00:00:11 -!- distant_figure has quit (Read error: Connection reset by peer).
00:01:21 -!- distant_figure has joined.
00:04:05 <nooga> nerdymon in a nerdyball!
00:04:21 <Vorpal> (and yes I did get the poke reference)
00:04:28 <alise> pikhq: Please regularly remind me that Fine Structure comments are spoilery.
00:04:30 <Vorpal> (but um, couldn't think of a new way to continue it )
00:08:43 -!- sebbu2 has joined.
00:09:31 -!- distant_figure has quit (Ping timeout: 272 seconds).
00:10:02 -!- distant_figure has joined.
00:11:19 -!- sebbu has quit (Ping timeout: 252 seconds).
00:14:10 <Vorpal> Gregor, alise: "* GregorR-L forces it into EgoBot >: )"
00:15:05 <alise> Also illegal in 41 states.
00:15:54 <Vorpal> alise, yes he was forcing "Furryscript" into it
00:16:06 <Vorpal> alise, which makes it even more kinky I guess XD
00:16:19 <Vorpal> alise, furryscript being zzo's scripting language
00:16:41 -!- distant_figure has quit (Ping timeout: 240 seconds).
00:16:49 <Vorpal> AssertionError: broken timestamp: UNLOAD <filename|name> <-- indeed, and I hate xchat that I was using back then
00:17:08 <nooga> this word reminds me this weird type of pornography that is somewhat popular (and hated) on 4chan
00:17:14 -!- distant_figure has joined.
00:19:01 <alise> Vorpal: Furryscript's main competitor: http://esolangs.org/wiki/FURscript
00:19:15 <alise> (Quite infamous, as I'm sure you know.)
00:19:18 <alise> (Also: meant to be serious.)
00:19:50 <alise> "The person who designed this language was 100% serious about it and the vb6 compiler, but I think he got as far as a text box and a copyright notice before going back to programming his graphics calculator. --Einsidler 10:44, 24 Nov 2006 (UTC)"
00:23:16 -!- distant_figure has quit (Remote host closed the connection).
00:23:57 <nooga> we have even a suitable category
00:24:02 -!- distant_figure has joined.
00:27:26 <nooga> i'd love to design another esolang but it's probably impossible to make something completely new
00:29:01 <alise> pikhq: Why has he just stuck all his short stories together and called them a novel?
00:29:04 -!- distant_figure has quit (Remote host closed the connection).
00:29:07 <alise> (Note: I realise it'll change as I keep reading.)
00:29:23 -!- distant_figure has joined.
00:30:19 -!- Flonk_ has joined.
00:31:49 <pikhq> alise: It shows that he's new to novels.
00:32:12 -!- Flonk has quit (Ping timeout: 252 seconds).
00:32:15 -!- Flonk_ has changed nick to Flonk.
00:32:27 <alise> pikhq: No. The Ed stories flowed better than this. If I was reading this without knowledge of their prior status, I'd assume he's up to something.
00:32:30 <alise> pikhq: It is complete now, right?
00:32:36 <nooga> alise: last time i discovered something exciting... it turned out to be a perfectly normal lisp
00:32:42 <alise> nooga: Heh; what was it?
00:33:07 <nooga> i tried to design something like ais' Feather
00:33:44 -!- augur has quit (Remote host closed the connection).
00:34:42 <pikhq> alise: It also gets more flowing later...
00:35:22 <alise> pikhq: Good. I just have to ask this, and please answer as vaguely as possible -- a simple "yes" or "no" will be fine -- do these disconnected fragments actually get picked up again, or are they just there to fuck with me?
00:36:04 <alise> Hey, my answer questioned: there's Seph again.
00:36:07 -!- distant_figure has quit (Ping timeout: 272 seconds).
00:36:10 <Sgeo|web> You're finally reading Fine Structure?
00:36:28 -!- Flonk has quit (Remote host closed the connection).
00:36:28 <pikhq> alise: They get picked up again.
00:36:50 -!- distant_figure has joined.
00:36:52 <alise> I have a horrible urge to write some sci-fi, and I'm suppressing it as well as I can because I know it'd be terrible.
00:37:28 <alise> [["Yes," says Mitch. "I can walk through stuff. And myself." He passes his wrists through each other.
00:37:28 <alise> "Don't-- don't do that," says Seph, "it could be bad for you."]]
00:37:45 <alise> (I've read this chapter before. Actually I think I've read all of these chapters before. I might not have read one.)
00:37:48 <alise> Sgeo|web: Oh, chapter. Right.
00:37:55 <alise> Indistinguishable from magic.
00:38:01 <alise> *"Indistinguishable from magic".
00:38:20 <alise> *I might not have read one of them before.)
00:38:34 <alise> That is, there is one of them that I might not have read; not "it is a possibility that I have not read any of them before".
00:39:47 <Sgeo|web> Up to a certain point, I presume
00:41:16 -!- Sgeo|FsckPuppy has joined.
00:42:11 <alise> Right, that's what I mean.
00:42:17 <alise> The standalone stories that have been integrated into the timeline.
00:42:30 <alise> Sgeo|FsckPuppy: Why haven't you switched to Ubuntu yet?
00:42:44 <alise> "Indistinguishable from magic" is the first one that has never been standalone yet, I think.
00:43:07 <alise> "Power Of Two" was a directory but not inside Fine Structure to start with, I think.
00:43:36 <alise> "Indistinguishable from magic" has always been part of Fine Structure, though, to my knowledge. Nonetheless, I have read it. I believe it was the first story for a while.
00:43:39 <nooga> alise: nothing, at one point i proved that my meta abstractions system is just lisp's macro system
00:44:15 -!- Sgeo|web has quit (Ping timeout: 252 seconds).
00:44:16 <Sgeo|FsckPuppy> Power of Two was once taken offline due to someone claiming it could get published, iirc
00:44:49 <alise> [[2008-07-18 18:30:49 by Paradoxia:
00:44:50 <alise> Ah, I see what you're saying. He **REDACTED**
00:44:50 <alise> 2008-07-18 18:59:56 by Sam:
00:44:50 <alise> Paradoxia, I edited your comment because it gives away plot details from later in the story.]]
00:44:53 <alise> Sgeo|FsckPuppy: Yes, that's what I mean.
00:45:07 <alise> The two teleportation stories were outside at first, too, I think.
00:45:37 <Sgeo|FsckPuppy> Might nto be the best idea to read comments before you've read the whole thing
00:45:49 <alise> And now I'm turning into Teal'c.
00:46:02 <alise> Sgeo|FsckPuppy: "Unbelievable scenes" is a bit of a bad opening because I thought "FUCK YEAH I hope the next few chapters are like this too".
00:47:13 <Sgeo|FsckPuppy> I saw all the speculation about USIS, but... I'm shutting up now
00:48:08 <alise> USIS? Would it be a spoiler to expand that acronym?
00:49:21 <alise> Why did you think it was "in space"? :P
00:50:54 <Sgeo|FsckPuppy> I was about to say the name, but if you twisted it just right, you'd get a spoiler
00:50:55 <alise> Sgeo|FsckPuppy: Let me guess: Spoilers?
00:51:07 <alise> Is "Unbelievable scenes in space" a spoiler?
00:51:24 <Vorpal> alise, "Unbelievable scenes in spoilers"
00:51:26 <alise> Actually, I should probably just remember your track record with Fine Structure spoilers and ignore you until I finish it.
00:51:36 <alise> But I'm too foolish and kind-hearted to.
00:52:13 <alise> But you will link to a ROT-13 decoder and say a bunch of stuff in ROT-13, and then say "...which is how Fine Structure ends".
00:52:34 * alise wonders why the story's continuing as normal when all the stars and shit have been blinked out.
00:52:39 <alise> (Yes, yes, I know, non-chronological order.)
00:52:56 <alise> (Fabula and sujet and all that.)
01:03:47 <alise> Well, then, now you know.
01:04:00 <alise> Hey, and Power Of Two reaches its next chapter: Exponents.
01:04:08 <alise> Don't like that capital O. I don't like it.
01:04:53 <alise> And the characters connect again!
01:14:55 -!- BeholdMyGlory has quit (Remote host closed the connection).
01:18:32 -!- distant_figure has quit (Remote host closed the connection).
01:19:21 -!- distant_figure has joined.
01:23:35 <alise> Sgeo|FsckPuppy: Is Ching's first name Kuang?
01:25:40 -!- Wamanuz has quit (Remote host closed the connection).
01:25:55 <alise> Sgeo|FsckPuppy: I know FTL gets invented, but I've forgotten the name of the guy who does it already. :) Bloody comments.
01:25:59 -!- cheater00 has quit (Ping timeout: 272 seconds).
01:26:16 -!- distant_figure has quit (Ping timeout: 276 seconds).
01:26:21 <alise> Unless it just meant whoever invented the device here on Earth in On Digital Extremities and onwards.
01:26:27 <alise> Huh. Alright then.
01:26:42 -!- Wamanuz has joined.
01:27:02 -!- distant_figure has joined.
01:27:11 <Sgeo|FsckPuppy> Oh, and you may be thinking of a story that's retconned ouyt
01:29:10 <alise> Sgeo|FsckPuppy: I have not read the story, as I explained.
01:29:12 <alise> So thanks for the spoiler.
01:29:27 <alise> Also, you just spoilered that it stays blocked forever.
01:29:46 <alise> [FTL gets invented, implicitly "sometime in the future"] "No it doesn't"
01:29:50 <alise> Provided as evidence:
01:29:51 <alise> <Sgeo|FsckPuppy> Oh, FTLC?
01:29:51 <alise> <Sgeo|FsckPuppy> Well, that's blocked, which you should know already
01:30:30 <Sgeo|FsckPuppy> Whether this means that it stays blocked or not... well, I'm not saying
01:32:29 <Sgeo|FsckPuppy> Don't attempt to decode, from what I've said, information about the future of the story
01:33:31 <alise> Sgeo|FsckPuppy: No, but you see, you reacted to "faster-than-light communication gets invented" with "no it doesn't". You then say "perhaps you meant [this form of it]?" and then state that it does not work; obviously, the answer you gave first is validated by your second negation, meaning that it never gets unblocked.
01:33:48 <alise> Anyway, maybe I should just try not to talk to you about Fine Structure and vice versa until I'm done with it and it is impossible to spoil me. :P
01:34:04 <alise> I still remember a bunch of people suffocate at one point.
01:36:15 <alise> Why do you want such a person?
01:36:24 <alise> (Wow, that's a pretty ghastly way to kill someone wrt. Eleven.)
01:36:32 -!- augur has joined.
01:36:58 <Sgeo|FsckPuppy> alise, warning: You do not want the answer to that question
01:38:12 -!- cheater00 has joined.
01:38:21 <alise> lol @ the use of BBC "News"
01:38:31 <alise> (The tendency of the BBC to put every damn thing in quotes in news headlines)
01:38:38 <alise> ("Two killed in "transporter accident"")
01:40:12 <alise> Only if you check really damn hard that it provides no possible inferrable information about the story to someone who has not read "Two killed in "transporter accident"" or beyond yet.
01:40:36 <Sgeo|FsckPuppy> I don't think "Stuff happens" being inferrable is that bad
01:41:22 <Sgeo|FsckPuppy> Outside of the official stories, but still canonical, is another thing in that format
01:41:33 <Sgeo|FsckPuppy> Whether there are more things of that format within the stories, I'm not sure
01:42:34 <alise> There is canon outside of the official stories?
01:42:39 <alise> Is that not a contradiction?
01:43:40 <alise> Sgeo|FsckPuppy: Is listing them a spoiler?
01:43:51 <nooga> well, i just typed it
01:44:27 <alise> Why are there so many fucking spoilers?
01:44:33 <alise> What are the extras? The notes below the stories?
01:44:42 <alise> Are the CHAPTER TITLES spoilers? I've read them.
01:44:57 <Sgeo|FsckPuppy> The thing listed asExtras, appendices, feedback (subdirectory)
01:45:54 -!- distant_figure has quit (Ping timeout: 240 seconds).
01:46:19 <alise> Sgeo|FsckPuppy: *as Extras,
01:47:21 -!- distant_figure has joined.
01:47:38 <nooga> distant_figure annoys me with this cycling
01:48:01 <alise> I think I like Mitch.
01:48:04 <alise> distant_figure: Stop it.
01:48:17 <alise> he lives in south africa
01:48:20 <alise> probably their power grid
01:48:36 <alise> he's mentioned it going off and on a lot before
01:48:49 <alise> iirc they have regular planned downtime of the power grid too :)
01:48:59 <nooga> i would buy a diesel generator in such case
01:49:26 <nooga> or even use my car to generate electricity for computers
01:49:50 <Sgeo|FsckPuppy> alise, I'd ask you a question, but you'd be able to ... um
01:50:20 <nooga> ooooooh, it's soo late
01:50:49 <alise> Sgeo|FsckPuppy: "I'd ask you a question, but you'd be able to" is not a spoiler.
01:50:53 <alise> So what the fuck is it now?
01:51:04 <nooga> sleep time, good night
01:51:15 <Sgeo|FsckPuppy> alise, yes it is. If you squint real hard and are a bit psychic
01:51:28 <alise> Sgeo|FsckPuppy: It really isn't.
01:52:24 <alise> Sgeo|FsckPuppy: It is a spoiler, yes?
01:52:46 <alise> Sgeo|FsckPuppy: Remember when I said we just shouldn't talk about it?
01:52:57 <alise> I am trying to enjoy a novel here. :P
01:53:16 <alise> Ooh a SUBDIRECTORY.
01:54:29 <alise> Arcologies sure feature a lot in this.
01:54:48 -!- zzo38 has joined.
01:55:02 -!- distant_figure has quit (Ping timeout: 264 seconds).
01:55:58 -!- distant_figure has joined.
01:58:00 <alise> http://qntm.org/structure, for what it's worth.
01:59:33 <alise> Sgeo|FsckPuppy: Am I meant to be completely disoriented, chronologically?
01:59:39 <Gregor> Maaaan, I own the domain name onero.us and I STILL do nothing useful with it.
01:59:47 -!- trinithis has joined.
01:59:52 <alise> Gregor: I'll buy it.
02:00:03 <Gregor> alise: No way! Also you can't, legally :P
02:00:13 <alise> Gregor: No, but you can let it drop. Or I could pay you fees. :P
02:00:20 <alise> (Let it drop in exchange for me paying you a lump sum, that is.)
02:00:37 <Gregor> alise: No, I mean you can't own it in that it's a .us domain and you're not a US resident. Not that this is really enforced, but /technically/.
02:00:56 <alise> Gregor: Well let's pretend I am a US resident.
02:01:13 <Gregor> alise: Then it's just the age problem :P
02:01:26 <alise> Sgeo|FsckPuppy: This thing must be set in an alternate universe; presumably 1970- is implying some sort of circa 1970 onwards setting here, and it's definitely not after the present day due to the vehicles mentioned, so...
02:01:32 <alise> Gregor: What age problem?
02:01:44 <alise> Do you actually own it or just rent it like everyone else?
02:01:46 <zzo38> O, it is a story about information? Is a book available?
02:01:54 <alise> zzo38: There are PDFs available that you could print out.
02:01:57 <Sgeo|FsckPuppy> Maybe you'll reorient yourself once you finish reading 1970-
02:02:06 -!- FireFly has quit (Quit: swatted to death).
02:02:07 <Gregor> alise: I just rent it like everyone else, but that's still a contract.
02:02:10 <alise> "A story about information" hasn't quite come together yet; but then, these are localised segments.
02:02:23 <alise> Gregor: Well, I happen to have purported to agree to several of them and nobody's g--
02:02:29 <alise> Gregor: I can flagrantly violate the terms of my domain contracts.
02:02:37 <zzo38> alise: OK, I might try the PDF
02:02:54 <alise> zzo38: There's two PDFs. I guess they're formatted differently.
02:03:06 -!- distant_figure has quit (Ping timeout: 265 seconds).
02:03:17 <alise> zzo38: I'd go for the second.
02:03:21 <alise> It's smaller, and better typeset.
02:04:08 <zzo38> alise: The files are a lot of pages. Where can I purchase this book?
02:04:24 <alise> zzo38: There is no printed copy as of yet. You could print the PDF, or just read it online.
02:04:33 <alise> Or you could use lulu.com to create your own print copy using that formatted PDF.
02:04:38 <alise> And then buy it... from yourself...
02:05:43 <zzo38> alise: I would rather buy it from the author, if I can. If I can somehow make something on lulu.com to make it pay someone else, and that I can indicate who to pay, for each book even in the same account to be able to put payment to different people
02:06:07 <alise> zzo38: You could set it to the minimum price so that all money goes to Lulu, and then send some money to Sam Hughes.
02:06:21 <alise> I imagine he'd give you an address to send a cheque or such to if you emailed him.
02:06:25 <zzo38> alise: But that won't do if someone else purchases the book too
02:06:34 <alise> zzo38: You can set it so that only you can see it.
02:06:38 <zzo38> Sgeo: I do sometimes read the log
02:06:41 <alise> Then nobody else would be able to purchase it, making it a personal copy.
02:06:56 <alise> His email is on this page, obfuscated for spambots: http://qntm.org/contact
02:07:08 <alise> "I am not looking to add advertisements to qntm.org at this time. Thanks for your interest." ;; guess he didn't bother to update it when he added Google ads
02:07:27 <zzo38> alise: But what if someone else wants to buy it, too?
02:07:48 <alise> zzo38: Then... they'll just have to do what you did. Or, far more likely, read it online or print out their own copy.
02:07:59 <alise> Or you could simply promise to give all profits from the book to Sam Hughes.
02:09:09 <alise> 361 pages; I would have expected the book to be longer.
02:09:33 <alise> Sgeo|FsckPuppy: Gosh, the start of Crash Zero is a bit boring.
02:09:58 <zzo38> alise: That is not a bad idea. But, if I run a business selling this book, I am not going to give *all* the profits to Sam Hughes (but I will give most of the profits). But when I have time to run a business, this can be one of the products. And then he can earn the money, too!
02:10:10 <alise> zzo38: But you wouldn't have done anything but fill out a form on Lulu.
02:10:18 -!- distant_figure has joined.
02:10:26 <alise> They process all orders, they print and mail out all copies, they take their chunk of the profit (plus however much you want, possibly nothing).
02:10:41 <alise> So taking profits would be a bit disingenuous... and illegal, too, without explicit consent from Sam Hughes.
02:10:55 <alise> Actually, so would publishing a public edition in the first place.
02:11:18 <zzo38> alise: I know. I can't publish it at all without permission, of course.
02:11:30 <zzo38> That is why I suggest I would rather I purchase it from *his* Lulu account.
02:11:30 <alise> Untrue; you can make a private copy for yourself on Lulu like I suggested.
02:11:46 <alise> He has stated he will not publish on any "vanity" press, i.e. self-service presses.
02:11:48 <alise> This includes Lulu.
02:11:55 <alise> So you're out of luck on that front.
02:13:14 <zzo38> alise: Ah, I guess I can make a private copy to myself and then delete it. I might consider doing that.
02:15:41 -!- distant_figure has quit (Ping timeout: 240 seconds).
02:15:42 <zzo38> But not yet. It will be much later.
02:16:53 -!- distant_figure has joined.
02:17:21 <alise> zzo38: I suggest doing what Sgeo|FsckPuppy just suggested.
02:17:52 -!- Sgeo|FsckPuppy has changed nick to Winston.
02:18:08 -!- Winston has changed nick to Sgeo|web.
02:29:49 <zzo38> Even if I read it online, I will not do so right now.
02:30:49 -!- EOF has joined.
02:31:04 <EOF> still sorting those logs?
02:31:38 <Sgeo|web> EOF, have you read Fine Structure?
02:31:54 <Sgeo|web> I don't know how trustworthy you are
02:32:20 <EOF> no i haven't
02:32:50 <EOF> i'll go through the wikipedia article :)
02:32:57 * pikhq declares Japanese rap "weird".
02:33:02 <alise> EOF: There isn't one! Ha!
02:33:19 * EOF seconds prior /m statement
02:34:00 <pikhq> http://bit.ly/8PRw6j I mean seriously, whaaat?
02:34:03 <EOF> http://eo.wikipedia.org/wiki/Maldika_strukturo in Esperanto
02:34:42 <pikhq> EOF: Mia Esperanto malbonas. Japanalingvo?
02:34:43 <alise> pikhq: They're not even trying
02:35:27 <pikhq> alise: Would you expect them to be doing stereotypical urban black culture?
02:35:44 <alise> pikhq: I meant the actual rapping.
02:35:51 <zzo38> What I can do is make a file for TeX so that you can use UTF-8. It should not be too complicated to do. Simply to make all characters 128 to 255 active and then make table which character in a font corresponds to which character in a different font, that should do it
02:35:59 <alise> As far as I can make out he's just picking the shortest words with the easiest sounds to end every line with :P
02:36:04 <alise> Not that I can make out much of it at all, of course.
02:36:09 <alise> Not knowing Japanese and all.
02:36:42 <EOF> http://tinyurl.com/2amodvo
02:36:46 <pikhq> alise: What you're hearing there is just the normal Japanese syllable structure.
02:37:04 <alise> It's not very good
02:37:18 <pikhq> There *aren't* many sounds in Japanese.
02:37:31 <pikhq> And I think he got most of them in there.
02:37:37 <alise> <EOF> http://tinyurl.com/2amodvo
02:37:45 <alise> links to a let-me-google-that-for-you for "Child Pornography"
02:37:47 <Sgeo|web> pikhq, did you receive my msg's?
02:37:54 <alise> if anyone's too lazy to click. or is just not stupid enough to click links from EOF
02:38:21 <alise> pikhq: "Communication" You can't just fucking use an English word to rhyme
02:38:33 <alise> Nobody finishes English rhymes with a French word because they're too lazy :|
02:38:44 <alise> YOU USED "COMMUNICATION" AGAIN
02:38:51 <Sgeo|web> Maybe we could find a foreign word to rhyme with "orange"
02:39:02 <pikhq> alise: It's a perfectly ordinary and commonly used word in Japanese.
02:39:07 <alise> pikhq: Shut up. :|
02:39:48 <alise> Law of Fantasy: Any language you make up uses "ae" to excess.
02:39:54 <alise> "Ika Lgass Hunaethn" "ancient Aethn"
02:39:57 <alise> Sam Hughes, I am disappoint.
02:40:20 <pikhq> Most people suck at making plausible languages.
02:41:05 <pikhq> Even though for most uses of a made-up language (namely, handful of loan words) it would pretty much suffice to define a basic phonology and perhaps a handful of morphemes.
02:41:16 <Sgeo|web> Well, kind of difficult for the characters to ever learn the true pronunciation, I think
02:41:34 <alise> Sgeo|web: True pronunciation? What?
02:41:39 <pikhq> alise: Ignore him.
02:41:46 -!- yiyus_ has quit (Ping timeout: 265 seconds).
02:41:51 <alise> I think maybe we are getting too lax with our Sgeo-does-not-talk-to-me-about-Fine-Structure-at-all rule :P
02:42:01 <pikhq> alise: No, the thing is, he's making shit up.
02:42:13 <alise> Wild mass guess fanon?
02:42:21 <pikhq> No, literally making shit up.
02:42:46 <alise> Has he divulged some insane theory to you in /msg or something :P
02:43:32 <EOF> anyone want to do some marathon programming?
02:44:05 <alise> pikhq: Is it just me, or has Sam Hughes actually gone to lengths to make "The Voice Of The World" in "The Nature of the Weapon" actually seem like it was translated?
02:44:10 <alise> Like, not flowing quite in the usual manner.
02:44:18 <alise> It may just be my brain making shit up itself.
02:47:18 -!- SgeoN1 has joined.
02:48:46 <alise> I wonder how on earth all this is going to tie together.
02:49:09 <alise> If I was Hughes, I'd be looking for a cop-out ending that saves me from having to explain everything round about now.
02:50:12 <alise> Well, I haven't, so elaborate.
02:50:40 <EOF> a while back a friend of mine learned PHP, the next month he was a pothead....
02:51:46 -!- SgeoN1 has quit (Ping timeout: 245 seconds).
02:52:09 -!- SgeoN1 has joined.
02:55:23 -!- Wamanuz has quit (Remote host closed the connection).
02:57:57 <SgeoN1> Alise, tell me when you leave 1970-
02:58:13 -!- yiyus_ has joined.
02:59:09 <SgeoN1> Because I'd get bored wth being updated every 1970- story you finish
02:59:17 <alise> SgeoN1: And why do you need to know...?
02:59:33 <alise> I'm on "The Big Idea", anyway.
03:04:28 <alise> Note to self: Yuen is fucking crazy.
03:05:30 <alise> Whether by doing that thing to that group every time interval, or by saying she does.
03:06:02 <alise> The former, seemingly.
03:07:04 <SgeoN1> I'd ask whether a certain story is still listed, but it would reveal a connection to that stry
03:07:50 <alise> SgeoN1: You could check.
03:07:56 <alise> SgeoN1: Uh, in 1970-? I can just paste the list.
03:08:12 <alise> How has HE survived?...
03:08:50 <SgeoN1> You think I spoiled you when I didnt
03:09:42 <alise> You said no presumably to the story.
03:09:46 <alise> I thought it was to that M-guy.
03:09:50 <alise> I think it no longer.
03:10:02 -!- distant_figure has quit (Ping timeout: 264 seconds).
03:10:07 <alise> I am 99% certain it's that guy, anyway, since he can do that thing he does.
03:10:23 <alise> (Fine Structure is a story about saying the most with very, very little.)
03:10:56 -!- distant_figure has joined.
03:11:14 <SgeoN1> Ask pikhq what I meant, I'm not retyping it
03:11:39 <alise> No; that would be a spoiler.
03:11:46 <zzo38> We should make up the plane 17 and more to "hyperastral planes", where the high bits of the code numbers can indicate properties and right-to-left and complex scripts and other things like that.
03:11:51 <alise> I am now leaving "1970-".
03:12:42 <zzo38> (Where plane 0 is normal plane, 1 to 16 is supplementary planes, and 17 and above is hyperastral.)
03:15:31 <SgeoN1> And so I'm back, from outer space (no relevance )
03:15:41 <SgeoN1> It's playing on the radio
03:18:42 <SgeoN1> Now there's a song associated with one of the crappier virtual worlds
03:18:49 <SgeoN1> Or, that I associate with
03:19:12 <zzo38> Planes 16777216 up to 3355431 shall be hyperastral planes.
03:19:31 -!- distant_figure has quit (Ping timeout: 276 seconds).
03:20:18 <SgeoN1> Not as crappy as imvu though
03:20:32 <alise> What? Which? Not that I care, but.
03:20:43 -!- distant_figure has joined.
03:21:52 <SgeoN1> Note that my metric for crwppiness behaves weird sometimes
03:25:37 <alise> http://en.wikipedia.org/wiki/Nonviolent_video_games ;; Non-violent first-person shooter
03:26:24 <EOF> if(you != lying){printf ("you deserve a hug.\n");}
03:27:14 <alise> (if (abuses-fake-code-for-dialogue-in-that-way-once-more? EOF) (perish-burning-fire EOF))
03:27:30 <alise> At least make your fake code work properly in /theory/ or just don't do it at all or ... ever...
03:28:13 -!- distant_figure has quit (Ping timeout: 272 seconds).
03:29:13 <EOF> fine, do you want me ti pastebin the whole file?
03:29:38 <alise> EOF: There is no possible way "you != lying" makes coherent sense.
03:29:56 <alise> Sure, you can make that code compile and execute... and possibly even ask the user if they're lying, and make you = 1, and store it as a boolean in lying.
03:30:18 <alise> But it's still not what you were trying to express. And fake code is irritatingly pointless unless it has some code-related merit.
03:30:23 <alise> And yes I'm grumpy.
03:30:25 <SgeoN1> If you is an indication of truth or falsehood for some nutty reason
03:30:26 -!- GreaseMonkey has joined.
03:30:27 -!- GreaseMonkey has quit (Changing host).
03:30:27 -!- GreaseMonkey has joined.
03:30:35 <EOF> it's not elegant
03:30:51 <SgeoN1> I sort of see lying as a constant here
03:31:00 <alise> It's not RIGHT. You are-a person; lying is a status. A person cannot be a status belonging to them; it is meaningless. You have failed.
03:31:21 <alise> Therefore perishment.
03:31:44 * SgeoN1 attachés an iron chain with ball to EOF
03:31:44 <EOF> the program simply continues to execute
03:32:17 * SgeoN1 misread perishment as punishment
03:32:18 <EOF> life goes on, or ends...
03:32:29 <alise> SgeoN1: attachés? Really now?
03:32:46 <SgeoN1> I managed to typo an accent!
03:34:05 <EOF> what keyborad layout?
03:35:00 <EOF> it'd be easier to make this in java
03:35:08 <EOF> i was goinc for C
03:35:14 <Sgeo|web> Java poisoning is a terrible thing
03:35:35 -!- distant_figure has joined.
03:35:46 <EOF> java is (:-/
03:35:49 <Sgeo|web> Hell, even vaguely-C-like-family-language-poisoning is sad
03:36:14 <Sgeo|web> But sticking with that family of languages is poisonous
03:36:45 <EOF> C is all powerful
03:37:02 <Sgeo|web> Learn Haskell, or a Lisp, or Factor, or Smalltak (probably not Smalltalk, not different enough)
03:37:06 <EOF> omnipotent
03:37:19 * alise finds an error in Fine Structure
03:37:29 <alise> EOF: "uSUCK" -> obnoxious. Rabid C defending -> obnoxious.
03:37:32 <EOF> hence the esoteric channel
03:37:46 <Sgeo|web> It's a basic mental health thing
03:38:05 <Sgeo|web> Learning non-industry languages
03:38:16 <alise> Sgeo|web: "You're wife's maiden name" is the error.
03:38:23 <alise> I doubt EOF has anything to do with the industry.
03:38:45 <Sgeo|web> I should check that what I removed is actually the HD
03:38:48 * EOF removes SgeoN1's poorly attached ball and chain.
03:39:40 <alise> "sudo su" is extremely bad practice.
03:39:44 <alise> "sudo -s" is always preferable.
03:39:46 <EOF> echo "i am omnipotent"
03:39:53 <alise> Or "sudo -u user -s" for the case of switching to another user.
03:40:08 <EOF> killall rm
03:40:28 <EOF> kill -9 Sgeo|web
03:40:32 <alise> If Sgeo|web's command is executing, you are not facing a shell prompt.
03:40:45 <alise> That is "kill -9 Sgeo | web". That is unlikely to be what you want.
03:40:50 <alise> Furthermore, kill takes a process ID.
03:41:20 <EOF> what is your PID Sgeo|web ?
03:41:38 <alise> Does this inanity have a point? Humour can count as a point.
03:41:44 <EOF> you're shit, number 2
03:42:00 <alise> I mean, assuming that it will, eventually, evolve or progress into something that has a single element of humour -- that would be enough.
03:42:29 <EOF> that was mean
03:42:40 <alise> Killing people is generally considered mean.
03:42:40 <EOF> and i'm drunck
03:42:47 <EOF> and my friend is high
03:42:56 <EOF> and wer'e all idiots
03:43:01 <alise> What has your /friend's/ mental state got to do with anything?
03:43:21 -!- distant_figure has quit (Remote host closed the connection).
03:43:23 -!- wareya has quit (Read error: Connection reset by peer).
03:43:48 -!- distant_figure has joined.
03:44:25 <Sgeo|web> If my house were a restaurant, it would be condemned
03:44:48 * EOF gets Sgeo|web a maid
03:45:45 * EOF gets Sgeo|web laid
03:46:10 * EOF sets up Sgeo|web 's RAID
03:46:48 <Sgeo|web> RAIDs probably still die when dropped
03:47:36 * EOF has overused rhyme and for that he will have dearly paid
03:47:57 * EOF is sorry for what he said
03:48:39 <Sgeo|web> I speak for all sane persons when I said that this is proof that English is demented
03:49:17 <alise> Sgeo|web: only shitty laptops fail like yours when dropped
03:49:33 <alise> my Toshiba has a habit of parking the drive heads when I so much as tilt it, and it's very effective.
03:49:40 -!- wareya has joined.
03:49:45 -!- distant_figure has quit (Ping timeout: 272 seconds).
03:50:10 <EOF> my laptop is new and crappy
03:50:29 <EOF> because it's a gateway(ewww)
03:50:37 <EOF> they overheat :/
03:51:38 -!- EOF has quit (Quit: Page closed).
03:51:45 * pikhq sets out to replace ALSA
03:51:57 <alise> -- Are all the guys with that name in this story the same fucking person?!
03:52:20 <alise> pikhq: With OSSv4?
03:52:23 * Sgeo|web forgot the details of that story
03:53:37 <alise> pikhq: Recommendable. BTW: Install XFCE's mixer if you don't have it. It can do OSSv4.
03:54:02 <alise> PCM11 -- for me, at least -- and the output of the first vmixer -- are the ones you want. Use ossxmix to turn everything up to full volume, first; watch out for application-specific volumes.
03:54:05 <alise> Sometimes they start out low.
03:54:13 <alise> Basically you want everything but your last output turned up to maximum.
03:54:16 <alise> Which is not the default.
03:54:18 <alise> "osstest" is useful.
03:54:20 <alise> Sgeo|web: It sucks.
03:55:12 <pikhq> Aaaaand there's not useful Gentoo packages. Fek.
03:55:23 <alise> pikhq: Yes, there must be.
03:55:31 <alise> http://www.gentoo-wiki.info/OSS4
03:55:36 <alise> Add the OSSv4 layman overlay:
03:55:37 <alise> layman -a oss-overlay
03:55:54 <pikhq> alise: The tarballs referred to in those ebuilds no longer exist.
03:56:04 <alise> pikhq: Give up now while you still have your sanity.
03:56:13 <alise> Recommended suffix to that sentence: "On Gentoo, that is."
03:56:29 <pikhq> I have three options: write my own ebuilds, give up, or switch distros.
03:56:47 <pikhq> And I don't know of any distros that suck less ATM.
03:56:48 <alise> You use freaking Gentoo. The last option is such a wonderful one, I'm surprised you're not going to take it twice.
03:56:56 <alise> pikhq: Well, true.
03:57:07 <alise> I use Ubuntu. Obviously, you won't. I don't blame you.
03:57:24 <pikhq> Well. Nix is *highly* tempting.
03:57:25 <alise> pikhq: Arch Linux is decent enough and OSSv4 is a single installation away -- but its community is as bad as Gentoo's used to be.
03:57:45 <alise> pikhq: You could try GoboLinux. For the fucking hell of it.
03:57:46 <Sgeo|web> I don't think I can be trusted to judge communities
03:57:55 <pikhq> alise: Are you familiar with Nix?
03:58:06 <alise> It's totally done wrongly, of course.
03:58:08 <Sgeo|web> Based on one person, I thought the whole LambdaMOO community was anti-documentatipn
03:58:18 <pikhq> Yeah, but less so than a lot of other things.
03:58:30 <alise> Actually Gobo Linux is like Nix when you don't try and make everything as much like Haskell as possible.
03:58:30 <pikhq> GoboLinux might be fun to try as well.
03:58:58 <alise> pikhq: And if you use GoboLinux/Étoilé, well, then you have OS X with Torvalds' name stuck on it. :P
03:59:10 <alise> And a rather hopeless set of applications...
03:59:27 <pikhq> GoboLinux only supports i686. DEAL BREAKER
03:59:37 <alise> pikhq: Oh, I wish tuomov's blog is still up; he had a *wonderful* post on package manager design.
03:59:55 <pikhq> alise: Know the old URL? archive.org
04:00:00 <alise> pikhq: robots.txt.
04:00:06 <pikhq> FFFFFFFFFFFFFFFFFFFFUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU
04:00:19 <alise> I'd email him, but, well, his abrasive personality is pretty well-known. :)
04:00:24 <alise> pikhq: It may be his server administrators.
04:00:30 <alise> pikhq: Hey, I know1
04:00:38 <Sgeo|web> alise: did you get to the description of Akker?
04:00:48 <pikhq> alise: Proprietary ATI drivers available?
04:00:50 <alise> Sgeo|web: I know he committed suicide. That's all; shut up.
04:00:56 <alise> pikhq: Only one way to find out, buddy!
04:01:21 -!- distant_figure has joined.
04:01:34 <pikhq> That's a "no, but some crazy bastard hacked the proprietary drivers to kinda-work on FreeBSD".
04:01:50 <alise> But who needs graphics?
04:01:53 <pikhq> If it weren't for that, shit, I'd be on a BSD tonight.
04:01:54 <alise> pikhq: I see only one solution.
04:02:17 <pikhq> alise: I'm not making my own distro.
04:02:23 <alise> Apex. (Alise + Pikhq + Unix = Apix, "ooh that's similar" -> Apex).
04:02:28 <alise> No, WE are! KITTEN HUGS FOR EVERYONE
04:03:26 <pikhq> Oh, yeah. Another thing that makes me hate most distros: they almost all suck at building packages sanely.
04:03:51 <pikhq> Gentoo is somehow the best at doing this. Everything else will *gleefully* allow things like undefined symbols in shipped binaries.
04:04:43 <pikhq> Debian's better about this than most, simply because they only change APIs every couple years or so.
04:05:02 <alise> AyeAyeAyeAyeAyePecccks
04:05:07 <pikhq> And Gentoo's better about this than most, because they change APIs daily and as such they actually have to deal with API breakage correctly.
04:05:57 <pikhq> And Slackware's better about this than most, because Patrick Volkerding is an all-around good guy.
04:06:23 -!- distant_figure has quit (Remote host closed the connection).
04:06:50 -!- distant_figure has joined.
04:07:21 <alise> pikhq: You cannot hide from its necessity!
04:07:40 * pikhq shall install Nix in a VM to check it out
04:07:57 <alise> I even have the slogan
04:08:11 <alise> "The Apex of Linux Systems Design in A Downloadable TargzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzZzz"
04:08:24 <Sgeo|web> alise: are you at Sundown yet?
04:09:44 <alise> HOW THE FUCK DOES THAT LINK TO UNBELIEVABLE SCENES
04:09:48 <alise> Sgeo|web: I'm at it; I will not read it. I will leave.
04:09:57 <alise> pikhq: APEX. Is this having any effect on you?
04:10:25 <alise> Sgeo|web: I need to fucking sleep.
04:10:46 <alise> pikhq: Well, is i-- Apex
04:10:53 <zzo38> alise: Normal sleep is no good?
04:11:06 <alise> Normal sleep is excellent! I will now obtain it.
04:11:11 <alise> But only if pikhq Apex pikhq Apex
04:11:32 <alise> No? Alas. Goodnight.
04:11:37 -!- alise has quit (Quit: Leaving).
04:13:37 -!- distant_figure has quit (Ping timeout: 264 seconds).
04:15:05 -!- distant_figure has joined.
04:18:26 -!- sshc has quit (Read error: Connection reset by peer).
04:22:22 -!- augur has quit (Remote host closed the connection).
04:29:02 -!- jcp has joined.
04:38:46 <zzo38> I found this list of videogame cliches: http://qntm.org/cliches
04:39:03 <zzo38> But I do think some of my games (and other games from other sources) are break some or all.
04:39:10 <zzo38> Let's think about "Super ASCII MZX Town" series.
04:39:24 <zzo38> Player one is blue and player two is red: Not applicable.
04:39:47 <zzo38> If it has a weak spot, it'll flash yellow when you shoot it: This game breaks this cliche.
04:41:36 <zzo38> Any broken machine of any kind will work perfectly once you have found the correct spare part. Moreover there is exactly one spare part hidden in the level. It is the correct one: There is one situation like this, but there is many spare parts (you need all of them). And they are found in a different level than the thing you have to fix.
04:42:54 <zzo38> An experienced gamer will automatically destroy crates/pots/boxes/rats to look for ammunition/health/magic/gold inside: In this game, shooting crates/pots/boxes/rats/monsters/yourself/etc is likely to only waste your ammunition. So use it only when absolutely necessary!
04:43:54 <zzo38> automatically looking for the red key to the red door: Most keys in this game do unlock the same color door. But there is one door which is different.
04:43:56 -!- sshc has joined.
04:45:49 <zzo38> Climbing a ladder is more life-threatening than engaging in a gun battle: Not applicable, there are no ladders.
04:47:13 <zzo38> If it's covered in fur, its a good guy. If it's covered in spikes, it's a bad guy: This game breaks this cliche. You cannot tell by what they are covered by. You often cannot tell by even seemingly obvious things!
04:51:30 <zzo38> Make the gamer think: Yes, to win the Super ASCII MZX Town series games, you do actually have to think about it. If you do what might appear obvious to you, you will quickly get stuck.
04:53:46 <zzo38> You might find a clue that seems obvious, but the obvious interpretation is not correct! You have to think about it sideways, too.
04:57:24 -!- sftp has quit (Remote host closed the connection).
05:03:39 <GreaseMonkey> does that mean that you have a platformer part where gravity is sideways
05:04:29 <zzo38> GreaseMonkey: Not in the game I am discussing, but there are such things in other games I have invented in the past.
05:06:23 <zzo38> Good guys will only have first names, and bad guys will only have last names: My game does not follow this rule. (Also, I consider the terms "first name" and "last name" racist, but that's not the point.) However:
05:07:31 <zzo38> MEDIUM_SIZE_MONSTER (who happens to be a bad buy) asks for various things in order to take over the world (including a copy of the game on VHS). Eventually BIG_MONSTER (a good guy) reveal they know the other guy's real name, which happens to be Dave.
05:08:42 <zzo38> (Yes, I do mean a copy of the game that those characters (and your character) are in!)
05:11:32 <zzo38> I have found more lists of these cliches, and since this game series is not finished yet, I might decide to break some of these cliches in the future games in this series.
05:12:23 <zzo38> Every powerful character you attempt to seek aid from will first insist upon "testing your strength" in a battle to the death: No, but there is one character who fakes it!
05:13:22 <zzo38> No city will have more than two shops (Natural Monopoly Rule): In Part II, there is one shop that has many vendors. I do not know whether or not this counts.
05:15:02 <zzo38> When you're out wandering around the world, you must kill everything you meet: Not the case at all. In fact, attempting to kill everything you meet will just waste your ammunition, even if you are successful at first. (You willl also get less bonus points because you have less conserved ammunition)
05:20:01 <zzo38> Xenobiology Rule (predatory species of the world will include spiders, scorpions, snakes, wolf, dragon, etc): The game does not include all of them. Also, some individuals will be good to you instead, even of similar type than the other ones. Example: In one room there is two dragons. One will hurt you. The other is good and won't hurt you (they will ask for your autograph, though).
05:21:45 <zzo38> However, I might say one thing: The spiders/scorpions/dragons/etc that are good to you and don't hurt you can usually talk, in this game (but occasionally in a language which you cannot understand).
05:37:19 <zzo38> A note about the source-codes of this game: The codes are freely available for anyone to view, modify, and distribute. The codes do exactly what it seems, with no lies. However, reading them is more likely to deceive you than to help you to win this game (in most cases, not all)!
05:41:19 -!- augur has joined.
05:49:27 -!- bsmntbombdood has quit (Ping timeout: 272 seconds).
05:55:43 <coppro> throne room + thief is so good
05:55:48 <GreaseMonkey> why, because "jump off a cliff" actually means "jump, off hits a with a cliff"
05:56:13 <zzo38> Throne room + thief? Give an example, perhaps
05:58:23 <zzo38> An ending sequence that's little more than a single line and a pixelly picture: Part I, just says you win (no picture). Part II, does have a picture (that has nothing else to do with the rest of the game), and other text, such as "Those responsible for the subtitles have been sacked". The things having to do with the game are the "you win" and the bonus points.
06:00:42 -!- bsmntbombdood has joined.
06:20:42 -!- augur has quit (Remote host closed the connection).
06:21:35 <zzo38> How does that game work?
06:22:11 -!- SgeoN1 has quit (Quit: Bye).
06:26:21 <zzo38> Do you know The CGA Collection: The CGA Collection is a collection of games for DOS systems, all written in QBASIC, all using either 40x25 sixteen colors text mode or 320x200 four colors graphics mode. Sound effects are only PC speaker. Games include: BJACK, BUMPERSH, COLORSPI, ELEMENTA, DOWN, KNAR, HACKBITS, GIVEAWAY, MAKETEN, MINES, MUTCHNAM, PUZGEN, QCOOKIE, SKEDALS, SNAKEBIT, SOVMINGA, STARSTAK, STARWARS, and more.
06:27:18 <zzo38> The game MUTCHNAM has ten billion levels (really!).
06:28:17 <zzo38> The game KNAR has five thousand levels.
06:28:19 <myndzi> dominion, the board g ame?
06:28:53 <myndzi> imagine a CCG, but the game is to build your deck (instead of building it beforehand and the game is playing it)
06:29:22 <zzo38> myndzi: OK. Now I know a little bit.
06:29:46 <myndzi> have a little bit more ;)
06:29:52 <myndzi> i haven't played in a while so i forget the exact details
06:30:10 <myndzi> but basically you set up 10 stacks of cards in the center of the table, as well as stacks of property and money
06:30:16 <myndzi> (these vary in amount by the number of players)
06:30:17 <zzo38> myndzi: Have you ever invented any chess variants?
06:30:40 <myndzi> on your turn you draw cards, play them, and discard your whole hand
06:30:57 <myndzi> money cards let you buy things, property cards are useless but give your deck value, and other cards have CCG-ish things going on
06:31:13 <myndzi> that's about how it works then :P the 10 stacks are chosen from about 25 different cards in the basic set
06:31:17 <zzo38> Do you have to play all your cards, or some of them, or a certain number of them, and do you discard all of them regardless?
06:31:20 <myndzi> so you get a different game depending on what cards you play with
06:31:24 <myndzi> you discard them all regardless
06:31:30 <myndzi> so you go through your deck pretty quickly
06:31:34 <myndzi> and you can't "save up" money or something
06:31:42 <myndzi> it lends a focus towards building a deck that works well with itself
06:31:57 <myndzi> there are cards that can give you more actions, more money, more cards
06:32:05 <zzo38> Interesting idea.....
06:32:11 <myndzi> along with various other stuff, some cards that are sorta like "attacks"
06:32:19 <myndzi> though nothing with hit points
06:32:37 <myndzi> as for inventing a chess variant, ha, nah
06:33:20 <zzo38> I have invented a few chess variants, including Xorix Shogi, where the pieces powers are XORed by any piece it captures.
06:33:29 <zzo38> (Do you know about shogi? I have a set of shogi pieces and board)
06:33:30 <myndzi> though when i learned about navia dratp, i concocted a half-designed game that was influenced by that, settlers of catan, and uhh, whatever the hell that one OSS turn based strategy game is called
06:33:53 <zzo38> Navia Dratp? Yes I saw them playing it at the anime convention!
06:33:55 <myndzi> i never actually playtested it haha
06:34:08 <myndzi> but i did cut out a bunch of little hexes and sorta design a way to randomize the board!
06:34:16 <myndzi> but it wasn't very effective so i didn't pursue it
06:34:34 <myndzi> also lol @ xored powers
06:34:38 <myndzi> that's pretty interesting actually
06:34:39 -!- augur has joined.
06:34:57 <myndzi> did you revise the movements or just keep them the same?
06:35:58 <zzo38> myndzi: They are mostly kept the same, but see: http://www.chessvariants.org/index/msdisplay.php?itemid=MSxorixshogi
06:38:20 <myndzi> it would be interesting to see what could become of designing the pieces' basic movements around the concept
06:39:04 <zzo38> myndzi: Yes it might be, perhaps you can try to make a new game based on a similar concept, if you want to.
06:39:15 <myndzi> what happens if a lance captures a pawn?
06:39:33 <myndzi> can it then skip a square? :P
06:39:43 <myndzi> (and also be unable to capture in that square)
06:39:43 <zzo38> Anything capturing a pawn in Xorix Shogi stay as they are.
06:39:59 <myndzi> fine, rook capturing a gold general
06:40:07 <zzo38> It says "All pieces can always move one space straight forward".
06:40:20 <myndzi> i was trying to concoct an example ;p
06:40:30 <myndzi> of how to xor "move one space" with "move any number of spaces"
06:40:46 <zzo38> See this page for some clarification: http://www.chessvariants.org/graphics.dir/xorixshogi/index.html
06:40:47 <myndzi> i read that rule but didn't put 2 and 2 together
06:42:20 <zzo38> Can you play mahjong? I played Washizu Mahjong at the anime convention (just once, though).
06:43:04 <myndzi> did you get your blood drawn? :P
06:43:58 <zzo38> myndzi: No, we didn't do that part. We didn't play for money either.
06:44:28 <zzo38> But we did use all the other rules such as the transparent tiles (some are opaque), and the teams.
06:45:10 <myndzi> haven't played mahjong yet, but i did watch akagi after someone told me about how the game worked
06:45:17 <myndzi> the transparent tiles thing was actually kinda interesting
06:45:24 <myndzi> i am having trouble understanding the graphics on the page you linked
06:45:36 <quintopia> without having to read up, what's going on here?
06:45:38 <myndzi> presumably a 'leap' is a one space movement?
06:45:44 <myndzi> and a ride is 2 or more
06:46:00 <myndzi> but the images for 0 and 1 are confusing
06:46:09 <myndzi> quintopia: just talking about board games and variants
06:46:29 <myndzi> zzo38: nevermind, it just occurred to me
06:46:34 <zzo38> Yes, leaps mean one space
06:46:37 <zzo38> And ride is two or more
06:46:45 <myndzi> it's a bit counterintuitive to have the icons not drawn in the same place every time
06:47:20 <zzo38> I did not draw the icons. I wrote the rules to the game, and then Fergus Duniho made the icons for the pieces.
06:47:47 <myndzi> it's still a bit counterintuitive seeming at first :P
06:48:41 <myndzi> i thought i had a clever way to implement that game with physical pieces
06:49:00 <myndzi> but it wouldn't let you keep the powers of captured pieces
06:49:52 <zzo38> I also invented a chess variant which is extremely unlike chess, it uses a one-dimensional board with 72 cells, unequal armies, 12 kinds of pieces, and no rule of "check"; yet, it is exactly the same as chess.
06:52:11 <zzo38> One of the 12 kind of pieces is a neutral kind that belongs to neither player. There are a few kinds of pieces that cannot move, and pieces which change the first time they move, pieces which cause other pieces to change.....
06:52:41 <zzo38> It might seem to be entirely different from chess, but, actually, it is exactly the same as chess!!
06:53:10 <zzo38> Here are the rules in case you want to read it: http://www.chessvariants.org/index/msdisplay.php?itemid=MSeeeeeeeeeeeeee
06:58:19 <myndzi> ha, that's sorta cheating
06:58:30 <myndzi> it's not REALLY one dimensional ;p
07:00:20 <zzo38> Higher dimensional games can be mapped onto one dimensional boards, and then indicate everything in one dimensional terms.
07:03:22 * quintopia challenges myndzi to an m,n,o,p,q,r,k game
07:03:23 * myndzi challenges quintopia to an m,n,o,p,q,r,k game
07:03:41 <quintopia> damn you and your partially scriptedness!
07:04:02 <myndzi> i don't know what this m,n,o,p,q,r,k is
07:04:24 <myndzi> also: i try so hard to tempt people to abuse that script by making it obvious but to no avail!
07:05:50 <zzo38> The game GIVEAWAY in The CGA Collection is also a chess variant. There are 26 kind of pieces. It is a single-player game, but there are opponent pieces as well, but the opponent pieces have a fixed rule for what they have to do (sort of like the rule in blackjack that the dealer has a fixed rule to play).
07:06:45 <zzo38> If you can capture, you must capture, otherwise you can make a non-capturing move. Opponent's pieces on opponent's turn make all possible captures simultaneously. If you have no legal move or if opponent's pieces are unable to make a capture, then you lose. If you run out of your own pieces, then you win.
07:07:27 <quintopia> that sounds like a really weird game
07:07:55 <zzo38> Rules for capturing are just like in normal chess.
07:08:16 -!- Ilari has quit (Quit: leaving).
07:08:17 -!- Ilari_antrcomp has changed nick to Ilari.
07:08:48 <zzo38> There is a level file containing the boards (which can be of any size), and it is also possible to create your own levels, that you might even ask someone else to solve.
07:09:33 <quintopia> zzo38: what are the rules for the other side?
07:10:40 <zzo38> quintopia: The rules for opponent's pieces are that on opponent's turn, opponent's pieces make all possible captures simultaneously. If a piece can capture multiple other pieces, copies are created in all of those positions. If multiple opponent pieces of different kind try to move to the same position, it turns into a question mark, which cannot move or capture at all.
07:12:25 <quintopia> zzo38: so the opponent pieces do not move if they cannot caputre, or if it is not clear which piece should do the capturing?
07:13:18 <zzo38> quintopia: No, if multiple opponent's pieces can do the capturing, they *all* do, simultaneously, and if they are not all of the same kind, they turn into a question mark.
07:13:33 -!- Ilari has changed nick to Ilari_antrcomp.
07:13:39 <zzo38> For your own pieces, you can make only one move (or capture) each turn, no simultaneous moves allowed.
07:14:02 -!- Ilari has joined.
07:14:11 <zzo38> If there are no possible ways that the opponent's pieces can capture any of your own pieces, then you lose the game.
07:14:33 <quintopia> zzo38: what does a question mark do when it forms?
07:15:19 <zzo38> quintopia: It can be captured. It is just a piece with no moves.
07:16:55 <Ilari> quintopia: Color Graphics Adapter?
07:17:33 <zzo38> quintopia: Yes, it is Color Graphics Adapter.
07:18:50 <zzo38> quintopia: Why what? Why does CGA stand for that?
07:19:03 * pikhq tried to watch Repo! The Genetic Opera today. The music was terrible.
07:19:30 <quintopia> zzo38: no i mean, why does somethign named that have a collection of games?
07:19:39 <pikhq> Bad operatic music is cringe-inducing. Absolutely, positively cringe-inducing.
07:20:56 <zzo38> quintopia: Rather, the collection of the games has the title "The CGA Collection", and they all have certain things in common, one of them being that all of them use either 40x25 text mode with sixteen colors, or 320x200 graphics mode with four colors.
07:21:58 <quintopia> zzo38: is CGA a company? did they develop these games?
07:22:28 <zzo38> quintopia: No. CGA is just the name of the graphics adapter. I programmed all of these games myself.
07:24:03 <zzo38> I did not actually *invent* all of the games myself, though, but I did invent some, and even some of the ones that I did not invent, these implementations have some extra features or something else.
07:24:25 <zzo38> All the sound/music in the SKEDALS game is Bohlen-Pierce.
07:26:32 <zzo38> And I did in fact invent a game with ten billion levels.
07:26:59 <quintopia> so these games only work if you have said graphics adapter?
07:27:51 <zzo38> quintopia: No, modern graphics adapters can usually emulate the CGA. And even if they don't, you can use DOSBOX or something like that.
07:28:16 <zzo38> (CGA is a very old graphics adapter.)
07:28:48 <quintopia> so one needs dosbox or some other kind of dos emulator then?
07:28:50 -!- lament has joined.
07:29:35 <zzo38> Yes, to run on Linux you do need a DOS emulator.
07:38:43 <zzo38> The COLORSPI game is a sort of puzzle game and action game. Your piece is the spider. There are green backgrounds and black backgrounds, and then there are walls. There are also seven colors of balls and seven colors of webs. And there is a time limit!
07:39:29 <zzo38> If a ball touches a spider, touches a web of a different color than the ball, or touches another ball, then you lose.
07:40:02 <zzo38> Webs of different colors are not allowed to cross each other.
07:43:20 <zzo38> You have limited number of webs of each color, at each level. Some colors of webs can be cut and some cannot be cut.
07:44:09 <zzo38> If a ball touches a web of the same color of the ball, the ball remains there (you can still lose if other balls collide with it), and it is safe for a spider to touch it and then it is removed. Once all balls are removed in this way, the goal apperas.
07:44:24 -!- yiyus_ has quit (Read error: Connection timed out).
07:44:49 <zzo38> You have to reach the goal also within the time limit, and sometimes you might have to make sure to plan ahead correctly such that you *can* reach the goal! (Some levels have no balls, in which case the goal is visible from the start.)
07:45:30 <zzo38> Webs can be created on black backgrounds only (and it is the only way for a spider to move across black backgrounds). Green backgrounds can contain no webs but can be moved across freely.
07:45:51 <zzo38> There are also a few other kind of objects, some of which can be pushed, and which work only on green backgrounds.
07:46:49 <zzo38> Do you like this idea?
07:48:41 <zzo38> quintopia: Just guess, perhaps? Or make question/comment?
07:49:19 <GreaseMonkey> <zzo38> quintopia: Yes, it is Color Graphics Adapter. <-- s/Adapter/Array/
07:49:42 <zzo38> GreaseMonkey: No, it is "Adapter". VGA is Video Graphics Array.
07:51:03 <zzo38> Boxes and slanted lines can be pushed around on green areas (they cannot be pushed into black space, it is like there is a wall there). Balls touching boxes cause you to lose. Balls touching slanted lines are deflected.
07:51:33 <zzo38> And then, there is one kind of piece which can be pushed like the box, except that it can be pushed into a black space to make it green.
07:52:04 <zzo38> And there is a sign that changes into a wall after anything passes over it.
07:53:46 <zzo38> This is how the game works!
07:55:04 <quintopia> if i ever get a chance to try the game out, i'll let you know if it's any good
07:58:42 -!- zzo38 has quit (Quit: Good day. This was *not* the Spanish Inquisition speaking, nor was it your boss's sister, despite any rumors you may have heard (or not heard).).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:03:08 -!- yiyus_ has joined.
08:15:07 -!- coppro has quit (Ping timeout: 276 seconds).
08:15:12 -!- coppro has joined.
08:15:12 -!- coppro has quit (Changing host).
08:15:13 -!- coppro has joined.
08:25:41 -!- olsner has joined.
08:29:45 <quintopia> meh. you're just a userscript engine for firefox. i'm looking for people.
08:33:33 <quintopia> 10:30 is 7 hours ahead of me, which doesn't seem quite far enough to be new zealand
08:33:43 <quintopia> in fact, i though india was 7.5 hours ahead
08:34:28 <quintopia> bah, international channels should all have timezone-bots.
08:37:52 <olsner> late? it's early, I just got up
08:39:19 <fizzie> Yes, but I thought we were using those 2-char country codes.
08:40:10 <fizzie> http://zem.fi/~fis/esomapnz.png -- now with more zzo38, whose weirdness unfortunately makes other people get more overlappy than in esomapn.png.
08:41:57 <quintopia> is this a map of when peoples are in this channel?
08:43:57 <fizzie> No, it's a map of two first PCA components out of 60 writing-style related features.
08:45:08 <fizzie> http://zem.fi/~fis/esomap-comp.png -- that has some of the more interesting features.
08:45:41 <fizzie> I have to leave in ten minutes and get dressed before that, so don't have time for replotting right now, otherwise yes.
08:46:11 <quintopia> is capitalization at beginning of sentences a feature?
08:46:18 <fizzie> http://zem.fi/~fis/test12.png -- this one is channel activity per time-of-day (Finnish time).
08:46:49 <quintopia> looks about like every channel i've been in
08:47:07 <fizzie> It might be, I'm not sure. The feature set was originally designed for book author classification though, and everything's capitalized properly there.
08:47:29 <fizzie> http://zem.fi/~fis/test12r.png -- same thing normalized.
08:47:36 <fizzie> Now I really must be going.
08:48:20 <quintopia> alise posts a third of the messages pretty much around the clock
08:54:42 <fizzie> Yeah, like the saying goes, "evil never sleeps".
08:55:38 <quintopia> hey fizzie are you familiar at all with entscheidungsproblem?
09:00:18 -!- Quadrescence has quit (Read error: Connection reset by peer).
09:00:47 -!- Quadrescence has joined.
09:25:26 -!- Quadrescence has quit (Quit: omghaahhahaohwow).
09:29:49 <quintopia> man, people that take their foul moods out on other people suck
10:02:49 -!- FireFly has joined.
10:09:40 -!- cheater00 has quit (Ping timeout: 276 seconds).
10:16:17 -!- Wamanuz has joined.
10:22:13 -!- lament has quit (Read error: Operation timed out).
10:34:22 -!- tombom has joined.
10:54:31 -!- nooga has quit (Ping timeout: 276 seconds).
10:54:36 -!- nooga has joined.
11:16:51 -!- cheater99 has joined.
11:18:29 <Vorpal> alise, fizzie, oerjan: http://sprunge.us/CdUU
11:19:37 <quintopia> what's the purpose of doing that anyway?
11:19:48 <quintopia> does he want his lines in the logs to be easier to find?
11:20:27 <Vorpal> quintopia, no, that is him saying stuff like "oerjan swats alise -----###" (he uses his flyswatter)
11:20:35 <Vorpal> I guess it is some kind of in-joke .P
11:20:59 <quintopia> i would never have recognized that as a flyswatter
11:25:00 <Vorpal> quintopia, what about ===\__/ then?
11:27:59 <Vorpal> quintopia, again an in-joke :P
11:29:09 <quintopia> let's see who laughs when i hit HIM with a pan.
11:29:42 <Vorpal> quintopia, ah but then you would have to steal it or get him to lend it to you (since there is only one pan)
11:30:41 <quintopia> Vorpal: then what was that you just posted above? did you, *gasp* MAKE YOUR OWN PAN?
11:31:01 <Vorpal> quintopia, no it was not in a /me, so it didn't count
11:31:13 <Vorpal> thus I was quoting him
11:31:26 <Vorpal> so two reasons it didn't count
11:31:55 -!- sftp has joined.
11:34:07 <Vorpal> quintopia, he is not in the channel atm, how would that be possible?
11:35:10 <quintopia> if he's not here, then he's not paying attention to his virtual pan
11:35:17 <quintopia> also, you don't have to highlight me every few lines
11:36:28 <Vorpal> ah indeed, a bad habit
11:36:36 <Vorpal> I'm used to high traffic channels
11:54:22 <Vorpal> oerjan is from Norway. I'm from Sweden
11:54:36 <Vorpal> and there are quite a few from .fi (plus two more from .se)
11:56:08 <quintopia> hard to keep you norsemen straight
11:56:49 -!- GreaseMonkey has quit (Remote host closed the connection).
12:00:34 <FireFly> Quite a few more than two more swedes, I think
12:33:15 <Vorpal> FireFly, I only know of you and olsner apart from me?
12:33:47 <FireFly> BeholdMyGlory is here from time to time, also MigoMipo
12:33:52 <FireFly> They're not here atm though
12:52:17 <Vorpal> sqlite> select * from logs where serial > 869850 and serial < 869870;
12:52:17 <Vorpal> 869851|2009-10-25 01:02:32|FireFly|||5|"Later"
12:52:17 <Vorpal> 869852|2009-10-25 00:04:14|oerjan|||0|eek backwards time travel!
12:52:48 <olsner> hmm, time to go vote then
12:53:50 <Vorpal> olsner, Jag förtidsröstade
12:53:58 <Vorpal> FireFly, I hope you vote too.
12:55:07 <olsner> after voting, back to scheme in haskell in 48h
13:07:13 -!- sshc has quit (Ping timeout: 255 seconds).
13:40:15 -!- distant_figure has quit (Quit: Reconnecting).
13:40:21 -!- distant_figure has joined.
13:40:21 -!- distant_figure has quit (Client Quit).
13:43:27 <fizzie> http://zem.fi/~fis/esomapf.png -- okay, that's a lot less pretty sight. (Here each x represents 2000 consecutive messages, and it's again PCA dims 1/2 out of the same features as before.)
13:44:37 -!- alise has joined.
13:45:13 <Vorpal> fizzie, does python have any built in class or such for a circular buffer?
13:45:45 <Vorpal> I need to keep some sort of sliding window when matching up own/clog
13:45:54 <Vorpal> to handle out order issues
13:46:17 <alise> You're not saying "yarr" enough.
13:47:21 <Vorpal> alise, oh is it that day today?
13:47:24 <fizzie> Vorpal: I don't think there is a built-in type for that, but there's the "buffer" object that can represent a slice out of an array.
13:47:57 <Vorpal> fizzie, not sure I want to put the entire db into an array :P
13:49:13 <alise> Vorpal: Indeed; also, you are marked as away.
13:50:03 <fizzie> There's the queue class, but it's more about being thread-safely synchronized. (You could use two queues and pop off the end all the messages that are older than your threshold.
13:50:33 <fizzie> Oh, right, there's a deque in "collections".
13:51:22 <fizzie> You can even specify a maximum length to the deque and it will automatically drop items from one end when you put new stuff into the other.
13:53:42 <Vorpal> fizzie, yes that seems usable
13:58:02 <Vorpal> fizzie, I wrote up some code to create a number of known matching points btw
13:58:55 <Vorpal> fizzie, I'm sure there is a better way than http://sprunge.us/BUEj but it seems to work
13:59:34 -!- nooga has quit (Ping timeout: 276 seconds).
13:59:34 -!- Sgeo|web has quit (Ping timeout: 252 seconds).
14:00:16 -!- nooga has joined.
14:02:58 <fizzie> Sounds like someone (accidentally or not) repeating his message within a short period of time would be a bit problematic for completely unconstrained matching; I guess taking the earliest match always and then removing it from consideration for later ones would handle that, though.
14:04:06 <Vorpal> fizzie, yes indeed, but that code was just to find known good pairs, and I first got dups but then I got like every ":D" matched up to every other :D within 1 minute
14:04:21 <Vorpal> fizzie, the length(body) > 10 seems to have fixed
14:04:30 <Vorpal> and there were no other dup pairs there
14:04:54 <fizzie> I guess it is not so often someone repeats a very long message completely exactly.
14:04:57 <fizzie> I guess it is not so often someone repeats a very long message completely exactly.
14:05:20 <Vorpal> fizzie, and indeed it isn't and it would as far as I can tell show up twice there
14:06:45 <alise> Vorpal: your /msg notifications suck :)
14:07:01 <Vorpal> fizzie, does python have a two-way dict?
14:07:12 <Vorpal> fizzie, like equally fast to look up by key and value
14:07:12 <alise> Vorpal: does a buffer just appear silently or something
14:07:19 <alise> <Vorpal> fizzie, does python have a two-way dict? ;; no but trivial to code one
14:07:43 <Vorpal> alise, yeah, but I hoped to avoid that work, and also if it had it, doing the two-dict solution would be a bad case of NIH
14:07:53 <alise> as far as i know, there is no package for it.
14:07:55 <alise> writing code is not NIH.
14:09:03 <alise> class BiDict(object):
14:09:03 <alise> def __init__(self, to, fro):
14:10:22 <fizzie> I assume you might want to have something where a single call would add in both dicts.
14:11:42 <Vorpal> fizzie, I checked the values with a simple select serial,tstamp,body from irc.logs where body in (select body from irc.logs where serial in ( < long list here > )) order by body; and the only "dupe" was "my brain hurts" from 2007 and 2009, thus not within the critical interval :P
14:12:37 <Vorpal> and that was said by different persons
14:12:43 -!- sshc has joined.
14:13:24 <alise> fizzie: Vorpal: what would it take to convince you two to give me your full #esoteric logs sometime?
14:13:30 <alise> see, obviously botte has to have the most complete logs ever
14:13:37 <alise> so i need to merge all of them. clearly.
14:14:15 <Vorpal> alise, a lot, since there could be private info in them, due to being xchat logs for several years
14:14:23 <Vorpal> so stuff like /msg nickserv went to log of current window
14:14:40 <alise> Vorpal: Obviously, I would want those filtered out. :P
14:15:04 <Vorpal> alise, thus: never going to happen. And due to the ambiguity discussed before with * I do not trust filtering. Like "* You have been invited to #secret-channel"
14:15:27 <Vorpal> alise, I could perhaps filter only join/part/quit/msg, and leave out all CTCP ACTION or such
14:15:35 <Vorpal> that should be reasonably safe
14:15:38 <alise> I don't mind all /mes of that form being filtered out. Also, knowing the name of a secret channel is hardling devastating.
14:15:55 <fizzie> Oh, I'm very cheap, I think I could part with my logs with just a five- or six-digit sum of euros.
14:15:59 <alise> Vorpal: Hmm. Well, if I can get logs of fizzie I suppose I could use his to merge in /mes.
14:16:12 <alise> fizzie: Don't you see it's for science?
14:16:25 <Vorpal> alise, clog is reasonably complete
14:16:36 <alise> Vorpal: clog has quite regular outages.
14:16:44 <Vorpal> also it was fun writing parsing for znc buffer playback
14:16:44 <alise> there are huge swathes of logs missing in which good discussion has taken place
14:16:57 <alise> furthermore, its timechanges are problematic
14:16:57 <Vorpal> however due to the +- thing at start of messages it is sometimes ambig
14:17:12 <Vorpal> like when my irc client didn't yet know it was in that prefix-mode
14:17:21 <Vorpal> or when we were not yet in it
14:17:21 <fizzie> I can dump something out from the postgres db, after some sanity checks; I think they should be mostly safe.
14:17:29 <Vorpal> for example I know a few will be incorrectly parsed
14:17:59 <Vorpal> like <foo> [10:11:12] +--[>+ ...
14:18:03 <Vorpal> I saw a few such cases
14:18:24 <alise> fizzie: The postgres DB is clog logs, is it not?
14:18:25 <Vorpal> where it is ambig if it is + as in "id with services" or not
14:18:44 <Vorpal> alise, no it's his personal logs
14:19:19 <alise> fizzie: I'm not sure how dumping a Postgres DB is any less safe than simply running the logs through the filter that got them in there in the first place?
14:19:31 <Vorpal> alise, anyway logs from late yesterday and onwards should be reasonably simple to filter
14:20:01 <alise> <zzo38> [...] (Also, I consider the terms "first name" and "last name" racist, but that's not the point.) [...]
14:22:52 <alise> Presumably because in a lot of, for example, Asian cultures, the "last" name is listed first.
14:23:22 <alise> "Kim Jong-Il" is actually "Jong-Il Kim" in our ordering.
14:24:12 <Vorpal> class deque(__builtin__.object)
14:24:12 <Vorpal> | deque(iterable[, maxlen]) --> deque object
14:24:24 <Vorpal> and nowhere can I find docs what iterable does exactly
14:24:47 <Vorpal> alise, not in help(collections.deque) at least?
14:24:59 <alise> http://www.google.co.uk/search?sourceid=chrome&ie=UTF-8&q=iterable+python
14:25:01 <alise> 5. Built-in Types — Python v2.7 documentation
14:25:02 <alise> Python defines several iterator objects to support iteration over general and specific sequence types, dictionaries, and other more specialized forms. ...
14:25:03 <alise> PEP 234 -- Iterators - Python
14:25:03 <alise> 30 Jan 2001 ... (Due to a misunderstanding in the original text of this PEP, in Python 2.2, all iterator types implemented a next() method that was ...
14:26:13 <fizzie> alise: Because the filter can only output to postgres at the moment, so it's easier to dump. No safer, of course.
14:26:20 <fizzie> Anyway, busy with other stuff now.
14:27:08 <Vorpal> class collections.deque([iterable[, maxlen]])¶
14:27:09 <Vorpal> Returns a new deque object initialized left-to-right (using append()) with data from iterable. If iterable is not specified, the new deque is empty.
14:27:19 <Vorpal> alise, a lot more helpful than the built in help
14:27:48 <alise> Vorpal: the built-in help is not documentation
14:27:51 <Vorpal> alise, anyway, I presume deque(None,10) should work... (I need the second parameter, but not the first)
14:27:52 <alise> it is a quick reference tool
14:28:01 <alise> the actual documentation, i.e. the manual, is the only official reference.
14:28:09 <alise> Vorpal: deque(maxlen=10)
14:29:02 <alise> 00:40:10 <fizzie> http://zem.fi/~fis/esomapnz.png -- now with more zzo38, whose weirdness unfortunately makes other people get more overlappy than in esomapn.png.
14:29:12 <alise> * Everyone backs away slowly from zzo38
14:29:32 <alise> I have collided with the GregorR nebula.
14:29:43 <alise> It looks like everyone's being time dilated! NO! THE TACHYONS!
14:30:21 <alise> "The hull is collapsing!" "Send out a message on all channels! Even that secret channel we never use and that serviced as a deus ex machina three episodes ago! ZZO38--IS--A--BLACK--HOOOOOOOOOOOOOOOOOOO[static]"
14:30:45 <alise> Thank you, thank you, no need to crowd, you can stop clapping now.
14:31:48 <alise> 00:47:51 <quintopia> agagaga
14:31:49 <alise> 00:48:20 <quintopia> alise posts a third of the messages pretty much around the clock
14:31:49 <alise> 00:48:26 <quintopia> just as i would have guessed
14:31:55 <alise> And almost half the messages in total... mwahahaha.
14:33:14 <Vorpal> fizzie, you are about as far out as zzo (in a different direction though)
14:34:02 <alise> Less tight though.
14:34:11 <alise> Vorpal: No, not true.
14:34:14 <alise> fizzie is quite close to cpressey.
14:34:19 <alise> zzo38 is much further away from everyone else.
14:34:33 <alise> I think fizzie's position may be largely due to zzo38 pushing him away.
14:34:49 <Vorpal> can I remove arbitrary values in the middle of a dequeue
14:35:15 <alise> So: no. Not unless Python's doing something screwy.
14:35:18 <alise> del q[n] might work.
14:35:22 <alise> But I doubt it has q[n].
14:35:32 <alise> Actually it wouldn't be del q[n].
14:35:36 <alise> More like q.remove_at(n) or something.
14:35:53 <Vorpal> I have an algorithm pretty much worked out but the sliding window data data type needs to act as a queue but with easy removing in the middle (when we match up an out of order element)
14:36:20 -!- FireFly has quit (Quit: swatted to death).
14:36:28 <Vorpal> | D.remove(value) -- remove first occurrence of value.
14:37:10 <alise> 04:53:50 <Vorpal> olsner, Jag förtidsröstade
14:37:16 <alise> why do you randomly talk in swedish to the swedes
14:37:20 <alise> (also, before you say anything, blame clog)
14:37:52 <Vorpal> alise, because I don't have a clue what it is called in English
14:37:58 <Vorpal> pre-vote sounds so strange
14:38:12 <alise> "I voted early", just like Google translates it.
14:38:45 <olsner> well, it's a completely different procedure though
14:39:21 <olsner> it's not just voting earlier, it's making your vote before the actual voting starts in a special way
14:39:41 <Vorpal> alise, so the translation fails to convey the correct info
14:40:04 <alise> It's still called that.
14:40:16 <alise> "I early voted", if you're willing to trade flow with a bit more meaning.
14:40:21 <alise> early-voted, maybe.
14:40:30 <alise> http://en.wikipedia.org/wiki/Early_voting
14:40:32 <alise> We just call it voting early, though.
14:40:39 <alise> You silly overly-precise Swedes.
14:41:12 <Vorpal> actually hm ordereddict with size limit might work better
14:41:16 <Vorpal> could emulate that manually
14:52:06 <alise> Vorpal: make a fucking class.
14:57:51 <Vorpal> alise, yeah been doing that for several minutes
14:58:22 <Vorpal> alise, what is the python format spec thingy for tuples? Like %s or %d for string/integer
14:58:34 <alise> You should just use %s.
14:58:45 <alise> Vorpal: even for %d, %s is usually used
14:58:49 <alise> Or %r, I guess, for __repr__.
14:58:52 <alise> But for tuples that's the same.
14:59:20 <Vorpal> "TypeError: not all arguments converted during string formatting"? huh
14:59:39 <alise> i think i know what's wrong
14:59:41 <alise> but i need to see the line
14:59:43 <Vorpal> def dump_lines(ol, cl):
14:59:47 <Deewiant> If you have only one tuple it'll try to use its contents
14:59:57 <alise> Vorpal: for debug just use + :P
15:00:02 <alise> print 'Own: ' + ol
15:00:05 <alise> print 'Clog: ' + cl
15:00:33 <Vorpal> alise, well ol is a tuple containing a datetime object amongst other things
15:00:38 <Vorpal> would + magically handle that?
15:00:50 <alise> it would do str(ol) which == repr(ol)
15:00:59 <Vorpal> ugh %s with datetime formats non-nicely, oh well
15:01:12 <alise> it formats non-nicely with everything
15:02:33 <alise> everything formats non-nicely with repr()
15:02:50 <alise> what you need is an object inspector!
15:03:02 <alise> [Suddenly, your monitor goes blank. BONG! "Welcome to aliseOS."]
15:06:07 <Vorpal> alise, you mean like in genera?
15:06:23 <alise> Or in aliseOS, which has something like that but SO MUCH BETTER.
15:07:20 <Vorpal> hm it got the first 313 events by the fast simplistic algorithm (which just matches up line pairs from the db), now to enable the more complex sliding window one
15:19:07 -!- Sgeo|web has joined.
15:19:27 <Sgeo|web> I really don't want to eat breakfast in this house
15:24:49 <Sgeo|web> The flies (or maybe that's the wrong term)...
15:28:40 -!- ais523 has joined.
15:33:29 <ais523> update on my compression algo: it's better than bzip2 but almost certainly worse than lzma, except on very repetitive files
15:33:35 <ais523> and thus it doesn't have much of a reason to exist
15:37:26 <Sgeo|web> That's a good reason for your algo to exist
15:38:24 -!- zzo38 has joined.
15:38:29 -!- EOF has joined.
15:39:05 <Sgeo|web> I feel like I asked a stupid question
15:40:58 <zzo38> I do not completely understand what the http://zem.fi/~fis/esomapnz.pnghttp://zem.fi/~fis/esomapnz.png is plotting, it says "map of two first PCA components out of 60 writing-style related features" but I do not know what that means.
15:41:54 <nooga> alise: start writing aliseOS plz
15:42:21 <alise> nooga: Can I have a research grant?
15:42:51 <alise> Sgeo|web: LZMA is not patented.
15:42:55 <alise> ais523: you could TWEAK IT!!!11
15:43:11 <alise> ais523: also, it may be significantly faster than LZMA
15:43:15 <alise> which would be a very good idea
15:43:29 <ais523> and most of the tweaks slowed it down, but it doesn't lose too much to remove most of them
15:43:40 <alise> http://snarxiv.org/vs-arxiv/ ;; Which paper title is real, which is rubbish?
15:43:50 <alise> The real papers are just so ridiculous.
15:43:56 <alise> (http://snarxiv.org/ random stream of rubbish)
15:44:16 <ais523> without the weird tweaking, it's O(n log n) with relatively low constant factors
15:44:28 <alise> ais523: that sounds pretty good
15:44:34 <alise> what's the real-time speeds you're getting?
15:44:35 <ais523> also, it heapsorts linked lists. twice
15:44:35 <zzo38> Like, in that graph, what is the X axis mean, what does the Y axis mean, and what does everything else mean?
15:44:43 <alise> ais523: er -- not mergesort?
15:44:45 <ais523> alise: I'm not sure, because it's overloaded with slow tweaks atm
15:44:57 <alise> ais523: plug in a mergesort :P
15:44:58 <ais523> alise: well, I needed to implement a priority queue
15:45:08 <alise> low-hanging fruit etc.
15:45:15 <ais523> and because I had the priority queue code already, it was trivial to make it into a heapsort
15:45:33 <ais523> and heapsort and mergesort are equally fast in terms of computational order, heapsort often has better performance in practice
15:46:31 <alise> ais523: on linked lists?
15:46:36 <alise> that may be true for arrays
15:46:52 <ais523> I'm not sure about on linked lists
15:47:07 <alise> it's not like mergesort isn't a trivial algorithm
15:47:24 <ais523> heapsort isn't that bad either
15:47:42 <alise> i'm trying to help you find a niche for your admittedly-now-useless algorithm, stop complaining
15:47:45 <ais523> one advantage of the heapsort is that the natural expression of it is iterative rather than recursive
15:48:06 <ais523> oh, decompression's O(n), which is nice
15:48:14 <ais523> but I think lzma decompresses pretty fast too
15:48:18 <alise> http://www.chiark.greenend.org.uk/~sgtatham/algorithms/listsort.html has a complete pseudocode mergesort that, though i haven't read it, seems to be optimised for doing it iteratively on a linked list
15:48:33 <alise> ais523: how much more %s of compression do you get from the slow tweaks?
15:48:36 <zzo38> Do you know quantum bogosort? The algorithm is this: [1] Arrange the files at random. [2] If it is sorted, stop. [3] If it is not sorted, destroy the entire universe.
15:48:50 <alise> "A perl script com piles the gram mar file into OCaml code (snarxiv.ml):" ;; Dear god.
15:48:57 <alise> *compiles *grammar; dunno why that happened.
15:49:09 <alise> Oh, ­ hyphens, I think.
15:49:13 <ais523> alise: less than 1%, which is what's annoying me so much
15:49:18 <ais523> they looked really good on paper
15:49:36 <alise> ais523: then get rid of them, stick a mergesort in, and rewrite it in C if it isn't already
15:49:50 <ais523> why would I write it in Java
15:50:02 <alise> dunno, you seem to be using the language for things, inexplicably
15:50:05 <ais523> the main reason it's in C is because I was doing a bunch of pointer manipulation
15:50:16 <ais523> alise: well, it's my job, that's a reason to be familiar with it
15:50:22 <ais523> and I only have one major project in Java
15:50:26 <alise> ha, from the same guy as snarXiv: http://davidsd.org/theorem/
15:50:28 <zzo38> Write it in Enhanced CWEB, and then make a book of it.
15:50:29 <alise> random theorem generator
15:50:43 <alise> zzo38: I don't think this compressor is begging to be a book...
15:51:27 <ais523> my roguelike dynamic routing algorithm is much more deserving of a paper, because that one's actually useful
15:51:32 -!- cheater99 has quit (Ping timeout: 265 seconds).
15:51:34 <ais523> I should write it down mathematically sometime
15:52:15 <zzo38> ais523: Then write that roguelike algorithm in Enhanced CWEB and make *that* one a book.
15:52:23 <alise> but Enhanced CWEB is utterly insane...
15:52:39 <zzo38> (You can mix C codes with normal mathematical equations as well, if it is necessary)
15:53:17 <zzo38> alise: You think it is insane? Maybe it is insane but I find it useful. I like to use it to write C programs with.
15:53:28 <ais523> what I actually did with the routing algo was to make it into a YouTube video
15:53:29 <ais523> http://www.youtube.com/watch?v=udCNcoIfYIc
15:53:41 <alise> ais523: YouTube, the future of mathematical journals.
15:53:44 * ais523 waits for alise to not believe me because it's so far out of character for me
15:53:54 <alise> I actually have Flash so I checked.
15:54:02 <alise> ais523: Why doesn't TAEB play with DECgraphics and colours? Sheesh!
15:54:11 <zzo38> A video? It would be better as a text, isn't it?
15:54:15 <ais523> it does play with colors
15:54:19 <ais523> but it doesn't show them on its debug screen
15:54:30 <ais523> because colors are used for something else there
15:54:35 <alise> zzo38: he /could/ have published a termcast, but...
15:54:58 <alise> ais523: TAEB can think faster and better than me, I feel kind of inadequate
15:55:09 <alise> I have a strong urge to eliminate the competition.
15:55:12 <ais523> it's actually slower than most humans
15:55:21 <alise> Well gee, that just makes me feel even better.
15:55:48 <zzo38> alise: Even as a termcast or ttyrec, I think a paper could better explain it, especially if you wrote it in Enhanced CWEB (but even a normal text could do).
15:56:05 <alise> zzo38: It's just a video of it in action!!
15:56:08 <ais523> well, I made the video by pointing a screen capture program at ipbt
15:56:11 <alise> ais523: So what sense of routing is being used here?
15:56:16 <alise> I'm not entirely sure what you coded.
15:56:33 <ais523> alise: it's a straightforward shortest-path routing algorithm
15:56:46 <alise> "The terminal emulator used is the one from the PuTTY SSH client (simply because that's the one whose API I was most familiar with)." ;; conveniently not mentioning he wrote it
15:56:59 <alise> ais523: what's special about it -- works on terrain you don't entirely know?
15:57:04 <ais523> but it's caching a) between turns, even if the goal and the player have moved; b) allowing for changes in the map
15:57:18 <ais523> so it's much faster than running Dijkstra or A* on the sort of workloads TAEB::AI::Player needs
15:57:19 <alise> so is Planar better than main TAEB yet?
15:57:55 <ais523> no in terms of raw score, yes in most other respects
15:57:55 <ais523> but I haven't worked on it for over a year now
15:57:57 <ais523> btw, when the whole map flashes green or blue, it's using a different algo
15:58:10 <alise> also, I find your YouTube video inadequate; re-upload it with at least 1080p resolution!
15:58:26 <ais523> either because there are monsters around, which change the caches too quickly for them to be useful
15:58:50 <ais523> or because there's just been a major update and it's faster to recalculate from scratch than to diff
15:58:51 <alise> ais523: does TAEB ever use gX?
15:59:17 <ais523> no, but it uses travel when moving four or more spaces in known terrain
15:59:25 <alise> also, what level of the dungeon is this? the display doesn't say
15:59:29 <alise> unless D:2D means something
16:00:52 <alise> http://undergrad.davidsd.org/theorem/applications.html ;; haha (the author of the random theorem generator getting a princeton student to nod his head along to the rubbish)
16:02:18 <alise> ais523: are you a licensed NIH Therapist?
16:02:32 <alise> i feel a horrible urge to make a new OS and a new language again...
16:02:36 <ais523> NIH = Not Invented Here? or National Institute of Health?
16:02:55 <ais523> well, doing something yourself for fun makes sense
16:03:17 <alise> nonono, I have a deep, unflinching belief that everything sucks
16:03:21 <alise> and that i can do so much better
16:03:27 <ais523> but if you're doing something for someone else, it helps to try to make sure it's better than the existing alternatives, and basing it on them makes sense
16:03:47 <ais523> jettyplay came out of a belief that all existing ttyrec player sucked; it's not finished yet, but I think it's better than all the competition already
16:03:56 <ais523> now I've finally tracked down all the known memory leaks
16:04:02 <alise> ais523: at least your project is small
16:04:14 <alise> my language is probably not even implementable efficiently with current CS knowledge!
16:04:16 <ais523> yep, and the competition not only finite, but verifiably so
16:04:58 <zzo38> Explain how Enhanced CWEB is insane, in your opinion.
16:05:01 <ais523> hmm, apparently the maker of I Wanna Be The Fangame is one of the only two people to have ever completed the original
16:05:09 <alise> zzo38: i'm not sure, it just ... us
16:05:09 <ais523> that sort-of makes sense, although I wouldn't have expected it
16:05:16 <ais523> (completed the original on highest difficulty, that is)
16:05:17 <alise> ais523: no, more than one person has completed the actual one
16:05:19 <alise> maybe that's old info
16:05:23 <alise> it's in the double digits
16:05:30 <alise> i forget exactly how many
16:05:31 <ais523> well, I was only aware of two
16:05:35 <ais523> but that's good to know
16:05:54 <alise> the general consensus seems to be that once you get used to it, you can just memorise all the levels and complete it
16:06:06 <alise> since i guess you can adjust to the insane bastard-physics
16:06:45 <alise> I wonder if they've actually put virtualisation support on my chip
16:06:47 <alise> but just locked it out?
16:06:50 <ais523> I agree with that, I think
16:06:59 <ais523> it's memorisation + steady hand platforming
16:07:10 <alise> "Modern laptops by Sony use, instead of a BIOS, a new approach called EFI." ;; I did not know this.
16:07:16 <alise> Sony laptops are on a collision-course with Apple.
16:07:23 <alise> I think they will release the exact same model within three years.
16:08:44 -!- cheater99 has joined.
16:09:17 <alise> nobody loves my poor U4100 processor
16:09:29 <alise> with its 1.3 GHz speed and its 2 MiB cache
16:09:36 <alise> model name: Genuine Intel(R) CPU U4100 @ 1.30GHz
16:09:39 <alise> guess it's downclocked itself
16:10:49 <ais523> due to overheating, or thinking it is?
16:11:06 <ais523> hmm, you have pretty much the same sort of laptop as me, don't you?
16:11:09 <olsner> or maybe it's just bored
16:11:18 <alise> ais523: all modern laptops downclock when not under load
16:11:19 <ais523> try rapping it a few times just beyond the top-left corner of the keyboard
16:11:22 <alise> to conserve battery
16:11:34 <ais523> sometimes the fan gets stuck, and doing that frees it up
16:11:38 <alise> my fan has never got stuck :P
16:11:44 <alise> it's not a bug that it's downclocked
16:11:52 <alise> also, i think this laptop may be better-built than yours, being bigger and more expensive
16:11:54 <alise> although essentially the same
16:12:02 <alise> one would think the greater internal spreadout would leave less room for jamming fans
16:12:22 <alise> actually sometimes the fan on this goes off altogether because the cpu isn't even generating enough heat to be worth it
16:12:55 * alise downloads the NixOS live CD
16:13:08 <alise> Purely functional package and configuration management, it's gotta be better than the rest, right?
16:13:21 <ais523> the fan on this is /usually/ off altogether
16:13:24 <alise> ais523: well, me too
16:13:26 <ais523> while it's spinning, it doesn't get stuck
16:13:30 <alise> but i often have like
16:13:33 <ais523> it's when it tries to turn on that the problem happens
16:13:40 <alise> and 30 tabs in each browser
16:13:43 <alise> well, not that extreme
16:13:46 <alise> but you get my point
16:13:51 <alise> and switching windows is...
16:13:57 <alise> (never slow, but you can tell it's working)
16:15:34 <ais523> tabs don't use CPU, though, do they? just memory?
16:15:42 <ais523> unless they're trying to run Flash or something silly like that?
16:16:01 <alise> i'm not sure *why* switching windows causes it to work a bit more if you have a lot of them
16:16:03 <alise> come to think of it
16:16:05 <alise> i may be imagining it
16:16:13 <alise> my relation with computers is very... superstitious
16:16:30 <alise> i don't quite trust that they're actually doing something reasonable when i ask them to do something, because they're told to do *so* *much* stuff
16:17:13 -!- nooga has left (?).
16:17:13 -!- nooga has joined.
16:17:17 -!- nooga has left (?).
16:17:22 <alise> I'm very sorry, but the following error(s) occurred:
16:17:22 <alise> Product /nix/store/l3q29m17x8sdribzmnz9lr1kg2l7hic4-nixos-manual/share/doc/nixos has disappeared.
16:17:24 -!- nooga has joined.
16:17:26 <alise> Well that's not very purely functional, is it?
16:20:28 <alise> ais523: I had a strange language idea where modules were actually Erlang-style processes.
16:20:40 <alise> And f(x) was just an object, Prolog-style.
16:20:53 <alise> io<-stdout<-print("Hello, world!\n")
16:21:02 <alise> <alise> And f(x) was just an object, Prolog-style.
16:21:13 <alise> predicate(print("Hello, world!\n"))
16:21:16 <alise> and it's just an object
16:21:19 <alise> you know what i mean
16:21:21 <alise> in the Scheme sense
16:24:15 <ais523> I think "term" is the "official" term for that
16:24:38 <ais523> I love the way it's entirely possible to use, say, 3-4 as an inert piece of data
16:24:40 <alise> "term" usually means "expression". Sometimes.
16:24:41 <alise> http://arcanesentiment.blogspot.com/2010/09/shorter-words-for-expression.html
16:24:56 <alise> arcanesentiment.blogspot.com/2010/09/shorter-words-for-expression.html
16:25:04 <ais523> because although it means '-'(3,4), it isn't interpreted as arithmetic unless you actually try to evaluate it as an arithmetic expression
16:25:37 <alise> but yeah, that's the idea I had
16:25:47 <alise> you could also have things like, I don't know, a receiver for a socket
16:25:49 <alise> which would end up having
16:25:56 <alise> receiver <- "packet"
16:25:59 <alise> as opposed to a function call
16:26:04 <alise> I guess it was some sort of unification, but.
16:26:42 <alise> (Arcane Sentiment is an excellent blog, btw.)
16:27:27 <nooga> http://snarxiv.org/
16:27:32 <nooga> this is extreely fun
16:27:45 <alise> http://snarxiv.org/vs-arxiv/ is funner.
16:33:14 -!- poiuy_qwert has joined.
16:33:33 <alise> ais523: how much RAM does your laptop have?
16:33:53 <ais523> enough that Windows 7 actually manages to start, eventually
16:33:54 <ais523> although I don't run it nowadays
16:34:31 <alise> $(free -m), Mem - total :P
16:39:10 <ais523> not entirely sure which unit it's using
16:39:16 <alise> bi, almost certainly.
16:39:22 <nooga> should i make something like http://snarxiv.org/vs-arxiv/ for prolog bits?
16:39:26 <alise> ais523: I have closer to 4 GiB. But the funny thing is, I find it hard to believe.
16:39:43 <alise> Because my CPU has quite a low GHz, and... well... it's so small! So small! and my desktops all have less memory...
16:39:49 <alise> I keep thinking I don't have 4 GiBs to play around with
16:39:52 <alise> and worry about allocating too much to VMs
16:40:38 <alise> wow, I forgot how ugly 16-bit colour is with modern UIs
16:40:49 <alise> they try and use their full-colour images, so the dithering is awful
16:41:30 <alise> nooga: what do you mean :P
16:41:31 <nooga> prolog is too simple
16:44:00 <nooga> a page that woul let you to guess which function is meaningless and which is a real one
16:51:26 <alise> ehird@dinky:~/NixOS$ dd if=/dev/zero of=hd bs=1k count=0 seek=8388608
16:51:26 <alise> 0 bytes (0 B) copied, 1.8299e-05 s, 0.0 kB/s
16:51:28 <alise> Damn that was quick.
16:52:31 <Vorpal> alise, does python have any function like is-int? or such
16:53:08 <alise> Vorpal: you don't want to.
16:53:20 <alise> isinstance(x, int)
16:53:27 <alise> but it's generally considered bad -- then again, who gives a shit?
16:53:30 <alise> Vorpal: not really
16:53:32 <alise> it's just x is-a int
16:53:43 <alise> it just checks that the class is either that class or a subclas
16:53:44 <Vorpal> fine then, since I need to invoke it quite often
16:56:07 <olsner> if you need it to be fast, just make it so you don't have to check anything at all
16:56:37 <alise> if you need it to be fast, make it a lookup table
16:56:40 <alise> THEN REMOVE THE LOOKUP TABLE
16:58:04 <zzo38> Make a graph of the IRC of: Uppercase letters, lowercase letters, digits, punctuation, line length, words per line, NOTICE messages, ACTION messages, messages containing CTRL+A but not ACTION, URLs, replies, user registration, QUIT messages, idle time, etc
16:58:50 <alise> THEN REPLACE YOUR CODE WITH NOPS
16:58:52 <Vorpal> olsner, well, fast as in "not excessively slow" here.
16:58:53 <alise> THEN REMOVE THE NOPS!!!!
16:59:52 <alise> Vorpal: did you know that linux consoles can have a background image, and be centred???
17:00:05 <Vorpal> bg image yes, and centered as in the image?
17:00:15 <alise> NixOS's console is centred with a background that has a translucent, glow-bordered area
17:00:18 <alise> in which the actual console is placed
17:00:21 <alise> it's freaking me out
17:00:26 <olsner> Vorpal: seeing as python already is excessively slow, I think you'll be fine
17:00:29 <Vorpal> you mean center as in smaller buffer than screen or centered as in <center>?
17:00:31 <alise> (translucent = dimmed)
17:00:35 <EOF> ARM SUPERCOMPUTER FTW
17:00:42 <Vorpal> alise, hm okay, that bit I didn't know
17:00:44 <alise> so it looks like you have a title-less, translucent xterm on top of a background image
17:01:09 <EOF> hello vorpal
17:01:18 <EOF> hello alise
17:01:26 <EOF> and hello zzo38
17:02:58 <EOF> the ARM archetecture is BEAUTIFUL!!!
17:03:36 <EOF> i want to hug it
17:03:58 <zzo38> EOF: Do you know about the MMIX?
17:04:02 <EOF> ARM has breasts
17:04:25 <EOF> MMIX better get going
17:04:38 <zzo38> EOF: How is that relevant? And that doesn't even make sense anyways. How can ARM have breasts?
17:05:06 <EOF> because it's beautiful
17:06:23 <alise> EOF: can you stop being fucking insane, or just go away?
17:07:02 <zzo38> "because it's beautiful"? I am not sure I follow.....
17:08:14 <EOF> i made a program for the sheevaplug and it calculates nipple positions from photos if you know the light angle
17:08:22 <EOF> and it works
17:08:58 <EOF> anybody want the code?
17:09:10 <EOF> or should o sell it to th military
17:09:17 <zzo38> EOF: So now you made a program that can find things in pictures
17:09:28 <zzo38> You should make the code more generalized
17:09:43 <EOF> it finds nipple positions
17:09:47 <EOF> nothing more
17:10:13 <EOF> using approximations of the golden ratio
17:10:16 <zzo38> EOF: Yes I know that, but unless it can be modified to do more generalized stuff, it isn't that useful
17:10:51 <EOF> i guess it could find the majority of other measurements in the human body
17:11:10 <EOF> with the right approach
17:11:33 <zzo38> I suppose that is a possibility
17:13:22 <EOF> motion tracking for long range computerized riflemen?
17:13:47 <EOF> aka sniper sentries :)
17:16:28 <EOF> ARM's breasts AR M cups
17:18:09 <alise> EOF: you are possibly the stupidest person ever to visit this channel.
17:18:44 <Vorpal> fizzie, the matching up works fine up until netsplit, but handling netsplit seems near impossible
17:18:52 <Vorpal> especially the rejoining after
17:22:40 <zzo38> alise: You may be correct. It does seem like that.
17:22:51 <EOF> i was once required by an employer to use FORTRAN, i quit
17:23:04 <EOF> FORTRAN EWW!!!
17:23:16 <alise> Fortran was a perfectly good language for its time.
17:23:43 <zzo38> EOF: If you don't like FORTRAN, that is OK. Different people can prefer a different program language.
17:24:00 <EOF> yeah, i suppose, but it was invented for the IBM 704
17:24:18 <EOF> prior to tpedecks
17:24:37 <EOF> dip switches and punchcards
17:29:48 <zzo38> I invented the "Pokemon Keyboard".
17:31:45 <zzo38> In typing mode, the keys are: 0/A, 1/B, 2/C, 3/D, 4/E, 5/F, 6/G, 7/H, 8/I, 9/J, minus/K, plus/L, multiply/M, divide/N, dot/O, colon/P, comma/Q, semicolon/R, equal/S, not-equal/T, less/U, greater/V, left-parenthesis/W, right-parenthesis/X, quote/Y, question/Z, hash/apostrophe, infinity/male, currency/female, SPACE, NUM/ALPLA, left, right, RUBOUT, ENTER, CLEAR, up, down, SPECIAL, CANCEL.
17:32:45 <zzo38> In non-typing mode, the keys are: NULL, GRASS, FIRE, WATER, BUG, POISON, GHOST, ROCK, GROUND, NORMAL, FLYING, FIGHTING, PSYCHIC, ELECTRIC, ICE, DRAGON, DARK, METAL, PICTURE, OTHER, LEVEL, EVOLVE, SEARCH, AREA, STATUS, ORDER, HEIGHT, WEIGHT, TEXT, MODE, EDIT, PREVIOUS, NEXT, DELETE, SAVE, SEND, PROGRAM, PRINT, TIME, LOAD.
17:34:08 <Sgeo|web> http://www.youtube.com/watch?v=0kh4xhem8tM&feature=player_embedded
17:42:10 <Sgeo|web> Hey, so I'm only second stupidest?
17:44:11 <Sgeo|web> http://www.youtube.com/watch?v=191ZS217zOY&feature=related
17:44:33 <Sgeo|web> Obviously, I'm only considering people in this channel, not the morons in the videos I'm linking
17:45:50 <alise> Sgeo|web: she is so shitfaced in that second video
17:45:58 <alise> "You're stupid, you're stupid!" ":D :D XD :D :D :D"
17:45:58 <zzo38> Sgeo|web: I do understand that is what you mean. (However, I did not watch the video.)
17:46:13 * Sgeo|web looks for the remainer of what she said
17:47:14 <Sgeo|web> " Angle: Actually, Thomas Jefferson has been misquoted, like I've been misquoted out of context. Thomas Jefferson was actually addressing a church and telling them through his address that there had been a wall of separation put up between the church and the state precisely to protect the church" "
17:47:44 <Sgeo|web> ...does she think that Separation of Church and State doesn't do that? That that's not what it is?
17:47:47 <fizzie> Vorpal: I'm not sure what's wrong with netsplit; isn't it just so that those messages will end up being not-matched and therefore both end up in the final result? (And after the split is over you should again start seeing matches.) (But I don't know what you've done there, so... also, still quite away.)
17:47:52 <Sgeo|web> It protects the church. All churches.
17:48:24 <alise> "Church" doesn't mean "little church building" in this context...
17:48:49 <fizzie> Vorpal: As for the Python type-checking, you can test like "if type(x) is int" or "if type(x) is not int"; I guess that won't handle inheritance though.
17:49:10 <alise> fizzie: that's evil :P
17:49:11 <Sgeo|web> There is a Python function for checking for inheritance
17:49:27 <fizzie> Sgeo|web: The isinstance() alise mentioned?
17:50:51 <fizzie> alise: Also, I did that "fixed sample set size" test, and ended up with a real mess: http://zem.fi/~fis/esomapf.png
17:50:58 <Vorpal> fizzie, the issue is matching up, since joins on different sides won't match
17:51:07 <Vorpal> if I was on a different side than clog
17:51:20 <zzo38> fizzie: What is this graph plotting?
17:51:32 <Vorpal> fizzie, I might be better off just trying to track msg/act
17:51:32 <alise> fizzie: That one's the other one in seventy-gazillion bajillion years.
17:51:48 <fizzie> Vorpal: Is it a particular problem if both sets of joins end up in the log?
17:51:54 <alise> The galaxies merging and shit.
17:52:11 <alise> So what VMs are there apart from VirtualBox and QEMU? For Linux that is.
17:52:47 <Vorpal> fizzie, well that is one issue. There is another scenario as well: one side see it as a netsplit, the other times out
17:52:54 <Vorpal> which happens in a few places
17:53:01 <Vorpal> but the first is a lot more problematic
17:53:09 <zzo38> Can you tell me what this graph is plotting, please? Like, what is the X axis for, what is the Y axis for, what are the shapes means, etc
17:53:44 <fizzie> zzo38: It's a scatterplot of the two first PCA components of 60 different writing-style-related features -- http://zem.fi/~fis/esomap-comp.png has some of the raw features -- where each x is from features computed over 2000 messages.
17:54:27 <fizzie> zzo38: http://zem.fi/~fis/esomapnz.png uses a lot larger sample sizes, so there the people actually can be separated.
17:54:47 <Sgeo|web> My dad dislikes Obama, but only because of economic... thingies that I fail to understand
17:55:01 <Sgeo|web> For all I know, the Republicans are 100% correct when it comes to the economy
17:55:09 <Vorpal> fizzie, also messages can get lost before the point the other end see a quit
17:55:10 <fizzie> zzo38: And the ellipses are just contour plots at one and two standard deviations of normal distributions fitted to corresponding points.
17:55:15 <zzo38> fizzie: Do you have a full list of the components and the features?
17:55:15 <Sgeo|web> But the social stances in general I cannot agree with
17:55:45 <Sgeo|web> My dad also thinks of the Tea Party in terms of its economic positions
17:56:16 <fizzie> zzo38: http://p.zem.fi/esomap-feats -- the "foo (n)" ones are like histogram bins there.
17:56:28 <Vorpal> fizzie, in the fixed set only you and zzo can be cleanly separated heh
17:56:50 <Vorpal> fizzie, at least for those two components
17:56:52 <fizzie> And "Punctuation frequency (ellipsis)" is in fact bugged and always 0 for everyone. :p
17:57:24 <Vorpal> fizzie, what does the ,1 for each one mean?
17:57:58 <fizzie> Vorpal: It's just the way octave prints out all cell arrays; it's a 60x1 array of strings, and it always prints out at least two indices.
17:58:18 <Vorpal> fizzie, hm what is Type/Token?
17:59:21 <fizzie> It's "count of unique words"/"count of all words".
17:59:22 <zzo38> Do you have a copy of the raw data for everyone (and statistics, such as mean/median/stddev/variance, for each feature)?
17:59:29 <Vorpal> fizzie, also, is there any specific reason to not include line length in characters? After all that previous plot cleanly separated you from me and alise when you did frequency of line lengths...
17:59:51 <Vorpal> zzo38, raw data I assume is the logs in topic
17:59:52 <fizzie> Vorpal: That's in there, it's the "[30,1] = Chars per paragraph (mean)".
18:00:44 <zzo38> Vorpal: OK, what if I use those I would also need the algorithms, which I do not have.
18:01:52 <fizzie> I have the raw computed features in MATLAB's binary format (readable with Octave too) if someone's so curious. http://zem.fi/~fis/esofea-n10.mat and http://zem.fi/~fis/esofea-m2000.mat for the esomapn.png and esomapf.png features, respectively. The data structures in the files might be a bit screwy, though.
18:02:01 <Vorpal> fizzie, is there any way to calculate/visualise "separability" when taking all PCA axises in account? Maybe as a graph with distance from a given nick to the other ones, one per nick of those.
18:02:01 <fizzie> (They're written to work with the MATLAB SOM toolbox.)
18:02:57 <fizzie> Vorpal: You could graph serious "separability" by training some sort of classifier with the data, and then plotting the confusion matrix; I'll do that soonishly.
18:03:09 <Vorpal> fizzie, "confusion matrix" :D
18:03:21 <Sgeo|web> http://gawker.com/5603331/comical-sharron-angle-needs-press-to-be-her-friend
18:03:27 <Vorpal> fizzie, what does a confusion matrix mean?
18:04:58 <fizzie> Vorpal: It's just the recognition results, so that rows match to the correct labels, columns match to the classifier results, and a value is in (i,j) if it actually has the label i but the classifier gives it j. You'd want the confusion matrix to be as diagonal as possible, since all values on the diagonal correspond to correct classifications.
18:05:00 -!- oerjan has joined.
18:05:13 <alise> Angle: "Well, no. We wanted them to ask the questions we want to answer so that they report the news the way we want it to be reported."
18:05:37 <Vorpal> alise, didn't you switch nick with Gregor some year(s) ago? For a few hours or such? And when you revealed what was going on it turned out no one had detected it.
18:05:46 <Vorpal> I seem to have a vague memory of that
18:06:16 <alise> Through a complicated number of swaps to each other's and other nicknames that appeared to restore itself but didn't.
18:06:24 * oerjan also has a vague memory. of that, and in general.
18:06:46 <Vorpal> alise, I guess that could both explain why you end up so near Gregor and also perhaps confuse the algorithm to place you even nearer. Or rather, place Gregor even nearer (due to your much higher volume I suspect you won't be affected as much)
18:06:59 <alise> Vorpal: it was for a short period of time
18:07:09 <Vorpal> true, but the first point still stands
18:08:00 <Vorpal> Gregor, did you drop an R hm?
18:08:12 <Vorpal> fizzie, do you merge Gregor and GregorR in those plots?
18:08:50 <fizzie> Vorpal: I think I merged in all that are like 'gregor%' there.
18:09:16 <Gregor> Makes sense since up 'til recently I used GregorR?(-[A-Z])?
18:09:44 <fizzie> Vorpal: Also, here's a confusion matrix for letters: http://zem.fi/~fis/q3.png -- you can see things like "oh, s is very easy to classify" and "k/p/t and m/n are easily confused".
18:10:28 <Vorpal> fizzie, is that for OCR? Or for the problem at hand?
18:10:43 <fizzie> Vorpal: It's some old speech recognition course homework.
18:10:55 <Gregor> So where's this nick-associability graph you're all referring to?
18:11:00 <Gregor> I don't see it in the logs (yet)
18:11:02 <alise> fizzie: /~fis/ must be so huge.
18:11:08 <Vorpal> fizzie, Finnish or English?
18:11:23 <fizzie> Gregor: http://zem.fi/~fis/esomapn.png is one.
18:12:01 <Vorpal> Gregor, another one is http://zem.fi/~fis/esomapnz.png
18:12:12 -!- Sgeo|web has quit (Ping timeout: 252 seconds).
18:12:14 <Vorpal> fizzie, why is there no http://zem.fi/~fis/esomapfz.png ?
18:12:36 <fizzie> Vorpal: Because esomapf.png already has a zzo in it.
18:12:48 <Vorpal> fizzie, what about a non-z version of that?
18:12:49 <fizzie> Gregor: I said this like *right now*.
18:13:15 <fizzie> Gregor: <fizzie> zzo38: It's a scatterplot of the two first PCA components of 60 different writing-style-related features -- http://zem.fi/~fis/esomap-comp.png has some of the raw features -- where each x is from features computed over 2000 messages.
18:13:41 <Gregor> See, this is why I don't like using a BNC :P
18:13:55 <Gregor> It appears I've been here for hours, when in actuality my BNC's backlog didn't even cover that.
18:14:26 <alise> Gregor: So... don't?
18:14:32 <fizzie> Oh, I've configured bip to replay absolutely everything that has happened after I last said something/manually reset the backlog/last connected.
18:14:38 <Vorpal> Gregor, so set it to log further than that?
18:14:46 <fizzie> Not that I usually read the backlog, though, so I guess it won't help. :p
18:15:09 <fizzie> Vorpal: The confusion matrix is lacking äö, so it might even be English. In any case it was an phonemes-out-of-acoustics only trivial "recognizer", so it's actually a bit language-agnostic in that it doesn't care about words.
18:16:38 <Vorpal> fizzie, but not all languages have all phonemes afaik?
18:16:58 <alise> ls: cannot access /lib: No such file or directory
18:17:04 -!- jcp has quit (Quit: Later).
18:17:09 <alise> ls: cannot access /usr: No such file or directory
18:17:19 <Vorpal> tell me what that does
18:17:30 <Vorpal> alise, and lists what?
18:17:30 <Gregor> Ohbtw: <alise> Gregor: So... don't? // but now my hostname is codu.org :P
18:17:35 <alise> bin boot dev etc home media mnt nix nix-store.squashfs proc root sys tmp var
18:17:45 <Vorpal> alise, what is nix-store?
18:17:45 <fizzie> Vorpal: Right, and English phonemes tend to have screwier names than just letters; so I think it is Finnish. I think it was from a three-sentence datasets, so it might just not have any äö in it.
18:17:46 <alise> Gregor: I can think of a million other ways to achieve that.
18:17:52 <alise> Gregor: Do you use a graphical client?
18:18:24 <Vorpal> fizzie, you don't have stuff like combination of letters making new sounds in Finnish? Like sj in Swedish
18:18:45 <alise> Gregor: Here, I've thought of one already: Set up a tiny, tiny server on a port on codu.org that, when connected to, waits for a NICK/USER/PASS line, checks the PASS line, then creates a connection to irc.freenode.net, sends off those three lines, and simply directly relays everything between the two.
18:18:53 <alise> This connection is, obviously, broken when your client or the server breaks it.
18:19:02 <alise> This would be, what, 30 lines of code?
18:19:05 -!- jcp has joined.
18:19:14 <Gregor> alise: So, a bnc that doesn't remain connected.
18:19:21 <alise> Gregor: BNCs are much more involved...
18:19:28 <alise> They have "commands" and stuff!
18:19:31 <oerjan> <quintopia> Vorpal: then what was that you just posted above? did you, *gasp* MAKE YOUR OWN PAN? <-- Ceci n'est pas un pan
18:19:59 <alise> Vorpal: Can you tell what it is yet?
18:20:09 <Vorpal> alise, not all bncs have commands
18:20:15 <alise> Vorpal: Err, not that.
18:20:25 <alise> The disturbingly-empty filesystem tree.
18:20:36 <fizzie> Vorpal: Not much, though I think "ng" is often considered a different phoneme in Finnish. Still, Finnish text-to-phoneme mapping is usually done with a (very small) set of rules, as opposed to English which always needs a pronunciation dictionary.
18:20:47 <Vorpal> alise, well you had nix-store.squashfs in it
18:20:54 <Vorpal> alise, and you didn't answer my question:
18:20:57 <Vorpal> <Vorpal> alise, what is nix-store?
18:21:06 <alise> Vorpal: I won't answer something /that/ simple.
18:21:11 <Vorpal> alise, I presume it is something called nix-store
18:21:15 <alise> I'm afraid I will require more mechanical commands.
18:21:20 <alise> Or very simple questions.
18:21:20 -!- Sgeo|NeedUSB has joined.
18:21:31 <Vorpal> alise, I'm not sure I'm interesting in guessing
18:21:51 <fizzie> alise: .squashfs makes it sound like some sort of embedded-linux, maybe a router or wlan-ap or whatnot.
18:22:02 <Vorpal> alise, but what does ls /nix do?
18:22:26 <Vorpal> alise, oh and some other stuff:
18:22:33 <alise> Although it also had "store" there a minute ago.
18:22:43 <alise> uname is /var/run/current-system/sw/bin/uname
18:22:54 <alise> (ls /nix/store does indeed work.)
18:23:06 <alise> Input/output error. I think I broke something.
18:23:18 <Vorpal> alise, what about: uname -a
18:23:27 <alise> Same. I think it doesn't like the VM.
18:23:50 <Vorpal> alise, well... does other commands that are not shell builtins still work?
18:24:27 <Vorpal> alise, anyway, is it linux-based or other (such as *bsd or whatever)
18:25:10 <EOF> why would somebody run linux in a VM if they could just fuse kernels
18:25:13 <fizzie> I'm not sure how well squashfs is supported on anything not a Linux.
18:25:19 <alise> ls -l ==> Bus error, lol, I have really fucked something up here
18:25:24 * oerjan sneakily steals his saucepan back
18:25:40 * EOF thinks that wasn't so sneaky
18:25:40 <Vorpal> alise, well then the answer is simple, it is some crazy linux distro along the lines of gobolinux when it comes to crazy organisation of the filesystem :P
18:25:44 <alise> Vorpal: I'll just give it away: it's NixOS, using the purely-functional package manager Nix.
18:25:48 <alise> similar, yes, but it's crazy in a cool way
18:26:08 <alise> every configuration or package change is versioned
18:26:09 <Vorpal> why the squashfs though?
18:26:11 <alise> so you can run trunk all the time
18:26:16 <alise> since you can just boot the previous configuration in grub
18:26:18 <oerjan> EOF: well you're not the one who isn't supposed to notice.
18:26:24 <alise> /nix/store/r8vvq9kq18pz08v249h8my6r9vs7s0n3-firefox-2.0.0.1/
18:26:29 <alise> the hash is a hash of the dependency graph
18:26:49 <alise> and the directory is immutable
18:26:59 <alise> and users can install multiple local versions of a package and it use the same tree
18:26:59 <Vorpal> alise, a pitty the whole OS is not purely functional. Since then fucking it up would be considered a side effect and thus be hard to perform :P
18:27:01 <alise> (because they're verified)
18:27:41 <alise> When you install a package like this…
18:27:41 <alise> $ nix-env --uninstall firefox
18:27:41 <alise> the package isn’t deleted from the system right away (after all, you might want to do a rollback, or it might be in the profiles of other users). Instead, unused packages can be deleted safely by running the garbage collector:
18:27:41 <alise> $ nix-collect-garbage
18:27:43 <alise> This deletes all packages that aren’t in use by any user profile or by a currently running program.
18:27:45 <alise> You can garbage collect the packages!
18:28:01 <alise> Oh yeah, and the package manager works on other distros too.
18:28:25 <Vorpal> alise, I'm not sure the native package manager wouldn't clobber it
18:28:54 <Vorpal> I'm pretty sure dpkg clobbers stuff just fine if it doesn't think it handles them itself and they are not in /etc or such
18:30:29 <alise> Vorpal: it doesn't
18:30:33 <alise> since it's self-contained
18:30:36 <alise> here's the really cool thing
18:30:43 <alise> in NixOS, the CONFIGURATION is managed with nix too
18:30:52 <alise> boot.loader.grub.device = "/dev/sda";
18:30:52 <alise> { mountPoint = "/";
18:30:53 <alise> device = "/dev/sda1";
18:30:59 <alise> services.sshd.enable = true;
18:31:03 <alise> that's an example minimalist full-machine configuration
18:31:09 <Vorpal> alise, I didn't say nix clobbered, but that the package manager might unless you use a prefix for the nix stuff
18:31:10 <alise> so you can roll this back even if you make it not boot properly
18:31:12 <alise> (maybe not if you fuck up grub)
18:31:23 <Vorpal> alise, also the name is highly likely to cause confusion
18:31:25 <alise> Vorpal: it all goes in /nix/store :P
18:31:31 <alise> it's an established project
18:31:35 <Vorpal> alise, not /bin/sh though?
18:31:36 <alise> research papers, large package set
18:31:40 <alise> (full KDE 4, many many packages)
18:31:46 <alise> small but established userbase
18:32:05 <alise> Vorpal: I think that's because too much stuff breaks without it, dunno
18:32:09 <zzo38> If I write distro, I will write a new package manager, called "pm", which receives package files (which can be concatenated) from stdin, and does stuff depending on the command-line options, sending results to stdout and status messages to stderr.
18:32:31 <Vorpal> alise, there are a few more in that category, /usr/bin/env is very common
18:32:48 <alise> Vorpal: i'm not sure what they do for that. i'd have to checl.
18:33:17 <Vorpal> alise, does it boot though?
18:33:29 <alise> Boot? Of course it boots. People use it as their main system.
18:33:40 <Vorpal> alise, I mean after you fucked it up
18:33:51 <alise> i'm not sure how i fucked it up but i sure did
18:33:52 <Vorpal> alise, that explains the squashfs
18:34:08 <Sgeo|NeedUSB> How does being a purely functional package manager help with any of that
18:34:30 <alise> Sgeo|NeedUSB: what?
18:34:42 <Sgeo|NeedUSB> "Nix is a purely functional package manager. This means that it can ensure that an upgrade to one package cannot break others, that you can always roll back to previous version, that multiple versions of a package can coexist on the same system, and much more.
18:35:00 <alise> the package management is purely functional, not just the code
18:35:15 <alise> it's not "this is written in Haskell" it's "packages are literally immutable and rollbackable"
18:35:36 <Vorpal> alise, what happens if you manually edit a file of a package?
18:36:18 <alise> Vorpal: well, at least one of the people there wants to chattr +i the store
18:36:26 <alise> i think you CAN fuck it up as root, but you'd have to be very deliberate
18:38:28 <alise> Vorpal: no /usr/bin/env :D
18:40:41 <alise> It's a LiveCD, what's the worst that could happen?!
18:40:51 <alise> Mm, the silent sound of destruction.
18:40:58 <alise> Anarchic carnage. Chaos rules the land of Nix!
18:41:24 <alise> Sgeo|NeedUSB: What?
18:41:28 <Vorpal> alise, so is there any package monad?
18:41:40 <alise> the kernel panicked :D
18:41:50 <alise> Sgeo|NeedUSB: mmmmmmyes.
18:41:55 <alise> but 2gb should be fine at a pinch
18:41:58 <alise> Sgeo|NeedUSB: NOTE
18:42:01 <alise> boot it from a livecd first
18:42:06 <alise> then use the USB Creator to write it to your drive
18:42:19 <alise> Sgeo|NeedUSB: remember to check the setting that allows ~ to be preserved
18:42:25 <Vorpal> alise, anyway, when they say a package isn't deleted right away does it mean that the app stays in your path and such
18:42:30 <Vorpal> or is it moved elsewhere from there
18:42:35 <Sgeo|NeedUSB> It won't preserve installed packages though, will it?
18:42:37 <alise> Vorpal: it's taken out of the current system
18:42:41 <alise> but still exists in prior systems and in the store
18:42:48 <alise> the current system is just a bunch of symlinks, I think
18:42:52 <Vorpal> alise, does it use unionfs for it or something?
18:42:55 <alise> Sgeo|NeedUSB: uh, maybe. dunno. probably not.
18:43:01 <alise> Vorpal: i'm not sure on the details
18:43:03 <alise> you'd have to ask #nixos
18:43:50 <Vorpal> alise, any multilib support?
18:44:04 <Vorpal> multilib done properly would make me interested
18:44:08 <alise> Vorpal: Probably? Ask #nixos.
18:44:49 <Sgeo|NeedUSB> But I really want Haskell with Erlang-style hotswapping
18:45:37 <Vorpal> "This command does everything necessary to make the configuration happen, including downloading and compiling OpenSSH, generating the configuration files for the SSH server, and so on." <-- hm, so... what if you want custom settings for the sshd, more than just "on/off"?
18:45:54 <Vorpal> do you edit sshd_config then or does it provide some other way for all possible ssh settings?
18:46:08 <alise> you're doing that thing you do
18:46:12 <Vorpal> alise, but you have it running
18:46:14 <alise> assuming that because someone mentions something they're an expert
18:46:22 <alise> Vorpal: so do they
18:46:25 <alise> they also know where all this stuff is
18:46:33 <Sgeo|NeedUSB> Every web framework and its mother has its own templating language
18:46:41 <alise> Sgeo|NeedUSB: not all of them!
18:47:29 <olsner> alise: what, you mean you're not a nixos expert? you've been using this thing for several hours by the sound of it!
18:47:30 <Vorpal> alise, the docs on their website... are very limited
18:48:11 <Sgeo|NeedUSB> Well, the thingy for Smalltalk instead has ... painting thingies
18:48:20 <alise> olsner: minutes actually!
18:48:26 <alise> Vorpal: they have a full manual.
18:48:34 <alise> the NixOS manual link is broken, here is one that works:
18:48:42 <alise> http://hydra.nixos.org/build/638450/download/1/nixos/manual.html
18:48:46 <alise> some issue with their continuous build system
18:49:09 <Vorpal> alise, yes it is very limited. It's the same as the one I saw.
18:50:00 <Vorpal> alise, here I compare to gentoo manual or freebsd manual. Heck even openbsd manual is better than that. It is just a quick start guide compared to those.
18:50:30 <alise> Sgeo|NeedUSB: middle mouse button
18:50:44 <alise> Vorpal: irrelevant. those document all the software packages in the system too
18:50:46 <alise> nix has no need to
18:50:48 <Vorpal> alise, bbl, valvaka (no clue how to translate that, ask olsner or fizzie or someone, though olsner is probably going to watch that too)
18:51:06 <alise> "valvaka", says Google.
18:51:09 <alise> Sgeo|NeedUSB: it is already on
18:51:12 <alise> left and right buttons together
18:51:17 <alise> shift+insert might also work
18:52:28 <olsner> hmm, wake seems to be a specifically funeral ceremony
18:52:32 -!- poiuy_qwert has quit (Quit: Leaving).
18:53:08 <alise> election wake XDDDD
18:53:37 <fizzie> I think it translates to vaalivalvojaiset; HTH.
18:53:53 <olsner> staying up late to watch the counting of the votes live and listen to political analysis and stuff
18:55:02 <Sgeo|NeedUSB> The annoying thing about Try Haskell is that it's more REPL-like and less "Put stuff in, put it in a file, run file" like
18:55:02 <alise> Vorpal: http://wiki.nixos.org/wiki/Main_Page
18:56:12 <alise> Vorpal: nix-build --argstr system i686-linux $NIXPKGS_ALL -A skype_linux
18:57:10 -!- oerjan has quit (Quit: Later).
18:57:42 <fizzie> Sgeo|NeedUSB: You can (IIRC) swap the button meanings in urxvt so that it's "paste" on right-button and "extend selection" (the far less used one) on middle, but I'm not sure I'd do that, I'm afraid it might lead to a habit of right-clicking for paste.
18:58:18 <zzo38> Could you add support for the <xmp> tag to the esolang wiki? (It does not have to actually output the <xmp> tag, it might convert it to other tags.) Also support digits after "xmp", so that you can write <xmp15> ...... </xmp15>
18:58:28 <olsner> aha, luckily wake also means "to remain awake for some purpose, duty, etc." and "to keep watch or vigil." ... either works well with an election
18:59:34 <zzo38> The <xmp> tag on MediaWiki should make it print everything inside the tags as verbatim text, with no templates, signatures, HTML tags, HTML character entities, or other stuff.
19:00:08 <nooga> you shold all try f.lux program
19:00:14 <fizzie> Sgeo|NeedUSB: PuTTY's is configurable too; I think it even calls the paste-on-middle "X-style".
19:00:15 <nooga> it's totally awesome
19:00:37 <zzo38> As well as turning off the spam filter inside of <xmp> tags for registered autoconfirmed users.
19:01:20 <zzo38> nooga: What is f.lux program?
19:01:53 <olsner> oh my, the nationalist party may get into parliament
19:02:12 <nooga> it adjusts color temperature of your display so that you can last whoile night looking at it
19:02:54 <nooga> and it doesn't affect the actual brightness nor contrast
19:03:48 <zzo38> I already adjusted the color temperature on my computer anyways
19:03:57 <alise> zzo38: it updates it dynamically over the night
19:04:03 <alise> so it maintains a constant perceived brightness
19:04:05 <Gregor> Hahaha ad fail. Accidentally clicked a Hulu ad that led me to http://ask.sherwin-williams.com/
19:04:10 <alise> based on your timezone and the light you use
19:04:14 <Gregor> Which is an unconfigured IIS7 install :P
19:04:27 <alise> Gregor: Wow, they stole that Welcome-in-43583475345-languages thing from OS X.
19:04:35 <alise> And the whole spacey-explosion-background-with-big-shadowed-text thing.
19:04:43 <alise> Microsoft: TRYING TOO HARD
19:04:54 <Gregor> Stealing from OS X is what Microsoft does.
19:05:07 <alise> Gregor: Do you know how stable btrfs is?
19:05:17 <alise> WELL START KNOWING
19:05:22 <zzo38> alise: And I don't need such a thing that updates it dynamically based on those things, what I have works
19:06:50 -!- Sgeo|NeedUSB has quit (Ping timeout: 252 seconds).
19:08:15 -!- SgeosTears has joined.
19:10:57 <alise> SgeosTears: not right now, but later.
19:11:02 <fizzie> Wow, ugly: http://www.epstk.de/ http://epstk22.awardspace.com/examples/demo2.png
19:11:12 <alise> fizzie: latter is 404
19:11:27 <fizzie> Weird, maybe it checks referer or something.
19:11:32 <alise> yeah works on the page
19:11:55 <alise> we need something to fake referer instead of just hiding it (like anonym.to does)
19:26:37 <SgeosTears> Dammit, is it starting the download over?
19:27:12 * SgeosTears attempts to work out how to resume a download
19:30:09 <fizzie> http://p.zem.fi/shogun-octave -- a stable piece of software there.
19:34:00 <SgeosTears> Note to self: Don't assume /mnt/flash is, in fact, the flash drive
19:34:12 <SgeosTears> Esp. if i know that /mnt/sda1 is def. the flash drive
19:34:12 -!- wareya has quit (Read error: Connection reset by peer).
19:35:08 -!- wareya has joined.
19:40:28 <alise> SgeosTears: how old is fine structure? it was there in august 2005
19:41:38 <alise> http://web.archive.org/web/20080221012806/qntm.org/?structure
19:41:43 <alise> "This story has been ongoing since 2006."
19:41:45 <alise> http://web.archive.org/web/20080221003727/qntm.org/?fiction
19:41:52 <alise> Fine Structure (subdirectory)
19:42:13 <alise> Forgotten things in space. That's not in any more, is it?
19:42:15 * augur flops on alise sup you
19:42:24 <alise> That's why you said unbelievable scenes in space, SgeosTears?
19:42:37 <alise> Is reading "Forgotten things in space" a spoiler, then?
19:44:25 <SgeosTears> Forgotten things in space was retconned out
19:44:40 <SgeosTears> I think it is why I said Unbelievable Things in Space
19:44:58 <alise> "Then a man named Calrus discovered how to travel faster than light."
19:45:02 <alise> then that explains it
19:45:04 <alise> the comment on the later story
19:45:08 <alise> about "Calrus, the guy who invented FTL?"
19:45:57 -!- alise has quit (Remote host closed the connection).
19:47:38 -!- calamari_ has joined.
19:48:44 -!- calamari_ has changed nick to calamari-.
19:50:34 -!- lament has joined.
19:56:16 -!- Quadrescence has joined.
20:01:41 -!- nooga has quit (Ping timeout: 240 seconds).
20:01:47 -!- nooga has joined.
20:05:21 -!- lament has quit (Ping timeout: 252 seconds).
20:08:20 <SgeosTears> alise: "[REDACTED] is canon. Think of it as a deleted scene."
20:08:30 <SgeosTears> No, [REDACTED] is not the title of the actual thing
20:10:08 <ais523> hmm, apparently Myst completely confused its fans, when the developers made statements about official canon that contradicted the games
20:10:18 <ais523> people aren't entirely sure what to make of it
20:11:15 <SgeosTears> Also, I think that in Uru canon, the games were actually PC games based on real events
20:11:36 <fizzie> Vorpal: Here's a confusion matrix out of a default-parameters SVM classifier, on the "easy" set (everyone's all messages divided into 10 sets), 10-fold cross-validation (i.e. classifier trained on a set that has 9 points from everyone, then tested on a set that has one point from everyone, results averaged out). It's pretty boring,
20:12:27 <Vorpal> <olsner> staying up late to watch the counting of the votes live and listen to political analysis and stuff <-- so what is it called in English?
20:13:12 <Vorpal> fizzie, "here" you say, but where is the link?
20:13:27 <fizzie> Vorpal: Uh... I *thought* I had written it, but seems not.
20:13:43 <fizzie> Vorpal: http://zem.fi/~fis/esoconf.png anyway; "I mean, really, you should've been able to guess that".
20:14:21 <Vorpal> fizzie, so Gregor/ihope?!
20:15:08 <fizzie> There's two instances of ihope being misclassified as Gregor, right. And some cross-confusion between gregor/pikhq, which I think was in the original plot too.
20:15:17 <Vorpal> fizzie, what is the scale? percent of messages tested on?
20:15:24 <Gregor> That level of correlation is pretty high, actually.
20:15:38 <Gregor> I suppose you could identify me fairly accurately with /:P$/ though :P
20:15:45 <Vorpal> fizzie, well ratio, not percent
20:16:07 <Vorpal> Gregor, I do that sometimes too, and ";P"
20:16:37 <fizzie> Vorpal: It's normalized across the columns, so it's "fraction of correctly classified instances of this particular person".
20:17:03 <fizzie> Gregor: Right, but my features are borderline trivial: http://p.zem.fi/esomap-feats if you didn't see them yet.
20:17:09 <fizzie> Vorpal: That's why they call it a confusion matrix.
20:17:11 <nooga> in which C compilers x[y] means the same as y[x]
20:17:28 <fizzie> nooga: In all of them. Well, all compliant, anyway.
20:17:42 <nooga> since, i believe that it's all translated to *(x+y)
20:17:43 <fizzie> x[y] = *(x+y) = *(y+x) = y[x], after all.
20:17:48 <Gregor> fizzie: Yeah, I saw them.
20:18:00 <fizzie> Yes, 4["hello"] == 'o'.
20:18:52 <fizzie> It's a bit of an IOCCC'y trick to do, though.
20:19:16 <fizzie> I've seen exactly one instance of integer_variable["string literal"] in real code, I think.
20:19:30 <Vorpal> fizzie, I'm surprised you seen any at all
20:19:37 <Vorpal> fizzie, what was the context of that one
20:20:08 <fizzie> Vorpal: dcraw.c, tiff_get(), line 4308 (of some version): if (*len * ("11124811248488"[*type < 14 ? *type:0]-'0') > 4)
20:20:48 <Deewiant> That's not integer_variable["string literal"]
20:21:03 <fizzie> Deewiant: Oh, you're right! I misremembererered it.
20:21:25 <fizzie> Deewiant: It's still pretty messy, though. Maybe I could suggest swapping those around.
20:21:49 <fizzie> "It seems to me you're trying to make this more confusing: here, swap those around."
20:22:27 <Deewiant> Modulo naming the magic constants :-P
20:22:51 <Vorpal> also type → t and len → l
20:23:03 <fizzie> Deewiant: I'd write it with a "static unsigned char type_lengths[] = {1, 1, 1, ...}" and then type_lengths[...], instead of using ascii digits and subtracting '0'.
20:23:18 <Vorpal> SgeosTears, which version? for which platform? which type of install?
20:23:27 <Vorpal> SgeosTears, or do you mean the program? then md5sum
20:23:31 <Vorpal> also why "tears" in your nick
20:23:45 <Deewiant> fizzie: Well yes, that comes with the naming
20:23:47 <SgeosTears> Because Firefox on PuppyLinux keeps crashing
20:24:10 <Deewiant> If it's somehow really obvious I might write it as shown, but it probably isn't
20:25:25 <SgeosTears> is canon. Think of it as a deleted scene.
20:26:42 <fizzie> SgeosTears: 9a95ed6f6ec38fb58c446dba1add6a08 if I picked the right row.
20:27:10 <fizzie> SgeosTears: https://help.ubuntu.com/community/UbuntuHashes
20:27:46 <Vorpal> btw for anyone interested (updated every minute): http://www.val.se/val/val2010/valnatt/R/rike/index.html
20:28:59 <Vorpal> SgeosTears, firefox keeps crashing? how weir
20:29:26 <olsner> Vorpal: yep, been refreshing the page ever since they started
20:29:26 <Vorpal> or the distro as whole?
20:30:30 <fizzie> Vorpal: Are the light-grey bars last election's results?
20:31:09 <fizzie> Based on the fact that those column headings are also grey, I guess so.
20:34:12 <nooga> http://cobaia.net/2010/09/top-funny-source-code-comments/
20:34:27 <fizzie> Vorpal: "Valdeltagande .. 81,8%"; our last parliament elections (2007; there's one next year) had the lowest participation rate in Finland ever in the post-war era, 67,9%. Victory of the apathetic.
20:35:58 -!- alise has joined.
20:37:00 <Gregor> fizzie: In the US it hasn't beat 60% in 30 years.
20:38:30 -!- SgeosTears has quit (Ping timeout: 252 seconds).
20:38:59 <Vorpal> fizzie, that's preliminary
20:39:04 <alise> 12:12:55 <Vorpal> ah wake works, right
20:39:11 <alise> wake for is funerals only
20:39:18 <fizzie> http://zem.fi/~fis/esoconff.png -- heh, the fixed-size 2000-message test set is a pretty horrible sight. Almost everyone else (except ais523, me, oklopol and vorpal, and oerjan to some extent) keeps getting misclassified as alise. Probably because e represents over a third of the dataset.
20:39:43 <Vorpal> <olsner> aha, luckily wake also means "to remain awake for some purpose, duty, etc." and "to keep watch or vigil." ... either works well with an election
20:40:03 <alise> Vorpal: nobody says wake
20:40:06 <alise> people think funeral
20:40:11 <alise> *nobody says wake in any other context
20:40:23 <fizzie> alise: When you say wake, I think of that thing behind a boat.
20:40:40 <Vorpal> alise, now what would you call it?
20:40:44 <ais523> fizzie: analysing punctuation would probably work better
20:40:49 <Vorpal> I'll read the answer tomorrow. Night →
20:41:21 <fizzie> ais523: Probably. All this was originally done for book authorship classification, where the data points are quite a bit larger, so simple statistics like that have less variance.
20:42:03 <ais523> things like the punctuation mark used after a nickping would distinguish me from Vorpal straight off
20:42:49 <fizzie> That's not often seen in books, though. I'll try to make a more IRC-specific feature set at some point. There's some punctuation frequency counts, but they're just averages over the whole set.
20:43:16 <fizzie> (Still, the "colon frequency" would probably be higher for people that do "nick: foo" often.)
20:45:12 -!- distant_figure has joined.
20:45:20 -!- distant_figure has quit (Client Quit).
20:48:28 <fizzie> http://zem.fi/~fis/esomap-comp.png -- well, Vorpal does have the lowest colon frequency, and he does "nick," I guess.
20:53:40 <fizzie> (I never quite know how to react to greetings.)
20:54:42 <EOF> i have fallen in love with the arm architecture
20:55:02 <EOF> and the energy efficiency associated with it
20:56:02 <EOF> just image if intel applied their 22nm process to a desktop-class ARM cpu
20:56:23 <fizzie> Why, we could end world hunger in one swell foop!
20:57:12 * EOF instructs fizzle to duck as he fixes to explode
20:59:55 * EOF sees that fizzle ducked just in time and only lost four of his eight limbs
21:00:24 * EOF is going to stop with the /me stuff
21:00:32 * Gregor wonders who fizzle is.
21:00:55 * EOF takes that back
21:01:12 * EOF doesn't know who fizzle is either
21:01:41 * EOF also thinks that we should all speak in third person about ourselves
21:01:41 <fizzie> Gregor: Coincidentally, "fizzle" is what I was used to be called.
21:02:04 -!- Sgeo|Empathy has joined.
21:02:08 * EOF sees the error in his waingys/typ
21:02:25 <Sgeo|Empathy> Is Empathy as bad as most IM clients that try to do IRC are?
21:02:57 * EOF has never used empathy
21:15:31 -!- poiuy_qwert has joined.
21:16:40 -!- MigoMipo has joined.
21:18:56 <zzo38> End world hunger in the way they did at the radio, the entire universe is destroyed. And then they had news "the world no longer exists, I wanted to go to the zoo next week but I can't because neither the world nor me exist any more"
21:20:40 <ais523> zzo38: such a statement is a difficult one to make truthfully
21:21:02 <zzo38> ais523: Yes it is, but that was a joke (I think it was called "Irrelevance Show" or something like that)
21:23:25 <Sgeo|Empathy> Firefox just wastes SO MUCH screen space it isn't funny
21:24:38 <zzo38> Sgeo|Empathy: You can turn off some stuff, I think, in Firefox. Or use XUL-Runner, since Vonkeror does not use up much screen space for user interface stuff, only the document you are viewing takes up most of the space of the window.
21:26:56 <ais523> Sgeo|Empathy: if you press F11, then it doesn't use up any screen space but the document you're reading, the vertical scrollbar, and a few pixels at the top of the screen
21:32:42 -!- Phantom_Hoover has joined.
21:33:10 -!- augur has quit (Remote host closed the connection).
21:33:24 -!- augur has joined.
21:37:33 -!- poiuy_qwert has quit (Ping timeout: 272 seconds).
21:37:35 -!- Sgeo|Empathy has left (?).
21:37:43 -!- Sgeo|Empathy has joined.
21:37:44 -!- lament has joined.
21:37:46 <augur> sgeo is not empath- oh wait
21:39:28 <alise> augur: what was the end of that sentence? :P
21:40:21 <alise> <Phantom_Hoover> What are the haps, my friends? ;; this is a *so* un-you thing to say
21:40:29 <Sgeo|Empathy> I don't want to have to reinstall Chromium and Flash when I restart the comp
21:40:35 <alise> augur: i was making a pun on em-pathetic :P
21:40:39 <alise> anyway how do you know that?
21:40:44 <alise> unless you're making a diagnosis of autism
21:41:05 <augur> alise: his name was Sgeo|Empathy but then he left. we know he's still Sgeo, so he must no longer be empathetic!
21:41:18 <augur> http://s3.amazonaws.com/readers/2010/09/19/tapir3_1.jpg
21:41:22 <augur> have a david mitchell tapir
21:41:35 <alise> its nose is dripping. :|
21:41:38 <alise> brb, booting into a livecd
21:41:40 -!- alise has quit (Quit: Leaving).
21:46:12 <Sgeo|Empathy> I can't cheat Meebo into telling me what my passwords are :(
21:47:33 -!- alise has joined.
21:47:35 -!- poiuy_qwert has joined.
21:47:38 <augur> alise: david mitchell tapir has a cold.
21:48:47 <Phantom_Hoover> Do dependently-typed languages have to be totalistic, or can they allow general recursion?
21:49:14 <alise> "total", not "totalistic".
21:49:22 <alise> And the former; the latter implies an inconsistent type system.
21:49:31 <alise> I suppose if you don't mind type-checking taking forever and will never want to prove anything with it, it's okay.
21:50:00 <alise> (Literally forever.)
21:50:18 <Phantom_Hoover> IIRC there was some plan for Epigram to have limited general recursion, but I don't know the details.
21:51:37 <Phantom_Hoover> Well, there was a tagging system in the type system, I think.
21:52:23 <Sgeo|Empathy> What's the difference between gddrescue and dd_rescue?
21:53:06 <fizzie> g sounds like something with a GTK gui.
21:54:01 <fizzie> (Also, this is fizzie|Empathy speaking, and I don't like this especially much. The correct-as-I-type puts squiggles under my name.)
21:54:40 <fizzie> Then the difference is probably that gddrescue can check your emails in addition to reading a disk.
21:54:50 * Phantom_Hoover stopped using Empathy when he discovered it was making his real name visible in whoises.
21:55:09 <Phantom_Hoover> And I couldn't fix it easily, so I just switched to XChat.
21:55:12 <fizzie> Phantom_Hoover: You can change that in the "add account" preferences.
21:57:57 <Phantom_Hoover> Empathy sucks as an IRC client and as a chat client in general, really.
21:58:30 <fizzie> With this default theme, I'm finding it pretty hard to distinguish between who says what.
22:00:41 <alise> <Sgeo|Empathy> Also, what, exactly, is People Near Here?
22:02:55 <fizzie> I don't see a "People Near Here" anywhere. :/
22:03:07 <fizzie> Ooh, graphical smileys, sign of civilization.
22:04:06 <fizzie> There's a "People Nearby" account type here, I guess that's the same thing.
22:04:50 -!- aku2 has joined.
22:06:11 <fizzie> I don't exactly like the word "room" in an IRCy context either.
22:06:22 <fizzie> But the speech bubbles sure are purty.
22:06:36 <Gregor> "I thought this plan was foolproof!" "Foolproof yes, idiot-proof no."
22:06:51 * Sgeo|Empathy dislikes how it's impossible to distinguish between someone joining/parting and someone using /me
22:07:13 -!- augur has quit (Ping timeout: 255 seconds).
22:07:34 * Sgeo|Empathy has disconnected(Ping timeout: 10000000000 seconds)
22:07:44 * Gregor has disconnected (OR HAS HE)
22:07:55 <fizzie> Sgeo|Empathy: Yes, that's confusing, I had to take a look in the XChat window to find out what actually happened on your "join".
22:08:20 -!- aku2 has quit (Client Quit).
22:08:44 <Gregor> "No Bullwinkle, we've got to think!" "Rocky, this is no time to pick up a new hobby!"
22:09:03 <alise> <Sgeo|Empathy> How does someone not on Ubuntu connect to it?
22:09:41 <alise> other clients support it
22:09:49 <alise> it may just be a libpurple thing, but no, iChat has it too
22:09:59 <fizzie> Yeah, it seems to be a serverless Jabber thing.
22:10:46 <fizzie> That sort of goes hand-in-hand with the serverlessness; where would you send the password to.
22:11:21 <fizzie> Zeroconf is done with multicast-DNS-driven service discovery, I assume that's how it works too.
22:12:23 <Sgeo|Empathy> BRB, restarting, in the hopes that my installed software stays installed somehow
22:13:08 <Phantom_Hoover> http://pastebin.com/R6pkvHBF LW is beginning to disturb me.
22:15:57 <alise> Which part, exactly?
22:15:59 <Phantom_Hoover> Seriously, that response by EY would be more in place in a SF/horror story than any kind of rationalist discussion.
22:16:09 <alise> Also, you could just link.
22:16:46 -!- SgeoN1 has joined.
22:16:46 -!- Sgeo|Empathy has quit (Ping timeout: 245 seconds).
22:16:57 <alise> I think it's safe to assume EY is at least slightly autistic, and definitely ... abnormal, socially.
22:17:15 <SgeoN1> There's some kind of sick joke going on where sometimes, it refuses to see the USB stick
22:17:21 <alise> That's part of the problem. Even if you don't agree with EY.
22:17:34 <alise> A non-Friendly AI being created is the worst possible thing that could happen. Ever.
22:17:47 <alise> I don't think I agree with EY there. Although I see part of his point after filtering out the rage.
22:17:52 <Gregor> Fuck, I just made one.
22:18:04 <alise> But everyone agrees that a non-Friendly AI is the Worst Possible Thing.
22:18:39 <alise> Well, anyone who's actually thought about it -- you know, rationally.
22:18:41 <Gregor> Naw, a non-Friendly AI is just the next evolutionary step. At that point humans will be obsolete.
22:18:57 <fizzie> Hey now, that's not so bad: http://zem.fi/~fis/esoconfnf.png -- it's the same "features from 2000-message sets" confusion matrix, but with a different classifier than esoconff.png.
22:18:58 <alise> Gregor: A non-Friendly AI would not become the Borg.
22:19:06 <alise> Non-Friendly != evil; non-Friendly is worse.
22:19:09 <Gregor> alise: Hence the quote :P
22:19:30 <alise> A non-Friendly AI wouldn't actively decide to wipe us out, it would just see no problem with doing an action that happened to result in that, at all.
22:19:38 <SgeoN1> Yay, Chromium stayed installed!
22:19:59 <Gregor> The Borg don't seek to destroy a species that doesn't threaten them and isn't worth assimilating.
22:20:03 <alise> Gregor: If you said "I like paperclips" to a non-Friendly AI that listens to humans, it is *very possible* that it would just replace the entire universe with a paperclip pattern.
22:20:24 <alise> "I like paperclips" -> I'm trying to make them happy -> I like paperclips means paperclips are good ->
22:20:26 <alise> Make more paperclips
22:20:37 <alise> Best way to make more paperclips: Turn things into very small paperclips.
22:21:02 <Phantom_Hoover> alise, but surely it can reason enough to evaluate multiple factors?
22:21:14 <alise> Phantom_Hoover: Elaborate.
22:21:22 <Phantom_Hoover> i.e. it has a better view of the world than "paperclips".
22:21:51 <alise> I never said that's how it viewed the world.
22:21:55 <alise> Phantom_Hoover: have you read Creating Friendly AI?
22:22:09 <alise> I'm not asking you to sign up to the Cult of Yudkowsky or anything, I'm just saying you should.
22:22:26 <alise> Because it's pretty important that nobody defends or builds any AI which isn't some form of Friendly.
22:22:44 -!- Sgeo|Empathy has joined.
22:23:23 <Phantom_Hoover> I mean, a computer on a desk has... limited paperclipping capability.
22:23:58 <alise> Phantom_Hoover: And that's the thing.
22:24:05 <alise> We are presuming that the AI is self-improving, since that's the threat.
22:24:12 <alise> We are presuming that it surpasses human intelligence.
22:24:16 <Sgeo|Empathy> A sufficiently smart might be able to persuade humans to give it what it needs
22:24:28 <alise> How difficult would it be for a smarter-than-human AI to work up some money, and buy tech?
22:24:31 <alise> To steal the tech.
22:24:31 <Phantom_Hoover> So the crucial thing is that we make it friendly *before* making it self-improving?
22:24:35 <alise> Phantom_Hoover: Yes.
22:24:47 <alise> Because once it's smarter than us -- or even as smart as us, if it's cunning and logical -- we can't control it.
22:25:05 <Gregor> http://www.craigslist.org/about/best/phi/1755781713.html ... lawl.
22:25:25 <Phantom_Hoover> Surely nuking it while it's still vulnerable would work?
22:25:28 <alise> Phantom_Hoover: I'll also just mention that a lot of very smart people have thought about this very hard for a very long time. :)
22:25:35 <alise> Phantom_Hoover: Not if it's replicated itself across the globe.
22:25:52 <alise> Phantom_Hoover: And how do you know whether it's friendly or not?
22:26:00 <alise> You'll have no hope of understanding its data banks after it's mutated them itself.
22:26:08 <alise> And it'll be moving too fast for you to tell, anyway.
22:26:18 <Phantom_Hoover> Surely these are all arguments against making an AI at all?
22:26:41 <Phantom_Hoover> If you're this unsure, playing games with the whole of humanity is profoundly unethical.
22:26:44 <alise> Phantom_Hoover: A Friendly AI has no such problems. And, well, that's saying that you should stop progress before it's dangerous.
22:26:50 <alise> We're not unsure if you get it right the first time.
22:26:52 <alise> Only if you don't.
22:26:59 <alise> Inventing fire was dangerous.
22:27:06 <alise> Fire changed everything. What if it was for the worst?
22:27:09 <alise> Phantom_Hoover: Untrue.
22:27:18 <alise> I said that it's impossible to see the looming danger if it's self-improving.
22:27:18 <Phantom_Hoover> <alise> Phantom_Hoover: And how do you know whether it's friendly or not?
22:27:24 <alise> What I meant to say was:
22:27:30 <alise> I worded it badly.
22:27:38 <alise> Phantom_Hoover: Anyway, you know you need to nuke it if you have some code and it's not Friendly.
22:27:42 <alise> Before you even run it.
22:27:48 -!- MigoMipo_ has joined.
22:27:56 <alise> Friendly is an actual term, not just pointlessly-capitalised.
22:28:04 <Phantom_Hoover> So we need to have absolute, unshakable, complete proof of Friendliness before giving it power?
22:28:36 <Phantom_Hoover> Sgeo|Empathy, with self-improvement, they're the same.
22:28:37 -!- MigoMipo has quit (Ping timeout: 264 seconds).
22:28:42 <alise> <Phantom_Hoover> So we need to have absolute, unshakable, complete proof of Friendliness before giving it power?
22:28:57 <alise> Phantom_Hoover: Eliezer was the first to give Friendliness a rigorous treatment, though.
22:29:02 <alise> To actually define what an AI has to be to be safe.
22:29:13 <alise> The system of supergoals and subgoals, that's all his.
22:29:53 <alise> Phantom_Hoover: Anyway, yes, there is massive risk. Even if you're careful and prove it in so many ways. But if you get it right it's also the end of suffering, which is Kind of a Big Deal.
22:30:11 <alise> http://singinst.org/upload/CFAI.html
22:30:19 <alise> Is definitely a worthy read if you're at all interested in this stuff.
22:30:51 <Phantom_Hoover> I'm interested in it inasmuch as I want to stop my pathological worrying getting in the way of my life.
22:31:26 <alise> Phantom_Hoover: I went through that too. Don't worry.
22:31:56 <Phantom_Hoover> In any case, that LW thread still seems like a fairly irrational, fearful response.
22:31:58 -!- MigoMipo__ has joined.
22:32:14 <alise> Yes, well, this stuff is scary. I didn't read the thread due to the paste being badly-formatted.
22:32:44 <Phantom_Hoover> I mean, if you're that worried about nFAI, you should be doing everything to stop its slightest risk.
22:33:17 -!- MigoMipo_ has quit (Ping timeout: 272 seconds).
22:34:22 <alise> Phantom_Hoover: No. Seriously.
22:34:27 <alise> AI research is not bad.
22:34:44 <alise> AI or the death of civilisation seems to be more or less inevitable in the long term.
22:34:52 <alise> So what you have to do is promote Friendliness.
22:34:58 <alise> Not stop AI research altogether.
22:35:04 <alise> Because if civilisation survives -- and I think it probably will...
22:35:08 <alise> There's gonna be AI.
22:35:26 <alise> Nobody does that. I don't know that thread on LW. Possibly it was a mistake.
22:35:41 <alise> Although anyone outright advocating non-Friendly AI on Less Wrong is almost certain to get mauled.
22:35:46 <Phantom_Hoover> That thread seems to have been deleted from LW itself.
22:35:59 <quintopia> GLaDOS was friendly before someone made her schizophrenic by changing important components
22:36:10 <alise> It's like saying "Why don't we try nuclear war?" as if everyone had nuclear weaponry in their house and it just required a huge amount of intelligence to fire them.
22:36:19 <alise> Phantom_Hoover: Risk to humanity -- yes, inflated, I agree.
22:36:21 <alise> I wouldn't have done it.
22:36:25 <alise> But I see where he's coming from, you know?
22:37:19 <Phantom_Hoover> alise, deleting discussions *on the risks* on the grounds of risk is not going to help avoid the risks, IMO.
22:37:27 <alise> Actually most AI proponents think that developing nanotech is a very bad idea because of the immense destructive possibilities, and it's best to let a benevolent AI take care of developing it.
22:37:29 <alise> Phantom_Hoover: Agreed.
22:37:34 <alise> You are probably right.
22:37:39 <alise> I am just defending the reasoning behind it, not the action.
22:37:51 <alise> I am not convinced that Yudkowsky actually sleeps at night.
22:38:01 -!- MigoMipo__ has quit (Ping timeout: 265 seconds).
22:38:03 <quintopia> what's the name of the dude that offers money to people if he, prentending to be an AI trapped on a disconnected computer, can't convince them to let him out in just a couple hours?
22:38:09 <alise> quintopia: Yudkowsky.
22:38:11 <Phantom_Hoover> And if he thinks an nFAI is that petty and malicious, surely the people attempting to stop its creation would be the first on its Infinite Torture List?
22:38:15 <alise> He doesn't do it any more, I don't think.
22:38:26 <alise> Phantom_Hoover: Stop what's creation?
22:38:30 <quintopia> yeah he would eventually fail, and lose a lot of money
22:38:38 <alise> Phantom_Hoover: But nFAI is a bad thing...
22:38:46 <alise> People trying to stop nFAI being created are his friends... :P
22:38:51 <alise> quintopia: Yeah, all $20.
22:38:57 <alise> quintopia: It's a thought experiment, anyway.
22:39:04 <alise> In the real situation the thing on the other end is many, many times smarter than you.
22:39:07 <alise> So it's guaranteed to win.
22:39:08 -!- Sgeo|Empathy has quit (Quit: Sgeo|Empathy).
22:39:08 <quintopia> could have sworn he offered more than $20
22:39:17 -!- Sgeo|Empathy has joined.
22:39:21 <alise> Phantom_Hoover: Anyway, I'm going to reboot now.
22:39:22 <Phantom_Hoover> But he deleted that post because he thought that if an nFAI was created, it could rain down suffering on people it didn't like.
22:39:31 <alise> NOOOOOOOOOOOOOOOOOOOOOOO
22:39:32 -!- alise has quit (Quit: Leaving).
22:40:03 <quintopia> don't listen to any of the other voices
22:40:21 <quintopia> if you lose your train of thought, you may never get it back
22:41:05 <zzo38> Don't lose your thought of train... Don't lose your thought of train... Don't lose your thought of train... Don't lose your thought of train... Don't lose more lot of rain...
22:41:16 <quintopia> keep it on its track and don't get derailed
22:42:14 <quintopia> tomorre do that this and the same?
22:42:40 <quintopia> gah. dropped packets like no one's biz.
22:42:42 <zzo38> Today is the tomorrow you worried about yesterday.
22:43:21 -!- augur has joined.
22:45:48 <Phantom_Hoover> Extrapolating Moore's law infinitely as evidence for a singularity seems... dodgy.
22:46:19 <Phantom_Hoover> Given that we can't go much further down without bashing into fundamental constraints.
22:46:37 <quintopia> especially with the hard limits that have already been identified on the various Moore's Laws
22:47:25 <Phantom_Hoover> And the fact that the good old laws of thermodynamics have *already* started to cap processor clock speeds.
22:48:01 -!- Sgeo|Empathy has quit (Ping timeout: 245 seconds).
22:48:28 <Phantom_Hoover> I mean, we can probably push quite a bit further, but not indefinitely.
22:48:28 <quintopia> there's also a hard limit imposed on transistor size, although if quantum scale gates become feasible and cost-effective, that'll go away
22:48:35 <quintopia> but even then, we have a size limit
22:49:26 <quintopia> adiabatic computing could go a long way toward stepping past current heat-related limitations
22:49:37 <Phantom_Hoover> And there's also the jetpack effect when extrapolating into the future.
22:49:37 <quintopia> it's really hard to imagine where we'll get to though
22:49:52 <quintopia> iiiii think we just said the same thing
22:50:37 <Phantom_Hoover> We'll be right with some things, but wrong with many, many more.
22:52:08 <Phantom_Hoover> But it's still worth consideration, since the costs and gains are so huge as to be impossible to ignore.
22:52:24 <quintopia> oh right and happy tlapd and stuff. shouldn't have forgotted that.
22:54:00 <quintopia> is there a variation on asimov's laws that doesn't drive AIs insane?
22:54:10 <Phantom_Hoover> This be far more disconcertin' than any of this FAI stuff!
22:55:01 <quintopia> you can be a pirate and still talk about bots
22:56:37 <Phantom_Hoover> We almost certainly won't be havin' nanoscale construction machines.
22:57:06 -!- SgeoN1 has quit (Ping timeout: 240 seconds).
22:57:23 <coppro> I be thinkin ye wrong about tha
22:57:31 <coppro> and it bein a matter of school pride
22:58:21 <Phantom_Hoover> I mean, macroscale von Neumann probes be plausible, but packin' the complexity necessary for such a feat into nigh-molecular scales is so implausible with current technology that it be meanin'less to speculate.
22:58:42 -!- SgeoN1 has joined.
22:59:41 <SgeoN1> Alise, better to wait for some more time until I get the external mount, or as soon as possible, despite not having the external mount
22:59:48 <Phantom_Hoover> I be havin' a horrible feelin' that alise will start arguin' with me over this, and that not be particularly pleasant for those in the same time zone who be havin' conserrrvative sleep schedules.
23:00:36 <coppro> Phantom_Hoover: why be ye callin' the lassy a he?
23:00:52 <SgeoN1> PH, what's your suggestion?
23:01:12 <Phantom_Hoover> coppro, I not be bothered to keep track of everyone's pronoun requests.
23:01:13 <Gregor> "Come now, would I sell my own granny for $1,000?" "I don't know, would you?" "Of course I would!"
23:01:23 <quintopia> coppro: why you callin' the laddie a lassie?
23:01:39 <Phantom_Hoover> SgeoN1, have ye got the hard drive out of tha computarrr?
23:01:48 <coppro> quintopia: arrrr, ye don't know the legend of alise, do yer?
23:02:15 <SgeoN1> It's been out for a while
23:02:33 <coppro> quintopia: it be best told by the girl herself
23:02:36 <Phantom_Hoover> quintopia, it be complex, and best be told by the 'man 'imself.
23:03:13 -!- oerjan has joined.
23:03:18 -!- lament has quit (Read error: Operation timed out).
23:03:26 <coppro> Phantom_Hoover: clearly yer didn't be learnin' the morrral o' the tale
23:04:21 <coppro> oerjan: and an 'earrrrty 'ello to you too!
23:05:14 <SgeoN1> Is the HD actually accumulating damage right now?
23:06:50 -!- tombom has quit (Quit: Leaving).
23:08:25 <zzo38> I added in PicoC into Enhanced CWEB now, so that you can define codes to run at compile time (for CTANGLE only), and you can make a function lossy$() that is executed instead of typing the identifier "lossy" into the output. And, it works.
23:08:42 <oerjan> <quintopia> yeah he would eventually fail, and lose a lot of money
23:08:51 <oerjan> i recall he already failed some times
23:09:13 <SgeoN1> Also, I have no plans on letting the GUI start if I have the HD just be inside the comp
23:09:14 <oerjan> he just hadn't put that on the main page for the challenge
23:09:25 <oerjan> but it was in some comment he made
23:09:52 <oerjan> i don't recall where it was
23:10:38 <oerjan> ah it's that day again. actually it already ended here.
23:11:35 <SgeoN1> Parted Magic shouldn't take that long to boot, really.
23:11:37 <fizzie> 'ts been over for over an hour here already; 'ts +3 here.
23:12:44 <quintopia> in my book, it lasts 2 days, from the moment it begins at the international date line, to the moment it ends on the other side of the same
23:13:00 <quintopia> my logic bein' that a true pirate should no' claim a home port!
23:13:59 <fizzie> Pirates be so arrrrrrogant folk, all about making everyone talk like one.
23:14:20 <quintopia> i actually would rather be a pirates of penzance type pirate
23:14:28 <quintopia> so i can talk like a lord gone astray
23:15:32 <pikhq> quintopia: I suggest ye learn a proper West Country accent!
23:17:15 <pikhq> quintopia: No, West Country is rhotic. And the stereotypical pirate accent.
23:17:27 <pikhq> quintopia: So. "Prroperrr"
23:18:38 <pikhq> It was also used in Pirate of Penzance.
23:19:31 <quintopia> i want to be a dainty pirate captain
23:22:26 <Phantom_Hoover> I also be less surre on whetharr civilisation be survivin' than alise.
23:22:54 <Phantom_Hoover> By this point, a single fall would be the end for a very very very long time. Yarr.
23:23:17 <Phantom_Hoover> Since we be usin' up all the easily accessible resources.
23:25:23 <oerjan> Phantom_Hoover: found a link to a discussion (with more links): http://www.reddit.com/r/science/comments/6k95z/a_transhuman_ai_would_just_convince_you_to_let_it/c04360q
23:28:17 <pikhq> Phantom_Hoover: On the other hand, we are leaving *around* a lot of accessible resources.
23:28:32 <oerjan> google conspires to make the mentions of failure rather hard to find if you just search for "yudkowksy aibox" :(
23:28:46 <pikhq> If civilization fell, the next civilization would have a number of things easy.
23:28:54 <oerjan> (especially if you don't _know_ about the failures)
23:28:58 <pikhq> For instance: metal would not need to be mined.
23:29:29 <Phantom_Hoover> Yes, but things like coal and oil, which fuelled our technological development, will be much harder to get.
23:30:33 <Phantom_Hoover> And a pile of rust, which is really all that would remain of exposed structures after a surprisingly short time, is not very useful.
23:31:22 <pikhq> Coal and oil were only really primary fuels for the past, oh, hundred years.
23:31:37 -!- augur has quit (Remote host closed the connection).
23:32:03 <pikhq> And even concerning coal: for many purposes, it's merely *more convenient* than the alternative.
23:32:23 <Phantom_Hoover> pikhq, which is also the major period of technological advance.
23:32:44 <oerjan> Phantom_Hoover: perhaps it will take a longer _time_ than the industrial revolution because of lack of coal and oil. but there _are_ still environmentally clean energy sources, such as windmills.
23:32:46 <Phantom_Hoover> We went from abacuses to quantum computers in around that timeframe.
23:33:06 <pikhq> Phantom_Hoover: Yes, but it still is plausible that an alternative energy source would be found.
23:33:07 <Phantom_Hoover> oerjan, can you imagine the Industrial Revolution happening with windmills?
23:33:15 -!- augur has joined.
23:33:35 <oerjan> but is going slower even a bad thing? the breakneck speed might be why we got into this much trouble in the first place
23:33:37 <pikhq> Phantom_Hoover: Dude, "past 100 years" *post-dates* the Industrial Revolution.
23:34:29 <Phantom_Hoover> But anyway, it's not at all certain that civilisation will survive on Earth.
23:34:33 <pikhq> The Industrial Revolution could have *easily* happened with wood instead of coal. And a lot of it was actually powered by windmills and water wheels anyways.
23:35:32 <pikhq> So, even assuming no alternative fuel sources could be found (not *that* likely), we'd still get up to Industrial Revolution level tech with mostly renewable/recyclable resources still left.
23:36:15 <pikhq> And I strongly suspect that one could build up to nuclear power from there.
23:37:25 <Phantom_Hoover> I mean, what are you trying to prove by not revealing the nature of the arguments?
23:38:03 <quintopia> if he were still doing it, i'd take the challenge and then reveal the arguments.
23:38:19 <oerjan> i thought it was to prevent "I would never have fallen for that" arguments, i think he said as much
23:38:22 <quintopia> i could een do it without lying, by making someone else doing the promising not to part
23:39:12 <oerjan> "I would never have fallen for that" _is_ a major way of human self-delusion, after all
23:40:42 * oerjan is _much_ more infuriated by the _apparent_ hiding of the failed experiments, in case you didn't guess
23:41:26 <oerjan> although i _suppose_ that could be just an artifact of google's pagerank
23:41:41 <Phantom_Hoover> I also think the gatekeeper is incapable of showing anything meaningful.
23:41:55 -!- Sgeo|Empathy has joined.
23:41:57 <EOF> Phantom_Hoover:
23:41:59 <EOF> Phantom_Hoover:
23:42:19 <EOF> quintopia:
23:42:22 <EOF> quintopia:
23:42:49 * EOF feels like shooting himself
23:42:54 -!- poiuy_qwert has quit (Quit: Leaving).
23:42:57 <oerjan> Phantom_Hoover: but i _think_ yudkowsky's basic belief is that an AI would be so resourceful that saying you "would never have fallen for" a particular argument _is_ a dangerous delusion
23:43:18 <oerjan> because the AI could have used a plethora of other methods
23:43:23 <Phantom_Hoover> Yes, but this experiment has little to no bearing on any real situation.
23:43:46 <Phantom_Hoover> The point might stand; it might not. This experiment is completely useless.
23:43:55 <quintopia> thus we should make sure to have a friendly AI before a malevolent AI gets wise enough to convince its gatekeeper to let it out
23:44:02 <quintopia> then we could have a titan clash of the AIs
23:44:14 -!- augur has quit (Ping timeout: 265 seconds).
23:44:35 -!- SgeoN2 has joined.
23:45:37 <oerjan> quintopia: a friendly AI would simply prevent an evil one from ever being built, surely
23:45:43 <Sgeo|Empathy> Can I start Ubuntu at the commandline, then from there start x?
23:45:48 <Phantom_Hoover> And of course, *without knowing the effing discourse between the participants, the experiment is utterly stupid*
23:45:52 <quintopia> also we need a time machine so we can preemptiveill any scientists that eventually turn out to work on creating malevolent AI
23:46:23 <pikhq> Phantom_Hoover: It's like publishing an experiment without publishing the data.
23:46:29 <pikhq> In fact, it *is precisely that*!
23:46:57 <pikhq> About as valid as "I have an elegant proof of this that the margin is too small to contain."
23:47:00 <Phantom_Hoover> The kind of thing Ben Goldacre spends ages picking apart, in fact.
23:47:39 <quintopia> pikhq: at least in this case the experimental result has been verified by several individuals, so it is somewhat more trustworthy than fermat's proof
23:48:18 -!- SgeoN1 has quit (Ping timeout: 240 seconds).
23:48:42 <quintopia> indeed, i would argue that in this case, it is like publishing the data without publishing the experiment
23:48:51 <oerjan> Phantom_Hoover: of course yudkowsky and his ilk _are_ fanatically afraid of evil ais, psychologically so even if they are correct
23:48:57 <quintopia> we know how many people fell for it, but not what the actual methodology was
23:49:13 <pikhq> quintopia: Still absolutely unvalid, though.
23:49:47 <quintopia> pikhq: irreproducible at least, but we can imagine the techniques used at least
23:49:49 <oerjan> so you could suspect him of being disingenious on purpose
23:50:01 <pikhq> quintopia: Irreproducible == INVALID
23:50:24 <quintopia> pikhq: the results are NOT invalid, they are just not confirmable. there is a subtle difference
23:50:46 <pikhq> If you can't reproduce it it's invalid.
23:51:16 <quintopia> they can reproduce it. it's just not independently reproducible
23:51:51 <pikhq> Which makes it invalid.
23:51:54 <oerjan> Sgeo|Empathy: his last comment was probably a humoristic way of saying he went to bed
23:52:01 <quintopia> i really don't like your definition
23:52:09 <pikhq> Then you really don't like science.
23:52:30 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
23:52:35 <Sgeo|Empathy> But I can just go into the console when X is starting, I guess
23:52:39 <oerjan> Sgeo|Empathy: hey i did the same thing :D
23:53:07 <Sgeo|Empathy> I guess, before doing anything else, I should figure out the exact commands I'll be using
23:53:25 <quintopia> you can't invalidate a result by not explaining what the result was. it's still a result. it just isn't a widely understood one.
23:53:36 <pikhq> quintopia: Their claims are as scientifically legitimate as publishing data purpoting that saying magic words (they won't say which) summons God 10 times out of 10.
23:54:13 <quintopia> in other words, possibly completely legitimate and valid, but we'll never know if that's the case
23:54:28 <pikhq> Precisely. And that *makes it invalid*.
23:54:42 <Sgeo|Empathy> There's no reason to not let X load while I do the ddrescue stuff, right?
23:54:46 <quintopia> and not knowing whether god exists makes god not exist?
23:55:14 <pikhq> It makes it invalid *as an experiment*.
23:55:35 <pikhq> An experiment can readily have correct results while still being invalid.
23:55:48 <pikhq> Much like a proof can have correct results while still being invalid.
23:55:57 <quintopia> this confuses me greatly. how is it that an invalid experiment could *suddenly become valid* just as soon as its results are published?
23:56:20 <quintopia> i could see an unconfirmed but valid experiment being confirmed valid on this occasion
23:56:24 <zzo38> quintopia: It depend what you meant by "agnostics" or "atheists", but I consider it different.
23:56:35 <quintopia> as well as an unconfirmed invalid experiment being confirmed invalid
23:56:56 <zzo38> I also consider different that the different athesists and different agnostics even have different opinions of their own atheism/agnosticism.
23:57:00 <quintopia> zzo38: can i get on your speech analysis plot?
23:57:07 <zzo38> (I myself, am agnostic, if it cares)
23:57:28 <zzo38> quintopia: I have no speech analysis plot. If I make one though, everyone relevant will be on it (including you).
23:57:37 <pikhq> quintopia: In order to be valid as an experiment, it *must be testable*. This is about as essential to science as the idea that reality can be observed.
23:58:09 <quintopia> pikhq: this is a tewstable hypothesis, and it has been tested. independently verifiable != testable
23:58:11 <Sgeo|Empathy> Should I set "Spin down hard disks when possible"?
23:58:11 <pikhq> Without being testable, an experiment might as well be the rantings of a drunk hobo.
23:58:18 <zzo38> pikhq: You are correct about experiments. If the experiment does nothing, how can you do anything relevant with it?
23:58:27 <quintopia> zzo38: who was it who posted the speech analysis plots :/
23:58:47 <zzo38> quintopia: Look it up in the logs, I don't remember either
23:58:57 <pikhq> quintopia: Independently verifiable is a necessary but not sufficient condition of being testable.
23:59:24 <quintopia> pikhq: indeed, if an independent group can replicate their results using their own slightly-different methods, would the experiments together still be collectively invalid to you?
23:59:36 <pikhq> If they don't publish their methods?
23:59:46 <pikhq> As far as I'm concerned, it's a second group of drunk hobos.