engine part just lets any web app use google accounts for authentication
16:28:47 <tusho> and then i'm building the openid on top of that
16:28:52 <tusho> but still, google accounts for any site is exciting too
16:33:05 <fizzie> Deewiant: At least in my tests FING won't swap two commands if one is unused, so I just loaded NULL at the bottom. I guess a more correct way would've been to use Z in that case (and Y to drop them), but that's the way it's written now.
16:35:06 <fizzie> (I currently do 'G'KX'P'BX to move STRN's G/P to K/B so that I can still use them while FILE is loaded on top; and an identical command to revert it. I guess I could've just used 'G'KZ'P'BZ and then 'KY'BY to unload.)
16:35:36 <Deewiant> hm, seems silly for FING to behave that way.
16:35:43 <Deewiant> but I guess it's intentional.
16:36:19 <fizzie> That's what it does on RC/Funge-98; the spec doesn't really say much. It's maybe a bit illogical, given that Z will push a reflect if "src" is empty.
16:36:35 <Deewiant> ah, "push a reflect", not "reflect".
16:36:42 <Deewiant> in that case it seems it's not intentional. :-P
16:37:22 <tusho> push a reflect? XD
16:37:38 <fizzie> Currently RC/Funge-98's X reflects if either one them is empty.
16:38:34 <fizzie> Haven't seen Mike Riley around in the last few days, there was that one clear RC/Funge-98 bug I patched that I don't even remember any more, and I was going to ask about the strange behaviour his FILE 'R' had at EOF.
16:39:02 <Deewiant> you might want to e-mail him
16:40:16 <fizzie> Actually I don't see any patched bugs. Maybe I imagined that one, since I have no idea what it could've been.
16:41:43 <fizzie> Oh, there is it; my 'diff' invocation was missing -r. His FILE 'C' only fclose()d the handle and did not set it to NULL, causing it to run out of file handles fast since fungot's nonsense-generator opens and closes two files for each phrase.
16:41:44 <fungot> fizzie: and slereah_ said it wasn't not. and i threw what i had in mind. irc wasnt designed like that, at least on my machine
16:52:14 -!- KingOfKarlsruhe has joined.
17:12:35 -!- Judofyr has joined.
17:33:57 -!- puzzlet has joined.
17:34:50 -!- puzzlet_ has quit (Remote closed the connection).
17:37:08 -!- sebbu has joined.
17:44:00 -!- Mony has joined.
17:45:07 <funktio> does Z leave leave src as it is, or pop it? i.e., is it a copy or a move?
17:51:20 <AnMaster> funktio, in what fingerprint?
17:51:48 <funktio> AnMaster: FING
17:52:11 <AnMaster> hm...
17:52:46 <AnMaster> copy it seems
17:53:15 <funktio> ok, thanks
17:55:15 <AnMaster> <fizzie> Deewiant: At least in my tests FING won't swap two commands if one is unused, so I just loaded NULL at the bottom. I guess a more correct way would've been to use Z in that case (and Y to drop them), but that's the way it's written now.
17:55:16 <AnMaster> indeed
17:55:45 <AnMaster> according to the test Mike Riley wrote it is intended like that too
17:55:55 <Deewiant> that's awfully stupid
17:56:16 <AnMaster> Deewiant, what do you suggest? push a reflect?
17:56:19 <AnMaster> or move?
17:56:22 <AnMaster> for that situation
17:56:25 <Deewiant> a reflect, of course
17:56:40 <AnMaster> push a reflect, or reflect?
17:56:50 <Deewiant> it's like the cell stack is an infinite pile of zeroes at the bottom: the fingerprint stacks are an infinite pile of reflects
17:57:21 <funktio> I push a reflect now, but it'd be trivial to change
17:57:57 <AnMaster> well easy to change for me too
17:59:00 <AnMaster> Z pushes a reflect if the original one is empty btw
17:59:10 <AnMaster> which means it is even more weird
18:03:43 -!- kar8nga has joined.
18:06:46 -!- puzzlet has quit (Read error: 104 (Connection reset by peer)).
18:07:18 -!- puzzlet has joined.
18:12:36 <AnMaster> Deewiant, did you get my mail about SOCK replacement?
18:12:39 <AnMaster> was a few days ago
18:12:47 <AnMaster> however I thought of another thing
18:12:53 <AnMaster> wait on socket group
18:13:12 <AnMaster> to allow making a single threaded server to handle multiple sockets
18:13:28 <AnMaster> like SCKE's P, but on multiple sockets
18:13:32 <Deewiant> yeah, something like select(2)
18:13:47 <AnMaster> Deewiant, yeah, or poll() or whatever
18:14:10 <AnMaster> I use poll() not select() in cfunge currently
18:14:20 <Deewiant> yeah, I'd expect you to :-P
18:18:43 <AnMaster> Deewiant, why?
18:18:50 <AnMaster> it was just that poll() is easier to use IMO
18:18:50 <Deewiant> because you're like that :-P
18:18:52 <tusho> AnMaster: because it's rather typical overengineering of you
18:19:02 <AnMaster> tusho, not really, select() is messy IMO
18:19:13 <AnMaster> on linux select() is just a wrapper for poll() anyway
18:19:25 <AnMaster> on *bsd poll() is a wrappper for select() iirc
18:19:26 <AnMaster> heh
18:19:39 <AnMaster> they are rather similiar
18:19:56 <AnMaster> using kqueue or epoll would have been overengineering
18:20:16 <AnMaster> tusho: so what do you mean poll() is "overengineering"
18:20:43 <AnMaster> and what did Deewiant mean?
18:21:33 <Deewiant> AnMaster: I expect you to use the lowest-level interface you can find :-P
18:22:38 <AnMaster> Deewiant, um? on *bsd it is higher level
18:22:46 <Deewiant> I didn't know that
18:22:46 <AnMaster> and I actually wrote the code on a *bsd system
18:22:50 <AnMaster> so...
18:23:16 <AnMaster> now I hope tusho will explain himself too
18:23:16 <AnMaster> ...
18:28:04 <tusho> i'm actually just being silent to irritate you
18:28:20 <AnMaster> tusho, so I will continue to bug you to irritate you
18:28:33 <tusho> cool, i'll /ignore you if you do :D
18:28:38 <tusho> but you can't /ignore my silence!
18:28:50 <AnMaster> tusho, I either expect an explanation or a sorry
18:29:04 <AnMaster> because I see no reason why you should attack me without justifying what you say
18:29:07 <AnMaster> *shrug*
18:29:10 <tusho> i never attacked you actually
18:29:11 <tusho> :D
18:29:14 <AnMaster> <tusho> AnMaster: because it's rather typical overengineering of you
18:29:17 <AnMaster> that was an attack
18:29:18 <tusho> yes
18:29:18 <Deewiant> he's just being the troll he normally is :-P
18:29:19 <tusho> is that an attack?
18:29:20 <tusho> :P
18:29:27 <AnMaster> Deewiant, well probably you are right
18:29:42 <AnMaster> after all while he may be 13, I think it seems rather like 3
18:29:52 <tusho> wow, i am one smart fucking 3 year old
18:29:59 <AnMaster> at least when trolling
18:30:19 <AnMaster> of course he must be older than 3 since he actually can program and such
18:30:24 <AnMaster> but sometimes he doesn't act like it
18:30:34 <tusho> AnMaster: You mean like Theo de Raat?
18:30:37 <tusho> ICE BURN
18:30:45 <tusho> raadt, whatever
18:31:05 -!- GregorR_ has joined.
18:31:18 -!- GregorR has quit (Nick collision from services.).
18:31:24 -!- GregorR_ has changed nick to GregorR.
18:32:25 <AnMaster> tusho, actually you are right there, though I think Raadt knows a bit more than you (no offence meant @ Raadt)
18:33:03 <tusho> he's programmed for like decades longer than me
18:33:06 <tusho> so no surprise :P
18:33:17 <tusho> i started programming at, uh, 8 i think
18:33:24 <tusho> though i'd done shit like html a few years before that
18:34:03 -!- puzzlet has quit (Read error: 60 (Operation timed out)).
18:34:07 -!- puzzlet has joined.
18:34:44 <GregorR> VirtualBox = awesome
18:37:59 <AnMaster> tusho, when you were 5?
18:38:03 <AnMaster> 6?
18:38:12 <tusho> 7.5 or so for htmling, i think
18:38:16 <tusho> perhaps a bit later
18:38:17 <AnMaster> a few years?
18:38:22 <tusho> meh
18:38:24 <tusho> :P
18:38:25 <AnMaster> 8-7.5 = 0.5;
18:38:26 <tusho> but yeah, my first language was php :-(
18:38:31 <AnMaster> 0.5 < a few
18:38:35 <tusho> AnMaster: whatevs
18:38:37 <AnMaster> because
18:38:40 <AnMaster> 0.5 < 1
18:38:42 <AnMaster> 1 = one
18:38:45 <AnMaster> one < a few
18:38:46 <AnMaster> thus
18:38:48 <tusho> AnMaster. i may be 13.
18:38:52 <AnMaster> 0.5 < a few
18:38:53 <tusho> but i am familiar with this thing called "arithmetic"
18:38:58 <AnMaster> tusho, yep
18:39:05 <tusho> you may shut up now.
18:39:06 <AnMaster> so how is half a year == a few years
18:39:14 <tusho> because my memory is not very good, ok
18:39:17 <AnMaster> ah
18:39:18 <tusho> so i had to think about it after saying that
18:41:37 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | but I don't know how much metadata PPM images have.
18:54:18 <GregorR> Although a metric few is definitely greater than 1, an imperial few can take any value from 0.1 to 1000.
18:55:23 <tusho> GregorR: you like weird little web things, don't you? (see: choosemyhat, extra-www)
18:55:30 <GregorR> Yes.
18:55:31 <GregorR> Yes I do.
18:55:56 <tusho> GregorR: i am making a little thing that isn't really 'funny', but it certainly is a little crazy in how it works
18:56:05 <GregorR> Fascinating.
18:56:09 <tusho> here it is:
18:56:23 -!- Mony has quit ("À vaincre sans péril on triomphe sans gloire...").
18:56:35 <tusho> with google's app engine your app can use google accounts for authentication. that's all well and good but what if I want google accounts for authentication but not use app engine? maybe there's a library I need, maybe I don't wanna use python, maybe I wanna have control over my app, etc
18:56:56 <tusho> well, this is a little app engine app that lets you do just that. you add an application, and give it a URL and you get two keys
18:57:22 <tusho> (pubkey and privkey). to login, you redirect users to http://gaccproxy.appspot.com/login?app=PUBKEY&token=TOKEN&return=URL
18:57:30 <tusho> return is just the url you want users to return to after login
18:57:43 <tusho> token is anything you want that you can uniquely identify a user short-term by
18:57:45 <tusho> ip or whatever
18:57:47 <tusho> then
18:58:02 <tusho> it gets the user to log in on app engine, and asks for deny/allow access to it to the user, etc
18:58:04 <tusho> and you get a post
18:58:08 <tusho> to the URL you entered for the app
18:58:18 <tusho> with the private key, the token, and the account details (nickname & email)
18:58:26 <tusho> you check the private key is right (this is how you know it's gaccproxy talking to you)
18:58:28 <GregorR> OK ...
18:58:33 <tusho> and then store {token=>details}
18:58:46 <tusho> THEN, when it finally gets back to the return URL, you look up the token and nab the details off
18:58:54 <tusho> it's just funny how many server roundtrips and keys you need to get it all working :D
18:58:57 <tusho> well, funny for me at least.
18:59:15 <tusho> it's certainly rather silly
18:59:32 <tusho> but useful, and i quite like silly but useful things
18:59:56 <GregorR> ... I believe my response is "Uh ... wtf?"
19:00:07 <tusho> GregorR: what is the wtf bit :D
19:00:41 <GregorR> So, it's sort of like a centralized user authentication service?
19:00:45 <tusho> kind of
19:00:51 <tusho> it just lets any web app use google accounts for authentication
19:00:58 <tusho> i'm planning to build a google account->openid bridge on top of it
19:01:10 <GregorR> OH, I see, it actually uses Google accounts.
19:01:15 <GregorR> I didn't see where that part came in :P
19:01:21 <tusho> from the user's point of view, they hit login, enter their google details (using google's handlers, i never see your password or whatever), it asks them if they want to allow it, then it puts them back to the sit
19:01:21 <tusho> e
19:01:25 <tusho> but yeah
19:01:31 <tusho> obviously, normal apps can't use google's login page
19:01:44 <tusho> but App Engine (their python web app framework/hosting service thing) lets you use it
19:02:02 <tusho> so it's an app engine app that solely lets a user login via google accounts and it reports back to your app
19:02:17 <tusho> i just thought that 1) it's a neat hack 2) the hoops needed internally to get it working were rather amusing/silly
19:02:20 <GregorR> Right.
19:02:29 <GregorR> Makes sense. What do I have to do with it?
19:02:39 <tusho> GregorR: nothing, you just like silly web stuff so :P
19:02:41 <tusho> it is possibly just my weird sense of humour kicking in :D
19:02:43 <GregorR> Heh
19:03:17 <tusho> *hacks up the last bit (an "Always allow" button)*
19:03:53 <GregorR> I have Windows XP apps running seamlessly on GNU/Linux.
19:04:00 <GregorR> (Not Wine :P )
19:04:03 <GregorR> VirtualBox = awesome.
19:04:39 <tusho> GregorR: I have Windows XP apps running seamlessly on OS X - truly seamlessly: windows on Windows (lulz) appear as windows on my os x desktop
19:04:46 <tusho> as in, there's no emulator window
19:04:47 <tusho> it's just all there
19:04:52 <GregorR> That's my point.
19:04:55 <GregorR> VirtualBox does that now.
19:04:56 <tusho> also, open windows applications appear in the dock
19:04:59 <tusho> GregorR: ah, does it?
19:04:59 <tusho> nifty
19:05:04 <GregorR> VirtualBox ripped that feature off of VMWare :P
19:05:04 <tusho> parallels had it first though ;)
19:05:07 <GregorR> Of course.
19:05:13 <tusho> parallels had it before vmware too, i think
19:05:16 <tusho> not sure
19:05:20 <GregorR> Oh, Parallels? I thought Parallels was VMWare >_>
19:05:22 <fizzie> And Wine has had it always. :p
19:05:23 <GregorR> I'm so confuddled.
19:05:26 <GregorR> Heh
19:05:37 <tusho> fizzie: All it needs to do now is run Windows applications.
19:05:42 <tusho> ICE BURN x2
19:05:57 <fizzie> It runs a couple. I think. Or so I have heard. I've played Diablo II with it. :p
19:05:57 <GregorR> Now if I can only get OS X installed in an emulator X-P
19:06:24 <tusho> GregorR: go get one of those OSx86 isos
19:06:32 <tusho> that should work in vmware, probably
19:06:36 <tusho> virtualbox probably does too much tricky stuff
19:06:43 <GregorR> tusho: I haven't found one that'll work in anything. I tried one in VMWare, but failed :(
19:06:55 <tusho> well, it's not really very work
19:06:55 <tusho> y
19:07:07 <tusho> as apple only support like, 30 pieces of hardware in total
19:07:08 <tusho> :P
19:07:10 <GregorR> Yuh, it's not like OS X runs on normal unmodified PCs.
19:07:32 -!- dogface has joined.
19:07:41 <GregorR> I wish OpenDarwin still existed. Then at least I could get binaries that would work >_>
19:08:02 <tusho> r.i.p. :P
19:08:07 <dogface> How many points do I get for connecting to an IRC server via some random Internet server via my own IRC client here?
19:08:10 <dogface> Using nc?
19:08:15 <tusho> dogface: 3.4
19:08:25 <dogface> Yay, that sounds good.
19:09:01 <tusho> dogface: if you'd written an ssh<->irc bridge then used that to shell in somewhere, then used nc from there to connect here, that'd be 7 points
19:09:16 <tusho> 8 or more points involve killing aardvarks, I'm afraid.
19:09:34 <GregorR> But not with your bare hands, you have to kill them with the power of IRC.
19:10:52 <dogface> I connected to the server with SSH, then ran nc -l -p 39528 -c "nc secret-stuff" &, then connected to the server with IRC.
19:16:19 -!- kar8nga has left (?).
19:16:27 <AnMaster> <tusho> GregorR: you like weird little web things, don't you? (see: choosemyhat, extra-www) <-- choosemyhat??
19:16:34 <tusho> AnMaster: choosemyhat.com
19:17:38 <tusho> brb
19:22:19 <AnMaster> how many dimensions does a html page have
19:24:13 <AnMaster> <GregorR> I wish OpenDarwin still existed. Then at least I could get binaries that would work >_>
19:24:16 <AnMaster> it doesn't?
19:24:20 <AnMaster> wtf
19:24:26 <AnMaster> it still must be open source
19:24:31 <AnMaster> because they reuse code from such
19:24:35 <AnMaster> right?
19:25:31 <dogface> What's the regex escape thing for a non-whitespace character? \S?
19:25:38 <GregorR> AnMaster: But not enough to build an OS out of.
19:25:50 <AnMaster> hrrm true
19:33:33 <fizzie> Neh, it is rather less fun to edit old Befunge code than to write new. I'm having trouble fitting in all these "is this string long enough to do 5L on it" length tests without allocating silly-looking extra lines.
19:46:48 <AnMaster> dogface, depends on dialect
19:46:57 <AnMaster> in some it is \w iirc
19:47:05 <AnMaster> and \s for space
19:47:22 <AnMaster> fizzie, see you learn it!
19:48:09 <AnMaster> hum I just got an idea
19:48:16 <Deewiant> dogface: [^ \t\r\n\f] :-P
19:48:43 <dogface> You know, I think [^ ] would be sufficient.
19:49:17 <AnMaster> I think using w to do string matching, would be the fastest way in befunge, because you could then take a char in the middle, compare it, if it is the same go forward, if it is less take the path that handles commands that begin in lower chars and if it is more take the other ones
19:49:19 <AnMaster> one*
19:49:41 <AnMaster> could be tricky to write
19:49:50 <funktio> where is \w equivalent to non-whitespace?
19:50:05 <AnMaster> funktio, not sure, I may remember wrong
19:50:05 * dogface suddenly realizes that he can use cat to concatenate
19:50:10 <AnMaster> I tend to use POSIX regex
19:50:22 <AnMaster> so that means no \w or \s last I checked
19:50:23 <funktio> \w is usually [a-zA-Z_0-9], I think
19:50:32 <AnMaster> ah maybe
19:50:48 <AnMaster> dogface, that is what it is made for
19:51:02 <AnMaster> it is what cat stands for
19:51:13 <AnMaster> it is infact almost the only use of cat
19:51:25 <fizzie> At least there's ] so I can merge these 'v' and '^' that want to occupy the same space.
19:51:30 <dogface> I've never seen it used elsewise. :-P
19:51:39 <AnMaster> in scripts cat used for anything else than to concatenate two files == bad idea
19:51:59 <AnMaster> cat foo | grep bar is a common beginner mistake for example
19:52:02 <AnMaster> grep bar foo
19:52:05 <AnMaster> is the correct way
19:52:18 <fizzie> There's the Useless Use of 'cat' award thing.
19:52:26 <AnMaster> fizzie, was just going to mention that
19:52:41 <funktio> I often use `cat file.txt` in Perl to read a file
19:52:44 <AnMaster> <greybot> Useless Use of Cat (cat foo | grep bar). See http://partmaps.org/era/unix/award.html
19:52:49 <AnMaster> bot from #bash btw
19:52:56 <Deewiant> AnMaster: how about "cat foo >> bar"
19:53:11 <Deewiant> that's not concatenating two files, strictly speaking :-P
19:53:16 <tusho> hehe those kill -9 haters will totally, er, hate me
19:53:23 <tusho> I use kill -9 all the time :)
19:53:24 <AnMaster> Deewiant, hm, that is concatenating, kind of
19:53:33 <AnMaster> Deewiant, you are sure it is bash?
19:53:41 <AnMaster> or?
19:53:47 <Deewiant> of course, that's even sh
19:53:55 <fizzie> Still, I tend to cat a lot in use-once-only pipelines; it somehow goes more naturally when starting with cat, plus you can then stick more items in the pipeline without having to change it.
19:53:58 <AnMaster> Deewiant, no I mean the shell you want to run the command it
19:53:59 <AnMaster> in*
19:54:07 <AnMaster> because I can replace it with a better one in bash I think
19:54:12 <AnMaster> but that isn't portable
19:54:16 <AnMaster> something like
19:54:22 <AnMaster> echo >> bar < foo
19:54:26 <AnMaster> I *think*
19:54:28 <AnMaster> actually no
19:54:29 <fizzie> When you do "cat foo | grep bar" it's easier to amend that to "cat foo | sed ... | grep bar"; with "grep bar foo" you need to actually think to turn it to "sed ... foo | grep bar".
19:54:31 <AnMaster> won't work
19:54:42 <AnMaster> err
19:54:52 <AnMaster> fizzie, sed | grep is probably wrong too
19:54:57 <AnMaster> you can do it all in sed
19:55:00 <AnMaster> or you may want awk
19:55:24 <Deewiant> AnMaster: what happened to the unix philosophy of one tool for one job
19:55:32 <AnMaster> fizzie, so sed | grep or awk | grep or sed | awk is just plain wrong
19:55:36 <AnMaster> Deewiant, it is still there
19:55:37 <AnMaster> but
19:55:41 <AnMaster> sed does have grepping
19:55:41 <tusho> Deewiant: it wasn't optimized enough
19:55:43 <fizzie> Well, it was just an example. And I do what it's simpler to write, not what uses less processes, when trying to write a use-once pipeline.
19:55:43 <tusho> :)
19:55:53 <AnMaster> /foo/s/bar/quux/
19:55:59 <Deewiant> tusho: I bet that grep is probably faster for searching than sed, anyway
19:56:01 <AnMaster> will replace bar with quux on any line containing foo
19:56:03 <AnMaster> I think
19:56:06 <AnMaster> yes it would
19:56:32 <AnMaster> Deewiant, why wear out the keyboard when you don't have to? ;P
19:56:37 <Deewiant> especially if you use grep -F or such
19:56:41 <fizzie> That doesn't do the same thing, though.
19:57:02 <fizzie> You want to first do s/bar/quux/ and then print only lines that _still_ contain foo.
19:57:11 <funktio> /foo/ && s/bar/quux/ # in Perl
19:57:17 <fizzie> Personally I would have to peek at the sed man page for that.
19:57:32 <Deewiant> AnMaster: it's more typing to change grep -> sed compared to just "cat foo | grep baz | bar" -> "cat foo | grep baz | sed qux | bar"
19:57:37 <AnMaster> another horrible thing I see is: cat foo | grep bar | awk '{ print $3 }'
19:57:41 <AnMaster> now that is just horrible
19:57:43 <fizzie> On the other hand, I can just write "sed s/bar/quux/ | grep foo" and any monkey will understand what that does.
19:57:51 <AnMaster> awk '/bar/{ print $3 }' foo
19:57:57 <AnMaster> is the "right" way
19:58:04 <fizzie> No, it's your way. :p
19:58:11 <Deewiant> I think it'd be "sed 'N; s/bar/quux; /foo/P;'" or something
19:58:15 <AnMaster> fizzie, in a script UUOC or similiar is bad
19:58:26 <Deewiant> possibly some pointless or missing semicolons there
19:59:11 <AnMaster> well it is true, doing it in that order I would probably just use awk
20:00:19 <AnMaster> or: cat foo | sort -n | uniq
20:00:24 <AnMaster> sort -nd foo
20:00:32 <AnMaster> assuming you don't want some other flags uniq provides
20:00:35 <AnMaster> like -d or -c
20:00:43 <AnMaster> err
20:00:45 <AnMaster> -nu
20:00:46 <AnMaster> not -nd
20:00:47 <fizzie> There's also the fact that when writing a pipeline, I tend to construct it one operation at a time. I don't know about scripts, since I usually just write Perl at that point.
20:00:48 <AnMaster> typo
20:01:04 <AnMaster> fizzie, I have written an IRC bot in bash unless you didn't know
20:01:08 <AnMaster> modular one
20:01:14 <fizzie> Yes, you tend to mention it. :p
20:01:19 <AnMaster> true
20:01:29 <fizzie> Have I mentioned the one I did in Funge-98?-)
20:01:34 <AnMaster> fizzie, I know it
20:01:49 <AnMaster> fizzie, oh and how does it's randomness generator work
20:01:53 <AnMaster> ^help
20:01:53 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf, code=text/str:N; ^str 0-9 get/set/add [text]
20:02:00 <AnMaster> fungot?
20:02:01 <fungot> AnMaster: any actionscript object could be wrapped in a dynamic-wind exit thunk?
20:02:04 <fizzie> Oh, ^bool is missing from the help.
20:02:10 <AnMaster> ^bool
20:02:10 <fungot> Yes.
20:02:12 <AnMaster> ^bool
20:02:12 <fungot> Yes.
20:02:14 <AnMaster> um
20:02:18 <AnMaster> ^bool false
20:02:22 <AnMaster> eh?
20:02:24 <fizzie> It's only "^bool".
20:02:24 <tusho> ^bool
20:02:25 <fungot> Yes.
20:02:27 <tusho> ^bool
20:02:27 <fungot> No.
20:02:34 <fizzie> fungot's a "yes" man.
20:02:35 <fungot> fizzie: ( quote cons) evaluates to the list." at http://paste.lisp.org/ display/ 14194 with " nantampukmas" at http://paste.lisp.org/ display/ 314
20:02:36 <AnMaster> randomness generator?
20:02:54 <fizzie> oklopol wanted a single bit; a single bit is what he got.
20:02:54 <AnMaster> err
20:02:58 <AnMaster> why those spaces
20:03:00 <AnMaster> fizzie, ^
20:03:05 <AnMaster> at http://paste.lisp.org/ display/ 314
20:03:09 <tusho> AnMaster: fizzie just talked 3 seconds ago
20:03:09 <AnMaster> some spaces too much there
20:03:13 <tusho> you do not have to highlight it.
20:03:19 <AnMaster> tusho, really?
20:03:22 <AnMaster> tusho, are you sure?
20:03:25 <AnMaster> tusho, ^
20:03:28 <tusho> /ignore AnMaster
20:03:47 <AnMaster> tusho:enable_humor().
20:03:56 <fizzie> AnMaster: Because my Funge-98 ported tokens-to-text routine does not have all the features the prototype had.
20:04:08 <tusho> AnMaster: AnMaster:enable_annoyance(). Oh wait, it's already on.
20:04:33 <fizzie> Currently it just adds a " " before each word-like word.
20:04:49 <AnMaster> fizzie, I see, how does it work though?
20:04:59 <fizzie> That is also why "( quote cons)" and not "(quote cons)".
20:05:26 <fizzie> The text generation or what? And how much detail do you want? The source is of course available, too. :p
20:05:58 <AnMaster> fizzie, forgot the link
20:06:10 <fizzie> http://zem.fi/~fis/fungot.b98.txt if I remember correctly.
20:06:11 <fungot> fizzie: but they took a survey and found that of the original befunge-93 ( pressey, 1998) which is in r5rs... and it's fnord est here, so i don't know how to use syntax-rules instead because it's simpler to write it
20:06:29 <fizzie> Whoa? Befunge-93 is in R5RS Scheme?
20:06:35 <fizzie> Why haven't I heard of this before.
20:06:50 <fizzie> fungot: Are you quite sure?
20:06:50 <fungot> fizzie: i think a similar argument could be that strict too, but i'm not
20:07:03 <Deewiant> ah, he was just kidding
20:07:31 <AnMaster> fizzie, same loader as before?
20:08:09 <AnMaster> fizzie, why all the fnord?
20:08:56 <AnMaster> also... "original befunge-93 (pressey, 1998)" looks like someone quoted my befunge-108 draft and then messed up years (should be 1993...)
20:08:58 <fizzie> AnMaster: I kept only words that occurred >1 times in the source text, turned all others to UNK tokens; and the Funge-98 code maps UNK to fnord. (Sorry for those who are seeing this explanation the third or fourth time.)
20:09:04 * AnMaster checks bibtex file
20:09:29 <AnMaster> fizzie, ok what about a factoid database next
20:09:38 <AnMaster> maybe we need a SQL fingerprint XD
20:09:50 * AnMaster lols at the thought
20:10:22 <AnMaster> base would be SQL, then SQLI would be sqlite for example, SQPS postgre sql, SQMY mysql and so on
20:10:30 <AnMaster> what do you think?
20:10:41 <tusho> AnMaster: I think that'll be even less useful than TRDS.
20:10:50 <AnMaster> I would probably only implement SQLI
20:11:26 <fizzie> I think it was fungot who quoted wrong; 4-grams mean 3 words of context, which means it only saw "opening parenthesis, 'pressey', comma", and that is commonly followed by 1998, since there are many lines talking about Funge-98.
20:11:27 <fungot> fizzie: just for boys? :)
20:12:18 <fizzie> fungot: No, Befunge's not just for boys.
20:12:25 <AnMaster> fizzie, hum?
20:12:43 <AnMaster> fizzie, what is the link to the loader again?
20:13:03 <fizzie> Uh, I think it's something like fungot-load-freenode.b98.txt in the same URL.
20:13:20 <AnMaster> k
20:13:22 <fizzie> Same directory, I mean.
20:13:29 <AnMaster> fizzie, does that bot handle STRN?
20:13:48 <fizzie> You mean your take of the STRN L/R? No. I'm currently fixing it to do that.
20:13:55 <AnMaster> ah ok
20:14:00 * AnMaster waits with wget
20:14:19 <fizzie> There are only two instances of L to fix, I think, and then some testing.
20:19:27 <fizzie> It seems to work now. I didn't fix the http:// handler since that's disabled anyway, as it currently just ends after the H hostname lookup.
20:19:47 <fizzie> Of course I just tested it with RC/Funge-98 only.
20:22:01 <fizzie> ^reload
20:22:02 <fungot> Reloaded.
20:22:19 <fizzie> fungot: Do you feel any different now?
20:22:20 <fungot> fizzie: so don't complain about lyx to text, will need to grasp to become a good programmer, it's the best
20:23:16 <fizzie> fungot: LyX as in the LaTeX editor? I didn't really like it.
20:23:16 <fungot> fizzie: i just implemented fisher-yates shuffler procedure. shuffle-vector! shuffle-string!, and cond switching irssi windows?!? to fnord fnord
20:24:20 <AnMaster> fizzie, I love LyX
20:24:37 <fizzie> Well, *you* would.
20:24:45 <AnMaster> fizzie, why do you say that?
20:24:51 <fizzie> (I don't know what that means, it just sounded like something to say.)
20:24:57 <AnMaster> hum
20:25:04 <AnMaster> fizzie, tell me when new version is up
20:26:02 <fizzie> It should be there... now. (And I guess I'm just used to writing LaTeX without any help.)
20:29:36 <AnMaster> fizzie, what does fungot use the owner thingy for?
20:29:37 <fungot> AnMaster: waiting for another ant to do something, your pointer is cast to the wrong thing here, because two fnord were cheaper than one ultra fast one. it preserves every step in the compilation hierarchy can be overriden with new words; one could, i'd have to consider this as an example
20:29:56 <AnMaster> "Bot owner prefix"
20:30:15 -!- funnygot has joined.
20:30:19 <AnMaster> there it is
20:30:27 <AnMaster> %help
20:30:28 <funnygot> %<lang> <code>; %def <command> <lang> <code>; %show [command]; lang=bf, code=text/str:N; %str 0-9 get/set/add [text]
20:30:45 <AnMaster> fizzie, feel free to test it
20:30:47 <AnMaster> it is in a chroot
20:31:21 <AnMaster> actually
20:31:24 <AnMaster> let me fix a ulimit
20:31:25 -!- funnygot has quit (Remote closed the connection).
20:31:26 <fizzie> The PRIVMSG sender nick!user@host prefix is checked against the owner prefix for the "code", "reload" and... maybe some other command too. Yes, "save" also.
20:31:55 <AnMaster> what does save do?
20:31:59 -!- funnygot has joined.
20:32:07 <AnMaster> fizzie, the %help doesn't mention those commands
20:32:16 <fizzie> Yes, because they work for the owner only.
20:32:24 <AnMaster> fizzie, raw too?
20:32:31 <fizzie> Oh, right, that too.
20:32:45 <AnMaster> anyway what commands are there for owners
20:32:50 <AnMaster> %reload
20:32:51 <funnygot> Reloaded.
20:32:56 <AnMaster> %help
20:32:56 <fizzie> And "save" stores the 10 "str" strings and all defined commands to the state file, "data/fungot.dat" or whatever it was.
20:32:57 <funnygot> %<lang> <code>; %def <command> <lang> <code>; %show [command]; lang=bf, code=text/str:N; %str 0-9 get/set/add [text]
20:32:57 <fungot> fizzie: where is the base case. what sorts of startup times do you need
20:33:14 <AnMaster> ^reload
20:33:19 <AnMaster> hm interesting
20:33:21 -!- tusho has quit (Remote closed the connection).
20:33:29 <AnMaster> %öööö
20:33:59 -!- tusho has joined.
20:34:04 <fizzie> The babble-generator should handle gracefully the case where the language model files it needs do not exist; so funnygot should not mind if someone mentions its name.
20:34:10 <fizzie> Haven't tested that, though.
20:34:51 <AnMaster> funnygot?
20:34:53 <AnMaster> ¤help
20:34:55 <AnMaster> %help
20:34:56 <funnygot> %<lang> <code>; %def <command> <lang> <code>; %show [command]; lang=bf, code=text/str:N; %str 0-9 get/set/add [text]
20:34:59 <AnMaster> hm true
20:35:07 <AnMaster> fizzie, how are those created?
20:35:08 <tusho> funnygot: bla bla.
20:35:13 <AnMaster> where are the scripts to do so
20:35:31 <AnMaster> fizzie, what does code command do?
20:35:34 <fizzie> Er, you need a Perl script and a C++ proggie to do it, and the Perl script handles my IRC log format only. I haven't put them anywhere, at least yet.
20:35:41 <tusho> code isn't a command, AnMaster
20:35:48 <AnMaster> <fizzie> The PRIVMSG sender nick!user@host prefix is checked against the owner prefix for the "code", "reload" and... maybe some other command too. Yes, "save" also.
20:35:52 <AnMaster> seems like it is?
20:36:11 <tusho> %commansd
20:36:13 <tusho> %commands
20:36:14 <fizzie> "code" appends something like "0R" to the argument, stores it somewhere, and jumps there with SUBR.
20:36:21 <tusho> fizzie: yes, but the thing in the help there
20:36:22 <tusho> wasn't code
20:36:25 <fizzie> It's pretty easy to mess things up with.
20:36:26 <tusho> that was saying what the code param could be
20:36:35 <AnMaster> fizzie, I can imagine
20:36:41 <AnMaster> fizzie, unless you are the author
20:37:01 <AnMaster> same reason why I mark eval module of envbot as "for debugging only"
20:37:12 <fizzie> About the only thing I have used it for is "^code 000f-p" to reset the babble-generator counter to make fungot talk to me again.
20:37:14 <fungot> fizzie: 11926: that's like taking people for granted
20:37:15 <AnMaster> and put it in a non-standard place
20:37:35 <AnMaster> babble-generator counter?
20:37:37 <AnMaster> what does that do
20:37:40 <fizzie> fungot: Well, you're not exactly "people".
20:37:41 <fungot> fizzie: when i mentioned locales and character sets. maybe thats what them meant by experimental eh? " steppin' out", if you want mathematical variables, you probably won't have those same teachers then, will you teach face-to-face?
20:37:55 <fizzie> AnMaster: It only replies four times consecutively to any one nick. That's to prevent infinite loops with optbot.
20:37:55 <optbot> fizzie: i'd use php :)
20:37:58 <fizzie> ^echo optbot
20:37:58 <optbot> fizzie: :-)
20:37:58 <tusho> those were lyrics
20:37:58 <fungot> optbot optbot
20:37:58 <optbot> fungot: ~exec while self.i<3: self.i=self.i+1;self.raw("topic #esoteric :"+self.topic[self.i:self.i+5]);time.sleep(0.7)
20:38:00 <fungot> optbot: t(x) x 3. there should be
20:38:01 <optbot> fungot: right?
20:38:01 <fungot> optbot: fnord/ fnord/ fnord/ test.html here? i was wondering how ofter dynenv are cloned.... or even in the presence of dynamic-wind?
20:38:01 <optbot> fungot: Just the fractional part?
20:38:02 <fungot> optbot: sarahbot, yow she says some random thing then? :) forgot? i read something from each in cml.
20:38:02 <optbot> fungot: even "where the memory pointer points" and "where the IP is" are global state, so a better definition is needed
20:38:03 <fungot> optbot: what would you have scheme", " a christmas fnord?) guitar riffs this band is metal)." fnord dixon either, but it doesn't feel very accurate. ' assembly' is a macro and you were waiting for someone to create a dogma and raise a bunch of
20:38:03 <optbot> fungot: it's not a funge interp
20:38:09 <AnMaster> fizzie, ah right
20:38:09 <fizzie> There it stops.
20:38:28 <AnMaster> lol
20:38:41 <tusho> "FUNGOT: It's not a funge interp."
20:38:42 <AnMaster> doesn't optbot have such a protection?
20:38:42 <optbot> AnMaster: what's stake mean there?
20:38:45 <tusho> AnMaster: no.
20:38:50 <tusho> optbot:
20:38:50 <AnMaster> and who is "sarahbot"?
20:38:50 <optbot> tusho: thats not an explanation.
20:38:50 <tusho> optbot:
20:38:51 <optbot> tusho: "what does your name mean?" "your mom"/"what does YOUR name mean"/"it's my name"/"don't wear it out"/"I'm borderline autistic"/"The last person to ask that is now fashioned into boots"
20:38:51 <tusho> optbot:
20:38:51 <tusho> optbot:
20:38:51 <AnMaster> fizzie, ?
20:38:51 <optbot> tusho: I do on different channels
20:38:52 <optbot> tusho: no
20:38:52 <tusho> optbot:
20:38:52 <optbot> tusho: the Game class was trivial to make
20:38:52 <tusho> optbot:
20:38:53 <optbot> tusho: $ ./ccbi --help
20:38:53 <tusho> optbot:
20:38:54 <optbot> tusho: though it must be said...
20:38:56 <tusho> etc
20:39:01 <AnMaster> tusho, bad
20:39:07 <tusho> AnMaster: nobody gives a shit :)
20:39:13 <fizzie> AnMaster: Sarahbot's on freenode #scheme. Or at least was there, haven't been on the channel lately.
20:39:34 <fizzie> AnMaster: The language model is built from my logs of freenode/#esoteric, freenode/#scheme and ircnet/#douglasadams.
20:39:39 <tusho> i stopped visiting #scheme 'cause of rihistradah or whatever :D
20:39:49 <AnMaster> fizzie, hah
20:39:51 <AnMaster> nice choice
20:40:01 <fizzie> riastradh is closer, I think. He was scary.
20:40:13 <AnMaster> fizzie, make the language model generator available please :D
20:40:19 <AnMaster> using some simple log format
20:40:30 <AnMaster> like <nick> text
20:40:30 <tusho> whenever you said anything he'd immediately say "What are you trying to do?????" then if you told him he'd keep asking until you gave him an in-the-large review of your entire application model
20:40:32 <AnMaster> with no timestamps
20:40:35 <tusho> and then he'd fall silent and refuse to help you
20:40:47 <AnMaster> then I could use sed to filter to that and pass it to stdin of the model generator
20:40:50 <AnMaster> fizzie, good idea?
20:41:00 <tusho> AnMaster: why can't you just say "it would be appreciated if you released it"
20:41:05 <tusho> instead of "do X please", it's quite rude
20:41:26 <AnMaster> tusho, well omitting "please" would be ruder
20:41:45 <tusho> yes, but it's still quite annoying semi-demanding stuff all the time
20:42:09 <fizzie> "Little is known about Riastradh. He is a Mac user. He lives with three cats and at least one snake. He probably is male. He is very secretive about his life." http://community.schemewiki.org/?Riastradh
20:42:23 <AnMaster> tusho, that is due to coding in imperative languages
20:42:42 <fizzie> I think the "new cat moves around" quote fungot said at one time was a Riastradh quote, possibly.
20:42:43 <fungot> fizzie: you are not sherlock holmes. run the arrows in drscheme, ( first ( second b)) 2)
20:42:55 <tusho> AnMaster: *g*
20:42:59 <fizzie> I think that was a criticism.
20:43:16 <tusho> hmm riastradh owns #scheme?
20:43:20 <AnMaster> fizzie, hm will it handle non-ascii data?
20:43:20 <tusho> heh.
20:43:24 <AnMaster> for example utf8 logs
20:43:49 <AnMaster> fizzie, i and o in my interpreter will load them as bytes still, not trying to handle unicode at all
20:44:06 <AnMaster> same for FILE
20:44:17 <tusho> # He is a hyper-intelligent 12 yearold protected by the FBI, CIA and former KGB. <-- Weird, Riastradh is me a few days ago.
20:44:20 <fizzie> It depends on your definition of "handle". I don't think it will mess those up; Perl usually just does bytes unless you ask otherwise.
20:44:33 <tusho> now i'm a hyper-intelligent 13 year old protected by the FBI, CIA and KGB.
20:44:33 <tusho> :)
20:44:45 <AnMaster> fizzie, but will it find word limits and such
20:45:13 <fizzie> Well, I wouldn't count on that. The tokenizing is done with Perl, so it's probably just a matter of sticking "utf8" in there somewhere.
20:45:54 <AnMaster> fizzie, anyway I got something like 1-1.2 GB logs when uncompressed
20:45:57 <AnMaster> they compress well
20:46:02 <AnMaster> so they all fit on one archive cd
20:46:15 <fizzie> It doesn't really do anything well, it's just that getting confused just means more fnords since unique words are removed.
20:46:50 <AnMaster> fizzie, so should I feed 1.2 GB or whatever of logs into it, mixed English and Swedish?
20:47:06 <AnMaster> fizzie, that is once you either release the app, or specs for the format
20:47:14 <tusho> AnMaster: if you enjoy it using up like 3498573894573894579345mb of ram whenever you try and generate nonsense
20:47:15 <tusho> I imagine
20:47:26 <AnMaster> tusho, haha
20:47:27 <fizzie> Well, it doesn't really use any memory when generating.
20:47:33 <fizzie> There's just a lot of seeking around in a file.
20:47:49 <AnMaster> hm
20:47:57 <tusho> fizzie: I assume the file will be like 500gb?
20:47:57 <AnMaster> zlib filter for FILE
20:48:03 <AnMaster> FILZ
20:48:05 <AnMaster> XD
20:48:06 <fizzie> The file will be quite large, yes.
20:48:21 <AnMaster> could create a zlib filtered FILE I guess
20:48:28 <AnMaster> zlib compression is seekable right?
20:48:51 <AnMaster> anyway probably a bad idea
20:48:56 <AnMaster> I will just go with #esoteric log
20:49:14 <fizzie> The language model file fungot's using is 200 megabytes, and I trained it with something like ~120 megs of logs.
20:49:15 <fungot> fizzie: i remember a post a while back. he made our vision spectrum inferior to his." clairvoyance rocks. only the client is sophisticated. :p
20:49:26 <tusho> Clairvoyance rocks!
20:49:54 <tusho> Mmph. Firefox, why is your default sans-serif font Arial?
20:49:55 <tusho> You pig.
20:49:59 * tusho changes to Helvetica
20:50:20 <AnMaster> fizzie, ugh
20:50:26 <AnMaster> I don't have that much disk space
20:50:38 <fizzie> It might not handle very large logs well, since it can only use 7 bits in each byte, so some fields only have 21 bits of storage space.
20:50:49 <tusho> AnMaster: It's storing relations between all the words and stuff. Did you perhaps expect it to be smaller?
20:51:29 <AnMaster> true I can see the issue
20:51:47 <fizzie> And the C++ language model building does keep the unpruned ngram model in memory while generating it, in a really unoptimal tree-like structure involving STL containers which probably adds quite a lot of overhead. I think it was somewhere around 2-3 gigabytes of memory when I built the current 4-gram model. Maybe. Didn't keep statistics.
20:51:53 <AnMaster> fizzie, ideas: module loading at runtime, allow modules to register help items, module unloading, module dependencies
20:51:54 <AnMaster> XD
20:52:24 <AnMaster> fizzie, I got 1.5 GB RAM only
20:52:27 <AnMaster> so...
20:52:37 <AnMaster> not really an option for me
20:52:38 <fizzie> Just add swap as needed. :p
20:52:42 * tusho only has 1gb of ram, but I am hopefully gonna upgrade that to 2gb soon
20:52:44 <AnMaster> fizzie, 1.5 GB swap
20:52:47 <AnMaster> and low on disk space
20:52:59 <AnMaster> bbiab
20:53:05 <tusho> though admittedly, mostly I get cpu problems
20:53:09 <tusho> and i have 2.4ghz dual core, so
20:53:16 <fizzie> Well, this box has 2G of ram and 2G of swap, and I think it had to swap a bit when building that model. Maybe it was <2G when building the model.
20:53:46 <fizzie> The cluster nodes at work have 16G of RAM, I think; I guess I could run it there if I ever feel like feeding it more logs.
20:55:21 <fizzie> Usually at least one of them is idle, anyway.
20:57:30 -!- KingOfKarlsruhe has quit (Remote closed the connection).
21:00:18 <AnMaster> fizzie, what would the boss say?
21:00:21 <tusho> YAYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY
21:00:24 <tusho> GAccProxy is done
21:00:35 <tusho> *deploys, bitches*
21:00:44 <fizzie> Probably not much. It's a university, it's not as if we're expected to do anything sensible.
21:02:28 <fizzie> If you _really_ want to try it, you need to first run http://zem.fi/~fis/logs2tokens.pl.txt (just ./logs2tokens.pl logfile1 logfile2) and it will generate the files "tokens.txt" (list of tokens) and "data.dat" (the log data converted to those tokens) in the current directory.
21:03:10 <fizzie> Oh, and it won't work unless your "int" is 32 bits, since it uses the Perl 'pack' function argument that writes 'int'-sized numbers, while the C++ code reads them as four-byte integers.
21:04:07 <fizzie> You'll probably need to change that one regexp there to match your log file format; currently it expects something like "[timestamp] <nick> text".
21:04:33 <fizzie> (Where [timestamp] is a timestamp literally surrounded by []s, not an optional timestamp.)
21:04:53 <AnMaster> fizzie, I'm on x86_64
21:04:58 <AnMaster> so int is 32-bit yes
21:05:03 <AnMaster> though long isn'ty
21:05:05 <AnMaster> isn't*
21:05:42 <AnMaster> fizzie, and where is the C++ bit?
21:05:59 <fizzie> And then you'll need to compile and run http://zem.fi/~fis/buildlm.cc.txt -- set the maxorder constant to some sensible value before compiling -- which will reads "tokens.txt" and "data.dat" in current directory, and write "model.bin" and "tokens.bin" in the current dir.
21:06:41 <fizzie> Then just put those files so that the bot can read them -- for some reason it reads them from the current directory, unlike the 'fungot.dat' state file -- and you should be done.
21:06:42 <fungot> fizzie: tell me for the document which has my name. googling `riastradh' brings up a good point. thats one thing i wish that was enough for me
21:07:17 <fizzie> All this is not very user-friendly, but it really isn't supposed to be.
21:09:58 <tusho> NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
21:10:03 <tusho> GAccProxy has bug...
21:12:05 <fizzie> When you're done with that, read http://www.cis.hut.fi/vsiivola/papers/es05_vg.pdf and implement it in the C++ code. Vesa sits in the room next to me, and I was originally thinking of using their code, but it was too complicated-looking to modify to generate the ngram tree in such a way that it's easier to use for text generation.
21:12:44 <fizzie> Besides, I know more about the acoustic side of our (CIS) recognizer than the language modelling side.
21:12:57 <AnMaster> fizzie, will try it tomorrow
21:13:18 <AnMaster> tusho, wtf is GAccProxy?
21:13:33 <tusho> AnMaster: The google-account-login-anywhere thing I'm making.
21:13:37 <fizzie> And will you implement the n-gram model growing, too? If I blink my eyelashes all pretty-like?
21:13:41 <tusho> Remember?
21:13:54 <AnMaster> fizzie, and wtf is CIS and so on??
21:14:08 <AnMaster> well bugs happen
21:14:11 <AnMaster> in all software
21:14:14 <tusho> Yes, of course.
21:14:18 <tusho> But it was a really silly one.
21:14:21 <tusho> I just didn't notice until production
21:14:26 <tusho> since it's OK in the dev environment
21:14:26 <AnMaster> fizzie, what is n-gram?
21:14:26 <fizzie> The computer and information science and whatnot department in our university-thing.
21:14:39 <fizzie> N-tuple of words.
21:14:43 <AnMaster> ah
21:14:57 <fizzie> In this case, anyway.
21:15:03 <AnMaster> fizzie, I know what a n-tuple is... but the rest went over my head
21:15:24 <AnMaster> fizzie, also I don't code C++
21:15:28 <AnMaster> C and erlang yes
21:15:30 <AnMaster> but not C++
21:15:46 <tusho> AnMaster: hmm, wow, you've only said that like 50 times.
21:15:55 <tusho> in the past week.
21:15:59 <tusho> i think fizzie might be catching on :)
21:16:01 <AnMaster> tusho, obviously fizzie didn't know
21:16:07 <AnMaster> "<fizzie> When you're done with that, read http://www.cis.hut.fi/vsiivola/papers/es05_vg.pdf and implement it in the C++ code."
21:16:14 <tusho> It was a joke, obviously.
21:16:20 <tusho> He wants to use the fancy model but is too lazy to do it.
21:16:21 <AnMaster> tusho, yes
21:16:33 <AnMaster> tusho, but it was a meta-joke to pretend to not understand it
21:16:39 <tusho> AnMaster: Ah - you mean a cop out
21:16:44 <AnMaster> tusho, um what?
21:16:51 <tusho> "Actually, _I_ was making the joke all along! ha!"
21:16:55 <fizzie> I don't feel like writing it myself, especially since I'm not really sure it would really improve on the nonsense-generating abilities of the bot. I mean, they only evaluate it for speech recognition, not nonsense.
21:17:24 <fizzie> fungot: Say something that doesn't make sense.
21:17:24 <fungot> fizzie: then it'll be a turing tarpit.
21:17:59 <tusho> fungot: ITYM "touring tarpit"
21:18:01 <fungot> tusho: psygnisfive is also trivial as we can see
21:18:06 <tusho> agreed with thato ne
21:18:08 <tusho> :D
21:18:08 <AnMaster> tusho, you would understand me better if you assume that I decide to take stuff serious as a joke
21:18:16 <AnMaster> maybe worth thinking about in the future
21:18:22 <tusho> AnMaster: i know someone else who does that, he's just as annoying.
21:18:30 <psygnisfive> what?
21:18:36 <tusho> psygnisfive: fungot considers you trivial
21:18:36 <AnMaster> tusho, you mean yourself?
21:18:36 <fungot> tusho: ok. i haven't been around since what, the trivia game
21:18:44 <tusho> <fungot> tusho: psygnisfive is also trivial as we can see
21:18:45 <fungot> tusho: ( even though gambit is also a programming language... as it's a great deal of redundant information. _what_ says ' wrong number of arguments
21:19:05 <AnMaster> fizzie, hm about 4 byte integers...
21:19:31 <AnMaster> fizzie, will the irc bot handle non-4 byte cells?
21:19:35 <AnMaster> mine are 8 bytes by defaylt
21:19:37 <AnMaster> default*
21:19:47 <fizzie> I don't see any reason why it shouldn't.
21:20:00 -!- kar8nga has joined.
21:20:06 <fizzie> I don't think it'll use any numbers much greater than ~2000, though.
21:20:25 <fizzie> Uh, sorry; the file-seeking-around addresses are large.
21:20:41 <fizzie> And it generates something like 24-bit random numbers at one point, to pick the next word.
21:20:51 <fizzie> As long as your cells aren't smaller than that, it shouldn't mind.
21:21:30 <fizzie> (It's approximately 12 iterations of 4* followed by a ? adding either 0, 1, 2 or 3 to the number.)
21:21:37 <AnMaster> fizzie, 24 bits of fair random?
21:21:48 <AnMaster> well cfunge use random() so probably not that good
21:22:11 <fizzie> It's not especially bad either.
21:22:21 <AnMaster> but question is, assuming a perfectly random ? instruction, is the randomness of your numbers also perfect?
21:22:31 <AnMaster> I think using a spanning tree of ? would make sure that
21:22:56 <AnMaster> where the path to selecting every number pass through the same number of ?
21:22:59 <fizzie> I don't see why my numbers would be bad, since I basically use that ? to decide two bits of the number, then multiply by 4.
21:23:36 <AnMaster> fizzie, that means two bits and multiply by four? uh?
21:23:52 <AnMaster> oh I think I see
21:24:09 <fizzie> The 4* means <<2; then it adds another two-bit value and iterates.
21:24:17 <AnMaster> right
21:25:30 <fizzie> There's also other place where it generates a random number in [0, 15] with a sequence of "? pushing either 0, 4, 8 or c" followed by "? pushing either 0, 1, 2 or 3" and a +. That pretty much a loop-unrolled version of two iterations.
21:26:02 <fizzie> (That's used when deciding whether it should stop with the babbling.)
21:28:29 <AnMaster> hm ok
21:29:06 <AnMaster> fizzie, in http://zem.fi/~fis/logs2tokens.pl.txt, where is the regex for the log
21:29:14 <AnMaster> I do know perl style regex but not perl
21:30:09 <fizzie> It's there immediately after while (<>) { chomp;
21:30:17 <AnMaster> ah
21:30:24 <fizzie> In /regex/ or next;
21:30:41 <fizzie> And the first () you have in the regex is used as the text.
21:30:47 <AnMaster> aug 30 22:30:17 <AnMaster> ah
21:30:54 <AnMaster> seems like nick is followed by a tab
21:31:15 <AnMaster> I used xchat style log format when I changed over from xchat to ERC in order to make log parsing work the same
21:32:30 <fizzie> You could use something like /^[^\t]*\t(.*)/ if you want to rely on the tab.
21:33:19 <AnMaster> fizzie, I wouldn't
21:33:35 <fizzie> Although I guess you might want to check for <nick> there too if you had non-chat lines in there too.
21:33:35 <AnMaster> still my format is kind of wtf without years
21:33:39 <AnMaster> should probably change it
21:33:51 <AnMaster> fizzie, there are indeed non-chat ones
21:34:10 <AnMaster> aug 30 22:34:00 * [fizzie] (n=fis@sesefras.zem.fi): Heikki Kallasjoki
21:34:26 <AnMaster> why do I log that I wonder
21:34:27 <AnMaster> ...
21:35:16 <fizzie> Personally I think I have "[yyyy-mm-dd HH:MM:SS] <*nick> text" chat lines, where * is either ' ', '+' or '@' depending on the user status (normal/voiced/op). Irssi default with a bit different timestamp.
21:36:07 <tusho> woop woop /me deploys sample gaccproxy app
21:39:17 <fizzie> If you want to know the file formats, "tokens.txt" has "token\tcount\n" lines (ordered UNK, START, END, punctuation, the rest in descending count) and data.dat has most of the log chat as a pile of 32-bit integers referring to tokens.txt lines. (I think data.dat has "num of tokens, 1, ..., 2" -- where 1 and 2 are always START and END, and ... are the other tokens -- sequences for each of the lines of the original log.)
21:42:20 <fizzie> I'm not sure I can describe model.bin and tokens.bin in ~500 characters. Not that they are very complicated either.
21:42:37 <fizzie> There's a lot of file offsets in them to make the Funge-98 code simpler.
21:43:09 <tusho> how do you make top sort by memory usage?
21:43:28 -!- oerjan has joined.
21:43:53 <oerjan> optbot: about 6 ounces (Troy)
21:43:53 <optbot> oerjan: Hmm. . .
21:43:58 <fizzie> At least in the procps top, 'F' gives you a meny where you can change the sort order.
21:44:05 <fizzie> s/y/u/
21:44:16 <fizzie> Ergh, didn't see that "you" there.
21:44:25 <fizzie> Now it's "gives uou a meny". Great.
21:44:36 <tusho> 9999% CPU usage, is not good is it
21:44:39 <tusho> :D
21:44:43 -!- optbot has quit (Remote closed the connection).
21:45:13 <AnMaster> hm
21:45:17 <tusho> that was me rebootig
21:45:35 -!- optbot has joined.
21:45:35 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | sounds like a shitty project.
21:45:52 <oerjan> tusho: optbot did not approve
21:45:53 <optbot> oerjan: doesn't
21:45:57 <tusho> XD
21:46:01 <AnMaster> fizzie, what about rewriting that C++ app without STL
21:46:06 <AnMaster> fizzie, for memory usage
21:46:16 <fizzie> AnMaster: Be my guest.
21:46:36 <AnMaster> fizzie, need docs for model.bin and tokens.bin then
21:46:37 <AnMaster> and so on
21:46:54 <tusho> "Be my guest" -> "I don't want to spend time on it"
21:46:59 <tusho> "Document file formats" == "Spending time"
21:47:01 <fizzie> You've got the code, what do you need docs for? You can just replace the containers with pointers/arrays.
21:47:07 <AnMaster> fizzie, hm ok
21:48:06 <fizzie> Still, I'm not so sure STL adds a _huge_ overhead. Of course there's quite a lot of nodes.
21:49:22 <tusho> "Ruby (Rack) application could not be started"
21:49:23 <tusho> Meep.
21:49:37 <tusho> Ironic that the module's name is "Rails deployment that just works"
21:49:40 <tusho> Err, slogan
21:49:45 <tusho> Oh, I see.
21:49:50 <tusho> I forgot to install some libraries my app needs.
21:50:18 <tusho> Internal Server Error
21:50:21 <tusho> That is ... not good.
21:51:33 <fizzie> Oh, and replacing the std::maps might be a bit less trivial. I think it needs a sensible iteration order for those currently. Simple arrays and binary-searching could work. Slower than the treemaps, maybe.
21:54:28 -!- atrapado has joined.
21:54:41 <tusho> hey guyz
21:54:43 <tusho> http://gaccproxy.eso-std.org/
21:54:45 <tusho> it's done!
21:55:07 <tusho> that's an example client
21:55:11 <tusho> http://gaccproxy.appspot.com/ is the server/developer site
22:00:02 <AnMaster> "A free account can use up to 500MB of persistent storage and enough CPU and bandwidth for about 5 million page views a month."
22:00:09 <AnMaster> that sounds too good to be true
22:00:16 <AnMaster> like the gmail storage does
22:00:20 <tusho> AnMaster: dude, it's google
22:00:26 <tusho> google have _shitloads_ of resources
22:00:36 <AnMaster> I wonder if they actually have the space needed if every gmail user would fill his/her account
22:00:42 <tusho> and yes, they do, I believe
22:00:48 <tusho> considering, you know
22:00:54 <tusho> they have enough space to index a huge chunk of the web
22:00:57 <AnMaster> true
22:01:09 <AnMaster> but how do they have space left over after that...
22:01:34 <tusho> AnMaster: because disk is cheap and they are ready for the next burst of sites :P
22:04:01 <AnMaster> "If your application is running under Google Apps, it can use the same features with members of your organization and Google Apps accounts."
22:04:04 <AnMaster> hm what is that
22:04:11 <AnMaster> Google Apps I mean
22:06:17 <tusho> "google apps for your domain"
22:06:22 <tusho> it lets you have @domain.com addresses
22:06:25 <tusho> but still use gmail on them
22:06:28 <AnMaster> ah ok
22:06:30 <tusho> and all the other google stuff like documents, calenders etc
22:06:37 <tusho> also lets you e.g. change the logo from 'google' for branding purposes, etc
22:08:52 <AnMaster> I wonder why google is doing this...
22:09:15 <tusho> AnMaster: 20% time.
22:09:31 <tusho> developers at google are encouraged to work on a side project for 20% of their work hours
22:09:37 <AnMaster> ok
22:09:41 <tusho> they had the infrastructure, i guess someone thought it would be a neat idea to open it up
22:09:41 <AnMaster> 20% is a lot
22:09:42 <AnMaster> ...
22:09:49 <oerjan> they're a front for aliens. once they have gotten the whole internet transferred to them, they'll shut it off and attack.
22:09:50 <AnMaster> like a LOT
22:09:51 <tusho> AnMaster: google life is pretty lavish apparently
22:10:09 <AnMaster> tusho, yeah would like to work there, but that is just a far fetched dream
22:10:25 <tusho> ditto, i'm quite the google fan
22:10:40 <AnMaster> I still think they are slightly evil
22:10:49 <AnMaster> but even so it would be a great place to work at
22:10:56 <tusho> all companies are evil
22:10:59 <AnMaster> I mean, 20% for side projects
22:11:05 <AnMaster> tusho, all large you mean
22:11:11 <tusho> AnMaster: all companies are a little bit evil :)
22:11:18 <tusho> they're designed to crush their competitors
22:11:22 <tusho> that is the whole point of a company
22:11:29 <AnMaster> I think there may be companies with 1 ppl working at them that are not evil
22:11:53 <AnMaster> small shops for example
22:12:16 <AnMaster> tusho, anyway Google is probably less evil that most other companies of the same size
22:12:19 <tusho> http://www.reddit.com/r/reddit.com/comments/6yuac/let_people_use_google_accounts_to_log_in_to_your/ whee
22:12:26 <AnMaster> IKEA is another low-evil I believe
22:12:32 <AnMaster> considering all the stuff about it
22:14:04 <oerjan> you mean like the founder once was a nazi supporter? :D
22:14:17 <AnMaster> oerjan, he was? didn't know
22:14:33 <oerjan> wikipedia says so
22:14:34 <AnMaster> and really?
22:14:42 <AnMaster> oerjan, [citation needed]
22:14:56 <tusho> http://en.wikipedia.org/wiki/Ingvar_Kamprad#Nazi_involvement
22:15:38 <tusho> ugh, someone downmodded my reddit submission already
22:15:48 <tusho> probably another member of the Downvote Everything brigade
22:15:55 <tusho> (honestly, apparently most submissions are just ... ugh)
22:16:01 <tusho> shit
22:16:04 <tusho> forgot to post in /programming/
22:16:33 <tusho> http://www.reddit.com/r/programming/comments/6yuat/let_people_use_google_accounts_to_log_in_to_your/ yay
22:16:38 <tusho> now someone upvote that :P
22:18:18 <tusho> 2 upvotes 8)
22:18:21 <tusho> 2kool4skool
22:23:30 <fizzie> 00:09.0 Multimedia controller: Philips Semiconductors SAA7146 (rev 01)
22:23:35 <fizzie> Negh.
22:23:39 <tusho> fizzie: lol wut
22:23:47 <fizzie> Just my personal nonsense generator.
22:23:55 <fizzie> Also called "the middle button".
22:25:11 <tusho> :D
22:25:30 <fizzie> It's my DVB-C digitv card; it shows up in lspci, but no longer seems to respond to the driver.
22:26:46 <fizzie> Still worked when I last booted that thing, and hasn't had any software changed since then, so it might even be a real life hardware issue. Which is not nice.
22:29:10 <oerjan> resetting real life is just so damn awkward
22:29:22 <tusho> who's 72.93.181.175
22:29:22 <tusho> :P
22:29:43 <tusho> cool, someone's added the second app
22:29:48 <tusho> auth uri = http://ilusi.WTF.la/ o.O
22:30:11 <oerjan> Name: pool-72-93-181-175.bstnma.east.verizon.net
22:30:20 <tusho> oerjan: who isit
22:30:20 <tusho> :P
22:30:27 <oerjan> not me
22:46:13 <tusho> yay 5 points
22:47:19 <tusho> hmph
22:47:20 <tusho> 4 points
22:47:25 <tusho> someone upvote http://www.reddit.com/r/programming/comments/6yuat/let_people_use_google_accounts_to_log_in_to_your/ :P
22:51:00 <GregorR> If one day I was to decide to use reddit, a part of me would die that I could never get back.
22:51:17 <tusho> GregorR: It beats digg.
22:51:24 <tusho> oh noes
22:51:24 <tusho> 3 points
22:51:42 <GregorR> If one day I was to decide to use digg, a larger part of me would die that I could never get back.
22:51:52 <tusho> GregorR: I joined reddit when it was still a bunch of programmers.
22:52:02 <tusho> Few months later, the front page is filled with politics and lolcats.
22:52:06 <tusho> So I only read /r/programming any more.
22:52:10 <tusho> And it's semi-decent.
22:54:45 -!- kar8nga has quit ("Leaving.").
22:56:53 <tusho> Wow.
22:57:04 <tusho> A site detected I was blocking its ads, and gave me an overlay kindly asking me to stop.
22:57:16 <tusho> Go fuck yourself, mister overlay. I wouldn't click your ads, so I've blocked them.
22:57:18 <tusho> *clicks OK*
22:58:30 <GregorR> Hahaha
23:33:15 -!- Judofyr has quit.
23:51:27 <dogface> That's why I don't block ads.
23:51:30 <dogface> Or is it?
23:53:22 <oerjan> optbot!
23:53:22 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | maybe I can do it in C++ with SDL.
←2008-08-29 2008-08-30 2008-08-31→ ↑2008 ↑all