00:00:04 <Vorpal> elliott, asteroids. Don't think I actually played that
00:00:16 <Vorpal> elliott, isn't it like Ambrosia's Maelstrom?
00:00:19 <elliott> Vorpal: It's that thing where you're a triangle and there are hollow rocks around you.
00:00:26 <elliott> http://upload.wikimedia.org/wikipedia/en/1/13/Asteroi1.png
00:00:32 <elliott> And you spin around shooting them.
00:00:53 <Vorpal> elliott, do large ones break into smaller when hit?
00:01:06 <Vorpal> elliott, if so it is exactly like maelstrom, except with worse graphics
00:01:19 <elliott> To be fair, it almost certainly predated Maelstrom by a long, long time :P
00:01:31 <Vorpal> elliott, you played the latter?
00:01:38 <elliott> Maelstrom is a 1992[1] clone of Asteroids with an improved graphics and interface.[2][3] Many of Ambrosia's subsequent shareware titles followed in a similar formula.
00:01:50 <Vorpal> elliott, http://en.wikipedia.org/wiki/File:Maelstrom_screenshot.png
00:01:56 <elliott> I have used exactly two Macs in my life for more than a few seconds, I think.
00:02:10 <elliott> One was someone else's eMac I used once, the other is my iMac.
00:02:16 <elliott> Although to be fair it's rapidly approaching three.
00:02:19 <Vorpal> emac... what one was that now again?
00:02:31 <elliott> Vorpal: you know the original iMac?
00:02:57 <Vorpal> elliott, I do know the original imac. And it came in a snow white edition later. And flower power
00:03:29 <elliott> http://www.extensions.in.th/post/emac/emac2.jpg
00:03:34 <elliott> It was short for "education Mac".
00:03:43 <elliott> The eMac, short for education Mac, was a Macintosh desktop computer made by Apple Inc. It was originally aimed at the education market, then available as a cheaper mass market option over Apple's second generation iMac. The eMac design closely resembled first-generation iMacs. It sports a PowerPC G4 processor significantly faster than the older iMac's G3 processor, and a larger 17" flat display.
00:03:47 -!- Sgeo_ has joined.
00:03:47 -!- cheater99 has joined.
00:04:20 <elliott> Although it did last for several years under OS X, yes
00:04:45 <elliott> Vorpal: Anyway we've ruled out MMO because MMO physics, synchronisation and servers sounds like FUN (read: pain).
00:04:56 <elliott> Vorpal: So it's either going to be a deathmatch game or [insert other thing here].
00:05:12 <Vorpal> elliott, I know a guy good at MMO
00:05:21 <Vorpal> elliott, I think he went by the name notch
00:05:29 <elliott> Vorpal: You have a strange definition of good.
00:05:39 <elliott> Vorpal: Also of MMO, Minecraft servers rarely have /that/ many people on them :P
00:05:54 <Vorpal> elliott, indeed. AND IT IS ALREADY BUGGY!
00:05:54 -!- cheater99 has quit (Read error: Connection reset by peer).
00:06:11 <elliott> Vorpal: I feel kinda bad doing copyleft ...
00:06:23 <Sgeo_> BZFlag servers have lots of people on them!
00:06:23 <Vorpal> elliott, alt: it is MMO for minecraft
00:06:38 -!- Sgeo has quit (Ping timeout: 240 seconds).
00:06:41 <Sgeo_> (and lots of cheating, but I digress)
00:06:48 <Vorpal> elliott, and copyleft rocks!
00:06:53 <Vorpal> elliott, you are seeing the light
00:07:01 <elliott> Vorpal: (As in, the game will be released under "sorta-copyleft": it's copyleft, except you can't distribute anything more than patches in public, basically.)
00:07:10 <elliott> Because otherwise people could just post the entire source :P
00:07:28 <Sgeo_> elliott, you're turning evil?
00:07:43 <elliott> Sgeo_: Yes; mwahahaha, look at my fangs.
00:07:59 <elliott> So evil that I'm considering putting it on a torrent site as a potential solution for slow sales.
00:08:08 <elliott> (Note: Phantom_Hoover__ is not nearly as insane as me.)
00:08:30 <Phantom_Hoover__> elliott, of the two of us, who thought of the negative mass engine?
00:08:35 <Vorpal> elliott, hm "On July 5, 2006, an "educational configuration" of the iMac Core Duo was introduced, discontinuing and replacing the entire eMac line. The new iMac has a Combo drive rather than a SuperDrive and a smaller hard disk of 80 GB."
00:08:40 <Vorpal> elliott, this sound feeble
00:08:54 <Vorpal> elliott, I mean, for education you want power machines
00:09:00 <elliott> Vorpal: Are you being serious.
00:09:01 <elliott> Phantom_Hoover__: Well, okay, you.
00:09:17 <elliott> Phantom_Hoover__: But I'm crazy enough to want to release the whole thing as free-as-in-beer and open source after sales trickle off!
00:09:43 <elliott> TAKE THAT, CAPITALISM! I AM GOING TO BE AS NONCONFORMIST AND HIPPIE AS I CAN WHILE STILL RELYING ON THE SALE OF DATA TO MAKE A LIVING!
00:09:51 <Vorpal> elliott, yes quite. We all have Core 2 Quad or better in most labs at university. (so there are a few old ones with p4 and CRT iirc, but they are an exception)
00:10:12 <Vorpal> (and they are only there because other computers are non-legacy)
00:10:25 <Vorpal> (and we need parport programmers for some stuff)
00:10:42 <Vorpal> elliott, yeah but that is software. There is no excuse
00:10:48 <Phantom_Hoover__> elliott, FUN FACT: as far as I can tell, the Edinburgh Council computers all have BT branding as their screensavers.
00:11:02 <Phantom_Hoover__> And they refer to their "customers", which annoys me intensely.
00:11:13 <elliott> Heh, customers of the government.
00:11:25 <elliott> Isn't it more like the government is *our* customer?
00:11:36 <elliott> Well, really, the government is our contractor.
00:11:40 -!- infanticide has joined.
00:11:42 <elliott> We just seem to forget that every now and then.
00:11:56 <Phantom_Hoover__> It's more like "commercial metaphors can only lead to naïvety".
00:12:29 <elliott> Phantom_Hoover__: But the government are definitely supposed to work for us. :p
00:12:34 <Vorpal> elliott, anyway, why athena?
00:12:41 <Vorpal> elliott, I mean.... there is no excuse to use that
00:12:51 <elliott> Vorpal: Because they invented it, and they have a LOT of machines with it.
00:12:58 <elliott> Vorpal: Hey, it runs gnome now. :p
00:13:03 <elliott> (And is based on Ubuntu, now, I think)
00:13:05 <Sgeo_> "og:url - The canonical, permanent URL of the page representing the entity. When you use Open Graph tags, the Like button posts a link to the og:url instead of the URL in the Like button code."
00:13:22 <Sgeo_> That doesn't sound the least bit abusable
00:13:30 <Sgeo_> http://developers.facebook.com/docs/reference/plugins/like
00:13:32 <Vorpal> elliott, wait, a second, isn't athena just a GUI toolkit?
00:13:57 <elliott> Vorpal: Athena is just what some people call Xaw, because Project Athena at MIT (involving DEC and IBM) created it
00:14:11 <elliott> Vorpal: Athena created X11, Kerberos and Zephyr.
00:14:27 -!- Zuu has joined.
00:14:29 -!- Zuu has quit (Client Quit).
00:14:48 <elliott> Vorpal: tl;dr: if MIT wants to have a bunch of semi-thin clients, they buy a bunch of expensive workstations, put Unix on them, invent X11 and Kerberos, connect them to a server, and call them thin clients.
00:14:53 <elliott> "Students would have access to (for the time) high performance graphical workstations, capable of 1 million instructions per second and having 1 megabyte of RAM and a 1 megapixel display." <-- THIN CLIENT
00:15:03 -!- Zuu has joined.
00:15:59 <Vorpal> elliott, well, if you need to do advanced computation then a lot of local resources might be good
00:16:07 <elliott> Vorpal: They're used to run Emacs :P
00:16:44 <elliott> eight megabytes and constantly swapping, not one!
00:16:49 <elliott> (okay, so they've been upgraded since then)
00:16:54 <elliott> Vorpal: http://blog.spang.cc/images/clean-athena.png Athena as of a few years ago.
00:17:03 <elliott> Vorpal: I understand it has been upgraded to be Ubuntu-based since then and probably the visuals too.
00:17:06 <pikhq> elliott: Well, when they started getting high-quality workstations was much more expensive than getting high-quality X terminals and UNIX systems.
00:17:11 <elliott> Vorpal: (That's GNOME 2.8. *In 2008.*)
00:17:25 <elliott> pikhq: X terminals didn't exist because X11 didn't exist.
00:17:31 <Vorpal> <elliott> Vorpal: http://blog.spang.cc/images/clean-athena.png Athena as of a few years ago. <-- still decent
00:17:35 <elliott> pikhq: They bought high-end workstations: 1 MIPS, 1 MB RAM, 1 megapixel display.
00:17:36 <Vorpal> compared to the original
00:17:44 <pikhq> elliott: Dear God.
00:17:46 <Vorpal> elliott, what version is gnome now. 2.16?
00:17:47 <pikhq> elliott: The overkill.
00:17:48 <elliott> pikhq: They /invented X/ to run on these.
00:18:05 <elliott> Vorpal: 2.8 came out in 2004.
00:18:06 <pikhq> elliott: Okay, Athena would have made sense if X terminals were easily available.
00:18:11 <pikhq> elliott: But holy God that's nuts.
00:18:23 <Vorpal> pikhq, awesome for the students there
00:18:26 <pikhq> elliott: Okay, yes, I was actually looking for "crazy awesome".
00:18:28 <elliott> pikhq: Anyway, they all have their own copy of GNOME I think.
00:18:33 <elliott> Just /some/ apps are on a server.
00:18:40 <elliott> I think Emacs is, for instance.
00:18:44 <Vorpal> elliott, except those at the AI lab. They had lisp machines
00:18:57 <Vorpal> (at some point at least)
00:19:03 <elliott> Vorpal: I'm pretty sure anyone could go down to the AI Lab :P
00:19:05 <pikhq> Vorpal: Lisp machines were falling out of vogue by the time Athena came into play.
00:19:25 <pikhq> Though MIT almost certainly still had plenty at the AI lab.
00:19:41 <elliott> But yeah, it didn't take long for them to decline.
00:19:57 <elliott> Were they Symbolics, though? I doubt it.
00:20:02 <elliott> Since Symbolics grew out of the AI Lab.
00:20:20 <pikhq> Probably the AI Lab-built machines.
00:21:02 -!- Phantom_Hoover__ has quit (Remote host closed the connection).
00:21:50 <pikhq> It amazes me that X actually has support code for widget libraries.
00:22:01 -!- cheater99 has joined.
00:22:08 <pikhq> Even though it was last used over a decade ago.
00:24:48 <pikhq> They called out to Xt.
00:25:06 -!- cheater99 has quit (Read error: Connection reset by peer).
00:25:28 <elliott> pikhq: People still use Xaw. And Motif.
00:25:40 <Sgeo_> If Vorpal moves left, will time move backwards back into day?
00:28:06 <elliott> pikhq: I NEED LANGUAGE SUGGESTIONS
00:29:28 <Sgeo_> elliott, Falcon Factor Haskell C C++ Java Newspeak Smalltalk Ur Python BancSTAR Brainfuck
00:29:34 <Sgeo_> ^^in no particular order
00:29:37 <elliott> Fail at interpreting my statement.
00:30:59 <elliott> pikhq: What would JESUS put in a language???
00:31:45 <elliott> elliott@dinky:~$ clang -fblocks bar.c -o bar
00:31:45 <elliott> /tmp/cc-FT8NmW.o:(.rodata+0x20): undefined reference to `_NSConcreteGlobalBlock'
00:32:40 <elliott> coppro: You never told me blocks don't work as regular function pointers :|
00:32:50 <pikhq> elliott: Jesus would put salvation into a language, and everyone else would turn it into C++.
00:32:52 <Sgeo_> elliott, something that prevents debuggers and statements used in a debugging context from working properly. You need faith that the program will work properly.
00:34:36 <elliott> "Clang doesn't yet provide an easy way to use blocks on platforms that don't have built-in operating system support (e.g., SnowLeopard)."
00:34:40 <elliott> pikhq: Can I see your copying code?
00:35:03 <pikhq> elliott: That line is bullshit.
00:35:13 <pikhq> elliott: http://compiler-rt.llvm.org/ You just need that library installed.
00:35:20 <elliott> pikhq: Right. Apparently it's not so mature though.
00:35:27 <elliott> (They linked to that in the same comment.)
00:36:13 <elliott> pikhq: But srsly, you can't use blocks as function pointers.
00:36:35 <pikhq> http://sprunge.us/eFJC There's a self-contained example.
00:36:53 <elliott> <elliott> pikhq: But srsly, you can't use blocks as function pointers.
00:37:26 <elliott> <pikhq> http://sprunge.us/eFJC There's a self-contained example.
00:37:31 <elliott> Any way to avoid being explicit with the closure?
00:37:38 <elliott> That's kind of a pain as a compiler-writer.
00:38:05 <elliott> What's the name of the compiler-rt lib?
00:38:20 <elliott> There's librt but it's not that.
00:40:31 <elliott> pikhq: Are you suuure I can't avoid explicit closing?
00:40:59 -!- cheater99 has joined.
00:41:31 <pikhq> elliott: If you can avoid ever popping the function off the stack, you can use GCC's automatic closing.
00:41:40 -!- zzo38 has joined.
00:41:56 <elliott> pikhq: Hmm. Well, can I just list all the variables mentioned in the function body and that would be sufficient?
00:42:12 <elliott> pikhq: Also: How do you avoid invoking gcc's automatic closing in that code?
00:42:17 <zzo38> Should C have a allocprintf command?
00:42:43 <zzo38> elliott: What do *you* think?
00:43:00 <pikhq> elliott: I do not ever use any variable from the enclosing scope.
00:43:07 <Sgeo_> zzo38, how difficult would it be to port MegaZeux to Flash?
00:43:12 <zzo38> elliott: You cannot guess what I meant?
00:43:29 -!- cheater99 has quit (Read error: Connection reset by peer).
00:43:29 <zzo38> Sgeo_: I don't know. But I suggest you don't port MegaZeux to Flash.
00:43:57 <zzo38> Can you compile C programs with SDL into Flash? If so, then it is easy. But I don't think you can do that.
00:44:08 <elliott> Well, you can compile C to Flash.
00:44:12 <elliott> A guy ported Doom to Flash that way.
00:45:07 <zzo38> Sgeo_: Why do you want it in Flash anyways? It is much better native (it will both be faster and will not require a Flash player).
00:45:34 <Sgeo_> So more people will play MegaZeux games!
00:46:11 <zzo38> Sgeo_: MegaZeux will compile on most operating systems, including Nintendo DS.
00:47:02 <Sgeo_> It's impossible to get lazy people to go through that effort to play what feels like a flash game
00:47:04 <zzo38> (But if you can find a way to compile C programs with SDL into Flash, you can try compiling mzxrun only into Flash, although I still don't like it.)
00:51:03 <zzo38> Sgeo_: Mainstream MegaZeux executables are available for most common operating systems. For my version, executable is available only on Windows to download, but it works on other operating systems too. If anyone wants to contribute executables of my MegaZeux in other operating system, I might post a link to those ones too.
00:51:53 <zzo38> elliott: What I meant by "allocprintf" is one that is like
00:52:15 <zzo38> "sprintf" but it automatically allocates it too.
00:52:43 <elliott> zzo38: allocsprintf would be a better name. GLib has that, by the way.
00:53:10 <zzo38> elliott: I didn't know that before.
00:53:17 <elliott> zzo38: http://library.gnome.org/devel/glib/stable/glib-String-Utility-Functions.html#g-strdup-printf
00:53:58 <zzo38> But that's for GNOME isn't it?
00:54:45 <elliott> zzo38: And works in programs with any type of interface; it does not depend on any interface library.
00:54:59 <zzo38> Well, either way, I am writing the program without such functions because it doesn't have it, I was just mentioning something that I might find useful. It is not absolutely necessary.
00:55:51 <pikhq> zzo38: Glib just happens to be heavily used by Gnome; it's a fairly generic "throw everything in that should've been part of the C library" sort of thing, though.
00:56:09 <pikhq> Well, it also has that revolting object system.
00:57:12 <elliott> Yes, but you don't have to link that in. :p
00:58:25 <zzo38> Sgeo_: Have you played my MegaZeux game yet, though? (It is recommended you download my copy of both the world file and the MegaZeux program, but you don't have to (for Part I).)
00:58:59 <Sgeo_> pikhq, what's horrible about it?
00:59:16 <Sgeo_> I mean, besides trying to use it in its native language?
00:59:31 -!- cheater99 has joined.
00:59:57 <zzo38> What I know is that when I need a object system in a C program, I will write one that is suited for that specific program that I am writing.
01:00:24 <Sgeo_> zzo38, that's because you're zzo38 >.>
01:01:28 <elliott> GLib is pretty awful in reality :P
01:01:29 -!- cheater99 has quit (Read error: Connection reset by peer).
01:01:47 <Sgeo_> elliott, to use from C, or even from a language like Vala/Genie?
01:01:58 <elliott> GLib is just awful full stop.
01:02:03 <elliott> But it has some useful functions.
01:02:48 <zzo38> I have written more of TeXnicard already. Including pattern matching.
01:03:48 <elliott> pikhq: srsly though, I don't think Debian has a compiler-rt package.
01:04:02 <zzo38> elliott: What is a compiler-rt package?
01:05:45 <elliott> split out by stupid debian
01:06:04 <elliott> pikhq: Are you sure there's no way to cast a block to a function pointer?
01:07:26 <zzo38> I have a idea of esolang that nearly everything must be done by solving a halting problem. What is the minimum number of other things needed?
01:10:17 <pikhq> None that I know of.
01:12:58 <elliott> pikhq: Bleh, I can't figure this out at all...
01:15:22 -!- zzo38 has quit (Remote host closed the connection).
01:16:21 <elliott> pikhq: Why is parsing such a pain.
01:17:01 -!- cheater99 has joined.
01:19:29 -!- cheater99 has quit (Read error: Connection reset by peer).
01:22:19 -!- ais523 has quit (Remote host closed the connection).
01:24:03 -!- calamari has quit (Quit: Leaving).
01:24:22 <elliott> pikhq: Maybe I should just write the compiler in Haskell.
01:24:42 -!- infanticide has left (?).
01:34:07 <elliott> pikhq: HOW DO YOU FEEL ABOUT NESTED COMMENTS
01:34:33 -!- cheater99 has joined.
01:35:05 -!- pikhq has quit (Ping timeout: 250 seconds).
01:36:03 -!- cheater99 has quit (Read error: Connection reset by peer).
01:36:52 -!- pikhq has joined.
01:47:35 <Sgeo_> And that's considering I didn't eat much today
01:47:41 <Sgeo_> Maybe I'll be ok on Feb 10th
01:51:31 -!- cheater99 has joined.
01:54:00 -!- cheater99 has quit (Read error: Connection reset by peer).
02:00:02 -!- FireFly has quit (Quit: swatted to death).
02:06:21 <elliott> pikhq: Even Parsec is a bit of a bitch.
02:09:59 -!- cheater99 has joined.
02:11:17 -!- wareya_ has joined.
02:12:26 -!- cheater99 has quit (Read error: Connection reset by peer).
02:14:34 -!- wareya has quit (Ping timeout: 276 seconds).
02:19:44 <elliott> pikhq: You should make a parser library.
02:28:33 -!- cheater99 has joined.
02:31:11 -!- cheater99 has quit (Read error: Connection reset by peer).
02:38:23 <elliott> Mathnerd314: yep. and every response to it. and the author's re-response.
02:38:32 <elliott> but it doesn't really help my current situation :)
02:44:56 <Mathnerd314> my lang. a pre-pre-pre-release version of what will eventually rule the earth
02:45:08 <elliott> Mathnerd314: brief overview? and no, /my/ lang will rule the earth.
02:46:04 <Sgeo_> Chrome's been acting up again lately
02:46:15 <Sgeo_> I'm almost tempted to switch back to Opera
02:46:50 <Mathnerd314> elliott: it has message passing, and pattern matching, and multiple dispatch, and... the rest is part of the front-end I have to implement
02:46:58 <elliott> Mathnerd314: strongly typed?
02:47:35 -!- cheater99 has joined.
02:47:42 <elliott> Mathnerd314: is it purely functional or impure?
02:48:21 <elliott> Mathnerd314: hmm. maybe not entirely worthless then
02:48:34 <elliott> Mathnerd314: message passing/multiple dispatch have a lot of overlap -- why not just pick one?
02:49:09 <Mathnerd314> it has both; they're implemented in the same lines of code
02:49:09 -!- cheater99 has quit (Read error: Connection reset by peer).
02:49:19 <elliott> Mathnerd314: that's not a reason
02:49:38 <elliott> careful, thought-out design taking into account what to exclude as well as what to include > "it didn't take many lines of code so I threw it in"
02:49:39 <Mathnerd314> umm... because I like power, and this is more powerful?
02:49:55 <elliott> so far every time you've used the word power it's been to justify a bad idea ... but very well.
02:50:29 <Mathnerd314> actually, it doesn't *really* have pattern-matching
02:50:37 <elliott> i didn't even object to pattern matching
02:51:25 <Mathnerd314> yeah, that's why it doesn't have it. because it has no data structures.
02:51:28 <Sgeo_> Is it "more power" to drop the strong typing from Haskell? More power in that case -> suck
02:52:10 <Sgeo_> Mathnerd314, I was commenting on the idea of "power", not your language speciically
02:52:10 <elliott> Mathnerd314: if it has no data structures, what do you send messages to?
02:52:14 <elliott> also: how do you decide what dispatch to use?
02:53:26 <Mathnerd314> an environment takes some code and evaluates it. I have an environment which can construct other environments
02:54:30 <elliott> i still think that having both message passing /and/ multiple dispatch is a design flaw. or at least an indicator of badly-thought-out design.
02:55:21 <elliott> coppro: wait, how do lambda expressions in C++0x compare, for compiling to?
03:02:28 <Sgeo_> I'm going to go watch a video on Ur
03:02:33 <Sgeo_> Maybe I'll warm up to it
03:02:45 <elliott> Sgeo_: You won't unless you have a decent grasp of dependent types.
03:03:00 <elliott> pikhq: In which a company announces that their BSD-licensed fork of the BSD-licensed LimeChat, which they do not name or link to in the press release, will be sold in the Mac App Store for $9. http://www.codeux.com/textual/future.php
03:03:39 <elliott> pikhq: New item on my TODO list: Offer free builds, encourage people to link to it with "Textual" as the link text for googlebombing.
03:03:39 <Sgeo_> I don't have that kind of time :(
03:03:46 <elliott> Sgeo_: No, you can spend it on IRC instead.
03:04:12 <Sgeo_> elliott, if I start thinking it's ok to watch lengthy videos when I should be working, I'm screwed
03:04:59 -!- cheater99 has joined.
03:06:52 -!- cheater99 has quit (Read error: Connection reset by peer).
03:11:19 -!- wareya_ has changed nick to wareya.
03:13:34 <Sgeo_> http://www.khanacademy.org/ good way to learn some more advanced math?
03:15:28 <elliott> um i don't really see anything there i would consider very advanced
03:15:47 <elliott> khanacademy seems to be universally popular though.
03:18:10 <Sgeo_> DEAR FLASH: FUCK YOU FUCK YOU FUCK YOU FUCK YOU
03:18:35 <Sgeo_> Sometimes, when playing YouTube videos, the audio will keep going and the video will stop
03:20:11 <Sgeo_> elliott, I see stuff about linear algebra
03:20:37 <Sgeo_> It's stuff I don't know
03:20:44 <elliott> That doesn't mean it's advanced stuff.
03:22:59 -!- cheater99 has joined.
03:25:21 -!- cheater99 has quit (Read error: Connection reset by peer).
03:32:56 -!- oerjan has joined.
03:36:55 <elliott> http://esolangs.org/w/index.php?title=Udage&diff=20547&oldid=20545 <-- i totally schooled that cpressey dude
03:37:47 <elliott> haha, and some classic Graue rage:
03:37:49 <elliott> [[I reverted your purge of Talk:Udage because that isn't the way wikis work. You do not own that page, nor do you own the Udage article. Do not delete valid information from this site again. --Graue 19:16, 10 Oct 2005 (GMT)]]
03:40:53 <Sgeo_> So, if I try to pretend PSOX never existed, and try to remove all traces of it ever existing, I'll be yelled at?
03:41:33 -!- cheater99 has joined.
03:42:34 -!- cheater99 has quit (Read error: Connection reset by peer).
03:42:50 <oerjan> Sgeo_: sheesh, obviously you'll be yelled at regardless
03:43:09 <oerjan> _maybe_ a bit more in that case
03:45:07 <elliott> the basic thing to understand, Sgeo_, is that people like yelling at you
03:50:03 <Sgeo_> I know you all secretly love me
03:50:08 <oerjan> Sgeo_: NOW DO YOUR HOMEWORK
03:51:55 <elliott> oerjan: you should ban Sgeo_.
03:52:45 <oerjan> also with my own procrastination that would be bad karma.
03:54:51 <elliott> oerjan: no, i just want you to ban him to make him go away
03:58:08 -!- cheater99 has joined.
04:00:21 -!- cheater99 has quit (Read error: Connection reset by peer).
04:16:01 -!- cheater99 has joined.
04:17:34 -!- cheater99 has quit (Read error: Connection reset by peer).
04:20:06 <elliott> http://en.wikipedia.org/wiki/Computable_number#Formal_definition computable dedekind cuts -- awesome!
04:22:49 <Ilari> Hmm... Last 30 days allocation lists for APNIC, 16x/14, 6x/13, 4x/12. That's about 11Mi addresses (plus then there are the <250k allocations...)
04:23:39 <Sgeo_> http://www.amazon.com/Wishing-Well-Making-Your-Every/dp/078686561X . This book. This horrible, evil book
04:25:06 <Ilari> Also intersting to look at raw allocation counts for 250k+ allocations: 5 for RIPE, 1 for ARIN, 1 for LACNIC, 2 for AfriNIC and 26(!) for APNIC.
04:26:33 <Ilari> Runner-up in large allocations: RIPE with 5.25Mi. That's not even half of APNIC.
04:29:42 <Sgeo_> As a kid, I had a bad reaction to the woowoo in that book
04:30:18 <Sgeo_> Some people find comfort in that sort of BS. I found abject fear that haunts me to this day
04:32:36 <pikhq> 5 forms of the final boss remaining.
04:33:02 <Ilari> Allocations this month: ARIN: 1 693 696. AfriNIC: 384 000. LACNIC: 1 095 936, RIPE: 3 953 920, APNIC: 12 063 232. Yes, APNIC allocated way more addresses than all the other RIRs COMBINED.
04:33:33 -!- cheater99 has joined.
04:34:32 <elliott> Sgeo_: You're fucking crazy.
04:35:09 <Sgeo_> elliott, the claims of that book have had such a profound negative influence on me
04:35:35 <elliott> Sgeo_: Did I mention that you're fucking crazy?
04:35:54 <Sgeo_> elliott, are you unaware of how easily kids can be affected by this sort of stuff?
04:36:07 <elliott> Sgeo_: No, no, I'm not quite sure you *understand*: you're fucking crazy.
04:36:47 -!- cheater99 has quit (Read error: Connection reset by peer).
04:37:23 <pikhq> 3 forms remaining...
04:37:31 <Sgeo_> pikhq, YouTube it all
04:37:55 <elliott> pikhq: Are you doing it with just the infinity-plus-one and mandatory bosses?
04:37:57 <elliott> pikhq: If not: wtf i hate you
04:38:56 <Sgeo_> There's different woowoo I believed as a kid, that wasn't so traumatic
04:39:11 <pikhq> elliott: I'm finishing my first playthrough dammit.
04:39:22 <pikhq> And the final boss is still a pain with the infinity-plus-one.
04:39:30 <Sgeo_> elliott, did you just call Ur lame/
04:39:33 <pikhq> You see, it has 12 forms. You have to kill each one.
04:40:07 <pikhq> Each one is a difficult boss in and of itself.
04:41:25 <pikhq> Except I one-shot each one.
04:41:37 <pikhq> I just need to have enough full-SP-heal items to get through.
04:41:58 <Sgeo_> I find no fear in the idea of an afterlife, only comfort. In the idea of being able to control reality with your mind, I find the most horrible painful fear.
04:46:08 <pikhq> elliott: Yes, I used Satan to defeat the end of the world.
04:46:12 <elliott> Sgeo_: You can do that, it happens by sending a few signals down to effectors which effect changes in the world around them.
04:46:42 <Sgeo_> pikhq, it would be better if it were an unwilling Anti-Christ
04:47:19 <pikhq> elliott: I could've used Odin instead; would that have made you feel better?
04:48:05 <elliott> Sgeo_: Why aren't you in abject terror?
04:48:16 <pikhq> Okay, then. Satan it was.
04:48:24 * pikhq rocks the devil horns
04:48:34 <Sgeo_> Because the only real way for my mind to affect reality is limited
04:48:58 <Sgeo_> My OCDish thoughts of people dying aren't going to kill them
04:49:02 <elliott> Not very limited if you're an atom.
04:49:09 <elliott> What the fuck has OCD got to do with that what
04:49:33 <Sgeo_> I thought OCD has to do with more severe unwanted thoughts than normal
04:49:48 <elliott> I don't think you have any idea what OCD is, shut up
04:50:12 <Sgeo_> Well, I'm pretty sure there is a name for what I just described. I don't think I have it, but
04:50:44 <elliott> Sgeo_: Um, thinking things you don't want to?
04:50:48 <elliott> Here's a name for that: existing.
04:50:57 <Sgeo_> elliott, but more severely and painfully than normal
04:50:59 <pikhq> Sgeo_: OCD implies obsessive compulsions. And that it's a disorder.
04:51:08 <elliott> Sgeo_: Just checking -- are we talking physical pain here
04:51:51 <Sgeo_> elliott, I don't know, ask the person I was talking to
04:52:15 <elliott> Sgeo_: YOU JUST TALKED ABOUT "[THAT]LY" THOUGHTS, IF THEY DON'T EVEN CORRELATE WITH YOUR ONE-LINE DESCRIPTION THEN YOUR STATEMENT WAS BEYOND MEANINGLESS >_<
04:52:33 -!- cheater99 has joined.
04:54:38 <Sgeo_> Well, you know what sort of thoughts I'm talking about, you said they were a part of existing
04:54:55 <elliott> Let's call it QRD. For the sake of argument.
04:55:04 <elliott> If your thoughts did not involve the mentioned pain, how can you even say they are QRDly?
04:55:36 <Sgeo_> Because the other attribute of being unwanted is still there
04:55:36 -!- cheater99 has quit (Read error: Connection reset by peer).
05:01:10 <elliott> Sgeo_: So let's say that a giraffe is an animal with spots and a long neck.
05:01:22 <elliott> Sgeo_: Therefore jaguars are giraffey.
05:01:46 <Sgeo_> Ok, maybe I shouldn't have said OCDish, or maybe my understanding is wrong
05:07:36 <elliott> whoa myndzi is pronounced mind-zy
05:07:41 <elliott> 02:19:02 <myndzi> ironic seeing as how i use a y in my nick for a long i sound
05:07:51 <Ilari> Yeah, in large allocations, APNIC allocated almost 3 times more allocations than rest of the world combined...
05:08:21 <oerjan> elliott: "mind's eye" hth
05:08:23 <elliott> Ilari: so how long until the end now
05:08:35 <elliott> oerjan: i thought it was just... "min-dzi"
05:08:43 <elliott> like mindy but with more z
05:09:01 <Ilari> IANA depletion in this or next month, APNIC depletion sometime in next year...
05:09:06 <oerjan> elliott: mind you i don't have real confirmation of this theory
05:11:35 -!- cheater99 has joined.
05:12:35 <Ilari> Those who model RIR depletions are starting to have their doubts to their models...
05:14:15 -!- cheater99 has quit (Read error: Connection reset by peer).
05:15:20 <Ilari> If you want to see example, look at the latest note to the IPv4 Address Report.
05:17:07 <Ilari> Hah: "[...]which, by the way, currently predicts that IANA will hand out its last IPv4 address blocks on 10 June 2011[...]". That was September this year. Now the prediction is February (and probably too optimistic)...
05:18:19 <Ilari> The main problem with predicting exhaustions: Badly behaved distributions and changes to distribution parameters...
05:19:02 <Sgeo_> I'll take door number e!
05:19:23 <pikhq> There's optional post-game content!
05:22:08 <Sgeo_> http://tr.froup.com/tr.pl?1671
05:24:43 <Ilari> The spread between highest and lowest daily allocation rates (for days with allocations) is more than factor of 1000.
05:26:57 -!- elliott has quit (Quit: Leaving).
05:29:14 -!- elliott has joined.
05:29:16 <elliott> 13:31:46 <Phantom_Hoover> I borrowed it without intent to return.
05:29:17 <elliott> 13:31:52 <Vorpal> Phantom_Hoover, that's nasty
05:29:21 -!- elliott has quit (Remote host closed the connection).
05:30:31 -!- cheater99 has joined.
05:33:59 -!- cheater99 has quit (Read error: Connection reset by peer).
05:37:58 <pikhq> It's been an hour since I beat the final boss and I'm still playing the game.
05:38:08 <pikhq> Not merely seeing cutscenes, oh no. Still playing.
05:39:38 -!- elliott has joined.
05:39:39 <elliott> 20:33:49 <cpressey> actually damn, everything can be mass-nouned
05:39:41 <elliott> 20:33:59 <cpressey> "we don't have enough mousepad yet"
05:39:44 -!- elliott has quit (Remote host closed the connection).
05:39:59 * pikhq wonders what elliot is doing
05:40:21 -!- elliott has joined.
05:40:24 <elliott> 20:59:39 <cpressey> C is an awesome language
05:40:25 <elliott> 21:01:03 <cpressey> it's like nuclear lego
05:40:30 -!- elliott has quit (Remote host closed the connection).
05:42:49 <pikhq> Going crazy. Got it.
05:43:24 * oerjan picks up a dropped t and stabs pikhq with it
05:45:23 <pikhq> oerjan: I was talking to elliot, not elliott, of course.
05:46:07 <oerjan> TOO BAD YOU DIDN'T SAY THAT BEFORE I FATALLY STABBED YOU
05:46:44 <pikhq> Aaand the game is finally freaking over.
05:50:01 -!- cheater99 has joined.
05:52:28 -!- cheater99 has quit (Read error: Connection reset by peer).
05:55:27 -!- Sasha2 has joined.
05:55:27 -!- Sasha has quit (Read error: Connection reset by peer).
05:57:01 -!- wareya has quit (Ping timeout: 276 seconds).
05:58:44 -!- wareya has joined.
06:08:33 -!- cheater99 has joined.
06:10:49 -!- cheater99 has quit (Read error: Connection reset by peer).
06:19:06 -!- oerjan has quit (Quit: leaving).
06:26:59 -!- cheater99 has joined.
06:28:12 -!- cheater99 has quit (Read error: Connection reset by peer).
06:44:33 -!- cheater99 has joined.
06:48:54 -!- cheater99 has quit (Read error: Connection reset by peer).
07:01:41 -!- Zuu has quit (Read error: Connection reset by peer).
07:04:34 -!- cheater99 has joined.
07:07:16 -!- cheater99 has quit (Read error: Connection reset by peer).
07:10:33 -!- Zuu has joined.
07:15:32 -!- augur has changed nick to Aug10.
07:16:43 -!- Aug10 has changed nick to augur.
07:23:34 -!- cheater99 has joined.
07:24:08 -!- zzo38 has joined.
07:25:27 <zzo38> Welcome to the esoteric programming channel!
07:25:33 -!- Goosey has quit (Quit: When the chips are down, well, the buffalo is empty).
07:28:35 -!- cheater99 has quit (Read error: Connection reset by peer).
07:33:55 -!- zzo38 has quit (Quit: No brain, no headache.).
07:44:34 -!- cheater99 has joined.
07:48:14 -!- cheater99 has quit (Read error: Connection reset by peer).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:04:06 -!- cheater99 has joined.
08:06:20 -!- cheater99 has quit (Read error: Connection reset by peer).
08:18:34 -!- roper has joined.
08:19:35 -!- oerjan has joined.
08:22:34 -!- cheater99 has joined.
08:26:01 -!- cheater99 has quit (Read error: Connection reset by peer).
08:33:31 -!- Phantom_Hoover has joined.
08:42:04 -!- cheater99 has joined.
08:44:53 -!- cheater99 has quit (Read error: Connection reset by peer).
08:49:44 <Ilari> The Lagerholm estimate jumed to "Today's IANA depletion date estimate: 2011-01-16".
08:51:14 <Ilari> What threw: It: 2Mi allocations to AU from APNIC.
08:54:32 <Ilari> 2.66 blocks in RIR pool...
08:59:32 -!- cheater99 has joined.
09:01:34 -!- Sgeo_ has quit (Read error: Connection reset by peer).
09:03:05 -!- cheater99 has quit (Read error: Connection reset by peer).
09:05:25 <Ilari> About 11M addresses left until APNIC gets another blocks, triggering X-day immediately.
09:16:02 -!- oerjan has quit (Quit: leaving).
09:18:34 -!- cheater99 has joined.
09:23:12 -!- cheater99 has quit (Read error: Connection reset by peer).
09:34:17 <Ilari> Sorry, insufficient data to estimate that. :-/
09:37:34 -!- cheater99 has joined.
09:39:36 -!- cheater99 has quit (Read error: Connection reset by peer).
09:52:52 -!- wareya_ has joined.
09:52:54 -!- nefasto has joined.
09:55:34 -!- cheater99 has joined.
09:55:40 -!- wareya has quit (Ping timeout: 240 seconds).
09:58:51 -!- cheater99 has quit (Read error: Connection reset by peer).
10:12:13 -!- nefasto has quit (Quit: http://irc2go.com/).
10:14:59 -!- cheater99 has joined.
10:17:06 -!- cheater99 has quit (Read error: Connection reset by peer).
10:32:35 -!- cheater99 has joined.
10:35:22 -!- cheater99 has quit (Read error: Connection reset by peer).
10:51:07 -!- cheater99 has joined.
10:53:41 -!- cheater99 has quit (Read error: Connection reset by peer).
11:09:35 -!- cheater99 has joined.
11:13:03 -!- cheater99 has quit (Read error: Connection reset by peer).
11:29:07 -!- cheater99 has joined.
11:33:04 -!- cheater99 has quit (Read error: Connection reset by peer).
11:48:35 -!- cheater99 has joined.
11:50:24 <Vorpal> <elliott> 20:59:39 <cpressey> C is an awesome language
11:50:24 <Vorpal> <elliott> 21:01:03 <cpressey> it's like nuclear lego
11:50:24 -!- cheater99 has quit (Read error: Connection reset by peer).
11:50:43 <Vorpal> the only problem with that analogy is that C is way less modular
11:51:21 <Vorpal> (so, it is like lego, except when it comes to one of the key defining properties of lego)
11:52:38 <Vorpal> Phantom_Hoover, at some point I presume
11:55:20 <Phantom_Hoover> 23:24:08 --- join: zzo38 (~zzo38@h24-207-49-17.dlt.dccnet.com) joined #esoteric
11:55:20 <Phantom_Hoover> 23:25:27 <zzo38> Welcome to the esoteric programming channel!
11:55:52 <Phantom_Hoover> I'd be baffled if anyone else did it, but zzo has desensitised me to his weirdness.
12:07:53 -!- cheater99 has joined.
12:13:02 <Ilari> Okay, why is stunnel trying to create localhost-only AF_INET port and then trying to connect to it?
12:16:47 <Vorpal> Ilari, what is stunnel for now again?
12:17:14 <Ilari> Similar to inetd, but listens for TLS connections instead of plaintext.
12:17:49 <Ilari> "trying" because it doesn't work and as result, stuff keeps failing...
12:17:55 <Vorpal> Ilari, must be config issues
12:18:14 <Vorpal> (I think that exhausted all possibilities :P)
12:22:01 <Vorpal> Ilari, has it worked before?
12:23:05 <Vorpal> Ilari, so what did you change (if anything?)
12:23:38 <Ilari> Hmm... Maybe it has always done that localhost connect thingy...
12:23:56 <Vorpal> Ilari, well, presumably it hasn't always been failing
12:24:44 <Vorpal> hm, is lo up when you get to userspace on linux?
12:24:52 <Vorpal> or does some early init script start lo
12:26:02 <Vorpal> ah, /etc/rc.sysinit has /sbin/ifconfig lo 127.0.0.1 up
12:26:24 <Vorpal> right after udev basically
12:27:07 <Ilari> Ah, apparently strunnel is configured to use AF_INET instead of AF_UNIX for temporary internal sockets. Bleh.
12:27:28 <Vorpal> Ilari, the whole "temporary internal sockets" thing sounds weird
12:27:42 <Vorpal> Ilari, however, connecting over lo should work on every sane system
12:28:10 <Vorpal> if someone has a crazy enough firewall that it doesn't work, then it is really their own issue
12:29:47 <Vorpal> $ unbound-control dump_cache | wc -l
12:30:08 <Vorpal> for wc -c it is 1286918
12:30:25 <Ilari> recompiling strunnel with saner settings....
12:30:49 <Vorpal> $ unbound-control dump_cache | grep -E 'IN[ \t]*A' | wc -l
12:31:40 -!- augur has quit (Ping timeout: 240 seconds).
12:31:45 <Vorpal> IN NS is 971, IN SOA is 1417
12:32:11 <Vorpal> $ unbound-control dump_cache | grep -E '^;' | wc -l
12:33:17 <Vorpal> wait, that seems wrong above
12:33:32 <Vorpal> quite, the \t didn't expand it seems
12:34:51 -!- cheater99 has quit (Read error: Connection reset by peer).
12:34:52 <Vorpal> unbound-control dump_cache | grep -Eo $'IN[ \t]*[^ \t]+' | awk '{print $2}' | sort -n | uniq -c | sort -n
12:35:07 <Vorpal> I think the parsing fails somewhat
12:35:25 <Vorpal> ah + there gives more reasonable results...
12:35:53 -!- augur has joined.
12:39:51 <Ilari> Hah... stunnel segfaulted.
12:40:32 <Vorpal> Ilari, why is it an issue that it uses AF_INET and lo?
12:43:10 <Vorpal> unbound-control dump_cache | sed '/START_MSG_CACHE/q' | grep -Ev '^;' | grep IN | awk '{print $4}' | sort -n | uniq -c | sort -n
12:43:32 <Vorpal> Ilari, they restrict lo?
12:43:38 <Vorpal> Ilari, if so they are utterly bonkers
12:44:19 <Vorpal> Ilari, you should setup a data diode or whatever it is called on lo. Just firewall is too insecure!
12:45:50 <Ilari> Stunnel crashes even if client doesn't try to send client certificate.
12:48:09 <Vorpal> Ilari, but why are you firewalling lo to this insane degree?
12:49:19 <Ilari> There's stunnel3 compat binary as well. Crashes just as good when connecting.
12:49:43 <Vorpal> Ilari, yeah, but uh, why are you firewalling lo to this degree. That is the real wtf
12:50:32 -!- cheater99 has joined.
12:53:30 <Ilari> Tried stunnel 3.26 too. Crashes on connect.
12:53:40 -!- sftp_ has joined.
12:54:02 -!- sftp has quit (Ping timeout: 245 seconds).
12:54:21 <Ilari> -> Stunnel sucks...
12:54:36 <Vorpal> Ilari, just fix the insane firewall rule
12:54:51 <Vorpal> Ilari, really, what are the rules for lo?
12:55:23 <Ilari> Same as global, plus few extra ports allowed.
12:55:34 <Vorpal> Ilari, why are you surprised stuff breaks then
12:56:08 <Ilari> I expect software to use AF_UNIX for local stuff internally, not AF_INET to loopback.
12:56:18 <Vorpal> Ilari, the usual way is to check for invalid state (and on ipv6 for rt type 0), then the rule after that/those is generally "if loopback, allow"
12:56:30 <Vorpal> Ilari, you expect too much
12:56:49 <Vorpal> Ilari, besides, what does it add to security? I doubt anything.
12:57:01 <Deewiant> Windows-portable software will likely not use AF_UNIX for anything.
12:57:26 <Ilari> I actually have built version of stunnel that uses AF_UNIX as it should. Except that that crashes if client tries to connect...
12:57:36 <Vorpal> Ilari, probably because few people use it
12:58:10 <Vorpal> Ilari, also, shouldn't you use something like selinux or similar to limit what unix sockets each program can use
12:58:14 -!- oerjan has joined.
12:58:18 <Ilari> It crashes inside SSL client certificate routines, which AFAIK come before it tries to use internal sockets...
12:58:40 <Vorpal> Ilari, I suggest you go back to the previous working setup and try again
12:59:04 <oerjan> damn rsi acting up again :(
12:59:33 <Vorpal> oerjan, (as in, which meaning of rsi)
12:59:45 <oerjan> the one which it's painful to type out
13:00:18 <Vorpal> oerjan, ah that would be "Register Storage Immediate"
13:00:27 <Ilari> It crashes instantly after receiving the client certificate...
13:00:38 <Vorpal> Ilari, even on the previous working setup?
13:00:54 <Vorpal> oerjan, use speech to text or something
13:01:00 <Vorpal> oerjan, it would be a good laugh
13:01:28 <oerjan> this laptop doesn't actually have a microphone. also i hate unnecessary sounds.
13:03:07 <Ilari> Okay, 3.26 works (including invoking the program) if I don't send a client ceritificate. 4.34 crashes even if I don't send it.
13:03:28 <Vorpal> Ilari, is this with or without insane firewall setup?
13:03:43 <Vorpal> Ilari, for purposes of debugging you should test if it affects anything
13:08:11 <Vorpal> oerjan, no external mic?
13:09:52 <Vorpal> oerjan, also, I never imagined you as one using a laptop.
13:10:02 <Vorpal> oerjan, I mean, a vt100 sounds more your style
13:10:08 <Vorpal> or maybe a workstation
13:10:26 <oerjan> sure, when i was at the university...
13:10:41 <Vorpal> oerjan, yeah, but laptop just doesn't suite you!
13:10:53 <Vorpal> oerjan, at least use a desktop
13:11:01 <Vorpal> even I use a desktop at home
13:12:58 <Vorpal> oerjan, besides laptops are not comfortable for long time use (say, more than a 4-5 hours / day)
13:13:53 <Vorpal> oerjan, the touchpad is useless. And the trackpoint finger gets sore after a few hours
13:14:14 <Vorpal> mouse ends up at wrong height often.
13:14:24 <oerjan> i don't mind the touchpad, there is no trackpoint
13:14:24 <Vorpal> and of course the monitor is in the wrong place
13:15:26 <Vorpal> oerjan, and the travel distance on laptops sucks
13:16:45 <Vorpal> oerjan, at least I hope it has full sized keys
13:16:45 <oerjan> well i don't actually bring it anywhere anyhow
13:17:28 <Vorpal> oerjan, anyway, they are inferior to a proper mechanical keyboard
13:18:18 <Ilari> Ah, found out why stunnel crashes: OpenSSL library/header mismatch.
13:18:49 <Ilari> Now it works even with client certs...
13:21:57 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
13:22:01 -!- Phantom_Hoover_ has joined.
13:36:48 -!- cheater99 has quit (Read error: Connection reset by peer).
13:39:49 <Vorpal> what was that site to find shortest path on wikipedia?
13:40:31 <oerjan> Gregor had a game based on it
13:40:41 <Vorpal> oerjan, I remember some automated tool for it
13:41:13 <Vorpal> fizzie, right, happen to know where to find one?
13:41:15 <fizzie> Quick googling found me http://thewikigame.com/ but not the one that had the psychedelic color-flashery and link-dropping feature.
13:41:35 <Vorpal> fizzie, the one I remember was just a pain white page
13:41:37 <fizzie> (That one has a few game modes though.)
13:42:00 <Vorpal> fizzie, and I don't want it as a game, I want it as a tool that tells me the shortest path.
13:42:29 <Vorpal> fizzie, yes, I googled but can't find it
13:42:38 <fizzie> http://www.netsoc.tcd.ie/~mu/wiki/ ?
13:42:42 <fizzie> (It's linked from Wikipedia.)
13:43:10 <Vorpal> fizzie, hm, I guess that works. Not the one I thought of. But any tool that does the job is okay
13:43:20 <fizzie> It's a bit dated (March 2008) now.
13:43:29 -!- Phantom_Hoover has joined.
13:45:17 -!- Phantom_Hoover_ has quit (Ping timeout: 245 seconds).
13:45:17 <fizzie> There's also http://www.xltd.com/WikiMindMap/WikiPath.htm but that's not one I remember having seen before. That one doesn't say which day's dump they're using.
13:46:02 <Vorpal> fizzie, it could be http://www.netsoc.tcd.ie/~mu/wiki/ if they redesigned that page. The result page looks similar to what I remember
13:46:21 <fizzie> I guess they could've just prettified it up a bit.
13:47:56 <Vorpal> fizzie, anyone know what the longest path is (excluding those of infinite length due to lack of any links between)
13:49:10 <fizzie> "Several people were asking about what's known as the "diameter" of Wikipedia, that is, the distance between the two articles furthest apart (the longest shortest path if that makes any sense to you). This was in fact the original goal of the project but it turned out not to be very interesting. Wikipedia has giant "tails", almost linear linked lists of articles that stretch out for 70 links."
13:49:16 <fizzie> That's on that six-degrees page.
13:49:40 <fizzie> "Even when I special-cased out that string of 70 boring articles, a new one appeared (I think it was linked pages about administrations of Myanmar or something)."
13:50:04 <fizzie> But the center of wikipedia (in March 2008) was "2007".
13:51:09 <oerjan> didn't they change policy to include less year links in articles?
13:51:26 <oerjan> i think i saw some being edited away at one time
13:53:05 <oerjan> don't recall if it was before or after 2008 though
13:53:21 <Vorpal> fizzie, hm, I wonder if wikipedia has an Eulerian path...
13:53:40 -!- yiyus has quit (Read error: Operation timed out).
13:54:22 <Vorpal> a Hamiltonian path would be interesting as well
13:54:52 <fizzie> There are pages with no incoming links and multiple outgoing links (like some disambig pages), so no.
13:55:00 <oerjan> both of those only need three articles with either no paths out or no ... right
13:55:42 <Vorpal> fizzie, what about the set of all pages that can be reached from the centre.
13:55:55 <Vorpal> still, dead ends I guess
13:56:06 <oerjan> you need paths both ways
13:56:15 <fizzie> Yes; I don't know about the strongly-connected component, though.
13:56:27 <Vorpal> so, we need to carefully tweak the pages so that such a graph is possible
13:56:36 <oerjan> even then it can easily fail i think
13:56:50 <fizzie> "A directed graph is Eulerian if it is strongly connected and every vertex has equal in degree and out degree."
13:56:58 <fizzie> That sounds like it's unlikely to hold.
13:57:04 <oerjan> for eulerian path you cannot have too many with an odd number of in/out neighbors
13:57:23 <Vorpal> fizzie, so Hamiltonian path might be more feasible
13:57:26 <oerjan> oh wait _directed_ too
13:57:39 <fizzie> Well, the links are quite naturally directed.
13:58:36 -!- Phantom_Hoover_ has joined.
14:00:38 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
14:05:34 -!- cheater99 has joined.
14:06:38 <Vorpal> ineiros, ??? at your server
14:06:45 <Vorpal> ineiros, I got in, but wtf
14:07:21 <Vorpal> ineiros, I'm waiting for the real world to be back before I try again
14:07:32 <Vorpal> ineiros, you have some completely weird place there now
14:08:18 <Vorpal> fizzie, try yourself. There were a number of monster spawner cages embedded in the ground near spawn. And it wasn't the usual world.
14:08:41 <fizzie> Well, it was also beta 1.1_02 and no hMod.
14:08:45 <fizzie> So I assume it's for testing.
14:09:01 <Vorpal> well, not much point in playing on it
14:25:33 <Phantom_Hoover_> How do I get Debian to let me install Sid packages if necessary, but use the Squeeze packages by default?
14:27:50 <fizzie> Add both to sources.list, then write an APT preferences file to prefer squeeze, but include sid with a lower priority.
14:27:55 <fizzie> Cf. http://jaqque.sbih.org/kplug/apt-pinning.html
14:31:15 <fizzie> Well, it has worked for me for a mixed-distribution Debian. How is it not working?
14:31:46 <Phantom_Hoover_> apt-get upgrade still lists insanely huge numbers of packages to upgrade!
14:32:15 <fizzie> Oh, http://wiki.debian.org/AptPreferences is the official version of above.
14:33:43 <fizzie> What about "apt-get --dry-run -t testing upgrade", does that want to upgrade the world too?
14:34:46 <fizzie> It should be no different than using the preferences-file driven pinning, though.
14:35:22 <Phantom_Hoover_> fizzie, assuming I can s/testing/squeeze/, since that's what's on all of my apt lines.
14:35:45 <fizzie> Well, you could try both. I think they should be usable interchangeably there.
14:42:09 -!- elliott has joined.
14:45:11 <elliott> 04:59:04 <oerjan> damn rsi acting up again :(
14:45:13 <elliott> oerjan: not carpal tunnel?
14:45:50 <oerjan> i don't know precisely
14:45:56 <elliott> 05:12:58 <Vorpal> oerjan, besides laptops are not comfortable for long time use (say, more than a 4-5 hours / day)
14:46:04 <elliott> Vorpal: i've been using this laptop exclusively for many months now
14:46:41 <elliott> oerjan: i hope you enjoyed vorpal criticising every aspect of your computer, i bet you're going to replace it now
14:46:47 <elliott> with all that new knowledge
14:47:36 <elliott> Phantom_Hoover_: se.mirrors.kernel.org
14:48:07 <elliott> 05:46:02 <Vorpal> fizzie, it could be http://www.netsoc.tcd.ie/~mu/wiki/ if they redesigned that page. The result page looks similar to what I remember
14:48:13 <elliott> http://mirrors.se.kernel.org/debian/
14:49:12 <elliott> 06:06:38 <Vorpal> ineiros, ??? at your server
14:49:12 <elliott> 06:06:45 <Vorpal> ineiros, I got in, but wtf
14:49:12 <elliott> 06:07:21 <Vorpal> ineiros, I'm waiting for the real world to be back before I try again
14:49:12 <elliott> 06:07:32 <Vorpal> ineiros, you have some completely weird place there now
14:49:12 <elliott> 06:07:51 <Phantom_Hoover_> Vorpal, screenshot?
14:49:13 <elliott> 06:08:18 <Vorpal> fizzie, try yourself. There were a number of monster spawner cages embedded in the ground near spawn. And it wasn't the usual world.
14:49:28 <elliott> 06:25:33 <Phantom_Hoover_> How do I get Debian to let me install Sid packages if necessary, but use the Squeeze packages by default?
14:49:31 <elliott> Phantom_Hoover_: YOU NEVER, EVER DO THAT.
14:49:34 <elliott> Phantom_Hoover_: DO. NOT. DO. THAT.
14:50:04 <elliott> Phantom_Hoover_: Rephrase. Do not do that unless you /really, really know what you're doing/. *REALLY.*
14:51:12 <elliott> Phantom_Hoover_: Just manually dpkg -i the few packages from sid you really want.
14:51:31 -!- FireFly has joined.
14:53:19 <elliott> 06:08:18 <Vorpal> fizzie, try yourself. There were a number of monster spawner cages embedded in the ground near spawn. And it wasn't the usual world.
14:53:26 <elliott> Vorpal: you didn't destroy them, did you?
14:55:08 <Vorpal> elliott, why do you ask
14:55:48 <elliott> In case you did. You said "were".
14:56:00 <elliott> fizzie: Is there a way to get into that blocked-off tunnel near spawn? :p
15:00:40 <elliott> Killed a fucking zombie with my bare fucking hands, chickened out, disconnected.
15:02:13 <elliott> Phantom_Hoover_: A creeper and a skeleton are currently molesting me from the game over screen.
15:02:38 <elliott> Phantom_Hoover_: So I should have rounded up a few enemies if anyone else wants to go on :P
15:02:41 <elliott> Phantom_Hoover_: Did I mention I ran into the creeper to avoid the horrible fate of zombies?
15:03:11 -!- Phantom_Hoover_ has quit (Quit: holidays).
15:03:12 <elliott> bitches can't climb up trees
15:04:48 <elliott> the mob spawners just spawn pigs :(
15:04:51 <elliott> fizzie: Got any explanation for this? :P
15:05:09 <elliott> fizzie: right clicking while holding a feather on dirt or sth ->
15:05:10 <elliott> 15:04:27 [INFO] Connecting to a322.org:25566...
15:05:14 <elliott> 15:04:23 [DIED] protocol.c: 359: Unknown packet id: 0xee (dir 2)
15:05:37 <fizzie> Probably desynced due to <insert reason>.
15:05:55 <fizzie> I still can't quite "get over" not adding a length field in the "protocol".
15:06:14 <elliott> So has anyone actually done anything other than build a blocked-off tunnel at spawn? :p
15:06:30 <elliott> fizzie: Holding down right click w/ feather -> pretty reliable crash if you walk
15:07:41 -!- MigoMipo has joined.
15:07:45 <fizzie> I'll take a look at home, hopefully. Though there's a movie-watching thing, so I might not have time.
15:08:03 <elliott> fizzie: So is the server staying like this? :p
15:08:21 <fizzie> I don't know; I haven't heard anything from ineiros except the "hello" on-channel.
15:09:08 <elliott> fizzie: Let's just say yes, 'cuz this is pretty awesome.
15:09:37 <fizzie> dir 2 is PACKET_TO_SERVER, though, so it got desynk'd in the client-to-server stream, which is nice, because that one tends to be less complicated and easier to anamalyze.
15:10:53 <fizzie> Packet ID 0x05 at least seems to have been completely repurposed, so I'll have to update for that.
15:11:16 <elliott> Iron at (17,-153) for whenever I'm not shitfaced.
15:11:27 <elliott> My terminology does not have to sense-make to be use-make.
15:13:47 <elliott> THINGS THAT DO NOT WORK: Jumping off a cliff to get down
15:14:05 <elliott> 15:13:29 [CHAT] //coords: x=41, z=31, y=63
15:14:05 <elliott> 15:13:30 [CHAT] //coords: x=41, z=31, y=63
15:14:11 <elliott> fizzie: There's a blank line after the chat prompt.
15:14:22 <elliott> 15:08:40 [INFO] Starting up...
15:14:22 <elliott> 15:10:36 [CHAT] //coords: x=17, z=-153, y=65
15:14:22 <elliott> 15:12:21 [CHAT] //coords: x=45, z=-242, y=85
15:14:26 <elliott> Blank line before that last one too.
15:21:20 <ineiros> elliott: The placed spawners apparently only spawn pigs. I would like to have the capability to edit them, though. But hey, free bacon.
15:22:20 <ineiros> I thought to have something else there for a few days until the hMod is stable again.
15:23:09 <ineiros> Also, if someone wants to have maps at some point, you should point me to a system that can handle the current mess.
15:25:22 <elliott> ineiros: You use cartomograph, right?
15:25:27 <elliott> Also: plz keep it like this, it's awesome.
15:29:22 <elliott> 15:28:41 [DIED] protocol.c: 359: Unknown packet id: 0x19 (dir 2)
15:31:05 <fizzie> ineiros: Why didn't you just patch the region?
15:32:38 <ineiros> fizzie: Setting the extents manually caused an exception elsewhere in the code. Didn't have motivation to look further into the code.
15:33:18 <fizzie> Also, wasn't it pynemap instead of pymap?
15:34:14 <fizzie> I haven't tried the others; Cartographer 5's Linux port didn't have extents-setting features.
15:34:32 <fizzie> Someone has done a google-maps driven map already, though, http://www.triangularpixels.com/Junk/TectonicusTest/map.html
15:36:03 <fizzie> http://www.minecraftwiki.net/wiki/Programs_and_Editors#Mappers says of one that it's the "The fastest mapper available; capable of processing very large maps. Supports multiple modes including cave maps and resource highlighting." but it seems to be Windows-only, according to them little icons.
15:39:37 <fizzie> I still want to do that google maps / openlayers -based web-tile-mapper at some point, though I'd prefer if it had quasi-real-time player markers and such from a server plugin too.
15:46:10 <elliott> fizzie: The Google Maps ones take up many-the-gigs of storage space, I gather.
15:46:55 <fizzie> They could have a limited-size cache and on-the-fly generation.
15:47:59 <elliott> fizzie: That'll not slow the server down at all, nope.
15:48:45 <fizzie> Well, it could be done on some other box with the world rsync'd every now and then. Not sure how easy it is to add real-life geodata on top of gmaps. (Or openlayers.)
15:48:49 <elliott> "Minecraft is about to go up in price, so today is your last chance to buy it for €10. Obviously I recommend it. When it officially reaches beta today, the main niggles with the mutliplayer will be gone and the price will be €15." -- yes, those niggles like not being able to duplicate shit :)
15:49:36 -!- Sasha2 has quit (Ping timeout: 240 seconds).
15:50:17 <elliott> "I stock up on wood. I stock up on stone. I stock up on metal. I even take some sand. Then I fret that the sand and wood is taking too much potential ham-space, and ditch it. Then I stock up on water. Then I wonder if I should go looking for more diamond to make armour with. Then I wonder if my beacon is tall enough. Then I wonder if I have enough pickaxes.
15:50:18 <elliott> Then I stop being such a dithering prick and set light to the portal."
15:50:39 <elliott> "I’m almost ready. The only thing wrong with this scene is that it doesn’t feel very dramatic. I’m going to *hell*, there should be some spectacle here.
15:50:39 <elliott> I turn around and look at the closest tree for a while, then set fire to it. It starts a forest fire that rages across the hill behind me, enclosing the portal between a crescent of fire and the water of the bay."
15:50:45 <elliott> I, too, solve all problems with forest fires.
15:52:11 <elliott> "This is when I discover the rock beneath me is not in fact rock, but a sea of screaming faces that stick to my feet in such a way that I can barely move. Shit like this is going in my TripAdvisor review, Nether. Two stars MAX."
15:52:59 <elliott> "I bolt back down as the fireball hits above. OK, I saw it. It’s about three blocks this way, then five that way, then a couple up and whunk! I’ve struck obsidian. Congratulations, Tom, you’ve discovered the thing you just came out of."
16:10:55 <elliott> fizzie: [[The Settlers 7: Paths to a Kingdom
16:10:55 <elliott> Super horrible DRM problems aside, this is a fresh air of old school game design, almost feeling board-game like at times. 4/5]]
16:11:05 <elliott> fizzie: Notch -- he hates stupid horrible DRM problems.
16:11:12 <elliott> Like, say, depending on a single server to play the game.
16:11:25 <elliott> And bundling forced updates with the validation.
16:16:18 <elliott> [[Sonic the Hedgehog 4: Episode 1
16:16:18 <elliott> Too little, too late. Sonic in an undead zombie by now. 2/5]]
16:16:26 <elliott> Coming soon to Minecraft -- zombies that aren't undead.
16:19:16 -!- pikhq has quit (Ping timeout: 255 seconds).
16:21:11 -!- pikhq has joined.
16:23:56 <elliott> ineiros: FWIW, there are supposedly-stable pre-compiled builds of the beta-updated development hMod.
16:24:40 <Vorpal> <fizzie> Someone has done a google-maps driven map already, though, http://www.triangularpixels.com/Junk/TectonicusTest/map.html <-- I seen at least two different programs that does it
16:25:45 <Vorpal> ineiros, when will the server go back to normal btw?
16:25:54 <Vorpal> I would guess when hmod is updated
16:31:01 <elliott> The maps would be cleaner. :p
16:31:08 <elliott> I'm even getting used to dying! A LOT!
16:34:55 <fizzie> elliott: Based on the hmod github commit-log, it's still seeing quite a lot of development, so I'd personally still wait a while. (Though of course for bleeding-edge testing...)
16:35:18 <elliott> fizzie: Isn't that the motto of the League of Notch Apologists? "YOU'RE the testers!"
16:38:43 <Vorpal> elliott, that is only because haven't built anything really complicated. Sure the stairs were massive, but they didn't actually take that long. Something like fizzie's house and bunker is a lot more work
16:39:06 <Vorpal> so yeah, we should go back to the proper map when we have a working hmod
16:39:30 <elliott> Vorpal: You could just say "MY HOUSE!! MY BEAUTIFUL BEAUTIFUL FORTRESS!!" and be done with it rather than coming up with complicated excuses.
16:41:33 <elliott> Deewiant: Well, no, he said "You wouldn't know! You're not like fizzie!"
16:41:37 <elliott> <Vorpal> elliott, that is only because haven't built anything really complicated. Sure the stairs were massive, but they didn't actually take that long. Something like fizzie's house and bunker is a lot more work
16:41:38 <elliott> <Vorpal> so yeah, we should go back to the proper map when we have a working hmod
16:41:45 <elliott> <Vorpal> I don't want that, I'd lose my house.
16:42:21 <elliott> YES! k3 is still available!
16:44:52 <Vorpal> elliott, in other words, you got annoyed because I didn't fit your baseless image of me as someone who constantly whines when I tried to discuss in a civil manner and with proper justification for the statement.
16:45:07 <elliott> Oh snap, time for some quality Vorpal indignation.
16:45:26 <Vorpal> elliott, actually I'm amused.
16:55:57 <elliott> elliott@dinky:~/k$ chmod +x lin/k
16:55:57 <elliott> K 3.2 2005-06-25 Copyright (C) 1993-2004 Kx Systems
16:55:57 <elliott> LIN32 2CPU 3805MB dinky.local 0 EVAL
17:05:58 <elliott> K, the language, version 3.
17:06:43 <elliott> and no. in fact i just found the interp against after olegfink linked me to nsl.com's copy
17:06:46 <Vorpal> elliott, I don't remember the details no
17:06:58 <elliott> it's like J but different.
17:16:25 <elliott> "K-Lite is a time-limited, reduced version of K which enables interested develop-
17:16:25 <elliott> ers to learn the language and develop small applications. K-Lite consists of the K
17:16:25 <elliott> language and interpreter, GUI software, and ASCII file read/write capability. It
17:16:25 <elliott> does not include connections, file mapping, interprocess communications or
17:16:25 <elliott> runtime capabilities." <-- hmm, this copy doesn't seem to be time limited
17:17:45 <elliott> whereas this is real K 3.2
17:18:40 -!- jix has quit (Read error: No route to host).
17:21:39 -!- jix has joined.
17:22:10 <elliott> hmm no it is the EVALuation version I think since it says EVAL
17:23:38 <elliott> "For example, Logical Or is denoted by | and Logical And is denoted by & in both K and C."
17:49:56 -!- elliott has quit (Remote host closed the connection).
17:52:14 -!- elliott has joined.
17:57:03 -!- cheater99 has quit (Quit: Leaving).
17:57:04 -!- jix has quit (Ping timeout: 276 seconds).
17:57:30 -!- jix has joined.
17:59:31 * pikhq now wonders how quickly it can take to go from 0 to the first boss in here.
17:59:44 <pikhq> Come on, absurdity.
18:03:38 <pikhq> All of a few minutes?
18:04:43 <pikhq> I'm just guessing.
18:05:19 <pikhq> Okay, bit more than that.
18:06:12 <pikhq> First hour or so skipped though.
18:06:44 <elliott> pikhq: are you doing the infinity-plus-one-all-that's-necessary run?
18:11:45 <pikhq> Speedrunning is easy with most of the game being optional.
18:12:25 <elliott> pikhq: Is this going to give you a terrible ending? :P
18:12:30 <Vorpal> text adventure speed run. How would you measure it?
18:12:38 <Vorpal> in number of commands?
18:12:45 <pikhq> elliott: Not necessarily.
18:12:53 <elliott> Vorpal: Yes, that's how it's done.
18:13:01 <pikhq> elliott: But the good ending would be PAINFUL.
18:13:08 <elliott> Vorpal: Take a look at this: http://www.the-spoiler.com/ADVENTURE/Infocom/trinity.1.html
18:13:56 <pikhq> elliott: There's a 260 floor dungeon. It is only mandatory at the very end of the good ending.
18:15:02 <elliott> pikhq: Can you one-shot it? :P
18:15:17 <pikhq> I can one-shot everything!
18:15:32 <elliott> pikhq: Can you do the bad ending, then go back and do the good ending?
18:15:47 <elliott> If so: do that, just so you can say you did the ABSOLUTE MINIMUM REQUIRED TO COMPLETE IT the first time.
18:15:49 <pikhq> I can load from a savepoint, yes.
18:16:04 <elliott> Save right before the bad/good diverge point :P
18:16:21 <pikhq> There's a few minutes from there to the bad ending.
18:17:12 <pikhq> You choose whether to accept that the world will end or try to fight Death.
18:28:30 -!- j-invariant has joined.
18:29:03 <elliott> http://en.wikipedia.org/wiki/Computable_number#Formal_definition
18:29:09 <elliott> defining computable reals by computable dedekind cuts
18:30:24 <elliott> (ok so that has very little to do with coq but still)
18:31:26 <j-invariant> that's so simple, I had always thought dedekind cuts would be really complicated to do
18:34:02 -!- roper has quit (Quit: Abandonando).
18:35:09 <elliott> j-invariant: I wonder what pi looks like
18:35:32 <pikhq> Okay, seems it takes 45 minutes to get through the intro stuff.
18:36:03 <elliott> fizzie: Not going to play on the HARDCORE SERVER?
18:36:10 <pikhq> And suddenly... SATAN!
18:37:21 <elliott> j-invariant: still, it seems simpler than the complicated usual approximation-function definition
18:37:27 <elliott> j-invariant: and also the continued fraction definition
18:38:16 <elliott> fizzie: I'm spending the night in a pitch black 1x1x2 hole with things making noises right next to me.
18:42:24 <Vorpal> <pikhq> elliott: There's a 260 floor dungeon. It is only mandatory at the very end of the good ending. <-- large floors?
18:42:36 -!- Sasha has joined.
18:43:20 <elliott> j-invariant: hmm i should implement computable dedekind cuts
18:43:24 <tswett> If it's pitch-black hole, mobs can spawn in it.
18:43:35 <elliott> tswett: As I said, it's 1x1. So they'd have to spawn... on me.
18:43:41 <elliott> j-invariant: me neither ...
18:43:54 <elliott> tswett: Technically, they have .3 m above my head to cram into.
18:44:17 <elliott> Aaand the server went down.
18:44:23 <tswett> Hm. Perhaps you and them are both really narrow, or something. I don't know if they can spawn there or not.
18:44:39 <Vorpal> pikhq, if they are large, is it randomly generated or something (or at least generated + hand edited, actually making that many floors, if large, by hand sounds painful)
18:45:52 <tswett> You know, I would probably get some enjoyment out of a game that is just a randomly-generated world that you can explore.
18:45:59 <tswett> Do nothing else. Just explore.
18:51:42 <Vorpal> elliott, local game: larger overhang than mt hoover. Also it turns 90 degrees halfway out. And there is no scenery cut off to explain it (that game is all post-halloween)
18:52:04 <pikhq> Vorpal: It is mostly randomly-generated.
18:52:37 <pikhq> Vorpal: And changes each time you go to a floor.
18:52:39 <Vorpal> pikhq, I guess it isn't very story-heavy then?
18:53:02 <pikhq> Oh, it's very story-heavy. The design of the single dungeon is just irrelevant to the story.
18:53:22 <Vorpal> elliott, wow there are 2 more overhangs like it visible from under the middle one
18:53:36 <pikhq> And aside from the very end of the good ending, it's entirely optional.
18:53:38 <Vorpal> elliott, and a rather large floating island
18:54:00 <Vorpal> pikhq, is there just /one/ good ending?
18:54:25 <Vorpal> pikhq, also I didn't mean if the game was story-heavy. I meant if the gameplay in that dungeon was
18:54:33 <elliott> This house should do me for the night.
18:54:39 <elliott> Although a creeper will fuck it up.
18:54:42 <elliott> Shit, I only have three hearts.
18:54:48 <elliott> How do you get health back again? It isn't coming back naturaly.
18:56:53 <elliott> j-invariant: I play it on Debian
18:57:00 <j-invariant> just wondering if you have to do any fixes to make it work?
18:57:09 <Deewiant> elliott: So you won't get health back.
18:57:32 <elliott> Deewiant: how do i get food again :D
18:57:34 <elliott> j-invariant: well you need the sun jvm
18:57:41 <elliott> or at least, that fixes it for some people
18:57:45 <elliott> j-invariant: what problems did you have?
18:57:59 <Deewiant> elliott: Mushrooms, fish, pigs, apples
18:58:08 <tswett> How do you get apples?
18:58:19 <tswett> Also, some mushrooms will hurt you, if I remember correctly.
18:58:34 <fizzie> I play on Ubuntu with the default openjdk, and it works just fine for me, but indeed for many it doesn't seem to.
18:58:54 <j-invariant> I think the graphics rendenring was completely wrong, but maybe I should try again with the upgrade
19:00:02 <elliott> Suspicious lack of noises tonight.
19:00:07 <elliott> Deewiant: Gonna play on the survival server?
19:00:17 <elliott> Deewiant: It's all ~hardcore~ now although it is night time so you may want to wait some minutes.
19:00:37 <pikhq> elliott: Fun fact: you only get the infinity-plus-one sword after the first boss.
19:00:39 <Vorpal> elliott, an exposed dungeon up in the very tip of the overhang!!
19:00:48 <pikhq> elliott: Said boss is entirely beatable without leveling.
19:01:29 <pikhq> elliott: It's essentially scripted up until then.
19:02:03 <pikhq> Vorpal: There's precisely two endings.
19:02:24 <elliott> j-invariant: how do you do "apply recordConstructor." in Coq without it complaining about not being able to find a value for a variable?
19:02:27 <elliott> I want it to become the new goal
19:03:25 <elliott> Aww, a zombie swimming, how cute.
19:03:34 <elliott> Wait, what? It's LIGHT now, why are you appaering now.
19:07:44 -!- wareya_ has quit (Ping timeout: 264 seconds).
19:09:27 <j-invariant> I have mn = id, nm = id, etc.. and I want to show that Ff = m Gf v
19:12:47 <fizzie> Yes, I foolishly tried out mcmap.
19:12:59 <elliott> j-invariant: eapply has me trying to prove things on ?7
19:13:04 <elliott> rather than getting me to specify ?7 first
19:13:08 <elliott> or is there a tactic for that too?
19:13:12 <j-invariant> elliott: yeah if that happens you are in trouble
19:13:16 <elliott> fizzie: can you please bring bacon? :{
19:13:32 <j-invariant> say you have category with identity compose and some axioms..
19:13:33 <elliott> j-invariant: hmm there's a command to declare a subgoal with a specified type right?
19:13:58 <j-invariant> then a good way is instead of eapply Build_Category. you can do apply (Build_Category <definition of identity> <definition of compose>). tehn you get goals left over for all the axioms
19:14:13 <fizzie> Strange stuff; mcmap goes into a 100% CPU utilization loop.
19:14:21 <j-invariant> but it can be difficult to write the definitions in situ
19:14:27 <elliott> fizzie: you sure do listen to my pleas :P
19:14:36 <fizzie> I don't have any bacon to bring.
19:14:45 <elliott> fizzie: there's a mob spawner at spawn.
19:15:08 <fizzie> Well, as soon as I get mcmap to actually let me in.
19:15:17 <j-invariant> elliott: I'mk stating to think that teh theorem isn't true
19:17:34 <elliott> fff now I need to refer to Z's ^
19:17:40 <elliott> and i've forgotten how to override scope
19:18:14 -!- Wamanuz4 has quit (Remote host closed the connection).
19:22:14 <elliott> fizzie: oh god spider noises
19:23:15 -!- Wamanuz has joined.
19:24:01 <Vorpal> elliott, I haven't had those in single player, but I have had strange noises
19:26:16 <j-invariant> elliott: the proofs are kind of a mess, and it's been hard work - but I got equality of functors defined
19:26:29 -!- wareya has joined.
19:27:52 <j-invariant> elliott: (if there is a neater way to define functor equality, I'll take it!)
19:30:38 <elliott> j-invariant: i don't know of one, way further than i've got
19:31:00 <fizzie> I threw out three bacon, I don't know if they disapparated somewhere. Hope not.
19:31:19 <elliott> fizzie: They were in my inventory when I spawned falling into nothingness.
19:31:41 <elliott> fizzie: You can stay at my house, FWIW. If you're staying on that is.
19:31:41 <fizzie> I'm not going to start actually playing there, though, I just wanted to check if mcmap worked.
19:31:55 <elliott> fizzie: See how boring you are?!
19:37:49 <j-invariant> elliott: want to do natural transforms for me? LOL this is making me exausted
19:38:15 <elliott> j-invariant: er let me think of an excuse :D
19:38:22 <j-invariant> its' probably going to be another 300 lines
19:38:42 <elliott> j-invariant: and they said Coq couldn't get any more academic!
19:39:22 <j-invariant> I really do want to define limits via cones though
19:40:33 <elliott> how do you do that if p's ^ is in Z_scope
19:40:36 <elliott> but q's ^ is in positive_scope?
19:41:40 <elliott> j-invariant: well no q is positive_scop
19:44:24 <elliott> (Z_of_nat (nat_of_P (3*(q^3))%positive))
19:45:31 -!- sebbu has quit (Read error: Connection reset by peer).
19:45:58 -!- sebbu has joined.
19:49:52 <elliott> Zle_bool (p ^ 3) (let q' := Z_of_nat (nat_of_P q) in 3 * q' ^ 3)
19:51:19 <elliott> what's the tactic that proves True again :D
19:53:42 <elliott> ============================
19:53:43 <elliott> Zle_bool (p ^ 3) (let q' := Z_of_nat (nat_of_P q) in 3 * q' ^ 3)) /\
19:53:53 <elliott> Zle_bool (p ^ 3) (let q' := Z_of_nat (nat_of_P q) in 3 * q' ^ 3)) ->
19:55:08 <elliott> it's just if x then True else False
19:55:16 <elliott> j-invariant: hmm i'd be better off using match rather than destructing with Qnum and Qden right?
19:55:48 <Vorpal> elliott, http://sporksirc.net/~anmaster/minecraft/screenshots/overhang/
19:55:52 <Vorpal> elliott, that beats mt hoover eh?
19:55:54 <j-invariant> if you destruct s it should simplify to the same thing
19:56:06 <elliott> gah, i've forgotten all the tactics
19:56:29 <j-invariant> yeah I wish there was another way to twirte proof, some sort of hand waving magic way
19:56:57 <Vorpal> elliott, take http://sporksirc.net/~anmaster/minecraft/screenshots/overhang/2010-12-23_19.57.42.png for example. That are two huge overhangs next to each other, not a full arch
19:57:00 <elliott> j-invariant: yeah like writing some stuff on paper, saying "the rest is left as an exercise to the reader", and calling it a day
19:57:05 <elliott> j-invariant: why do we need computers to do it anyway!
19:57:41 <Vorpal> elliott, the rest of the pictures in that dir are quite nice too, showing off the THIRD huge overhang which is hidden in the picture I linked.
19:57:58 <elliott> j-invariant: I think that maybe "x = tt" is easier to work with than "if x then True else False"
19:58:49 <Vorpal> elliott, you can't get up on one of them because it overhangs all around
19:59:00 <Vorpal> elliott, the other two (including the largest one) you can get up on just fine
19:59:28 <elliott> (if Zle_bool (p1 ^ 3) (3 * Z_of_nat (nat_of_P q1) ^ 3)
19:59:29 <elliott> (if Zle_bool (p2 ^ 3) (3 * Z_of_nat (nat_of_P q2) ^ 3)
19:59:32 <elliott> else False) -> p1 # q1 < p2 # q2
19:59:38 <elliott> "do arithmetic yourself computer."
20:00:16 <elliott> Error: Omega: Unrecognized predicate or connective: Qlt
20:01:44 <elliott> j-invariant: wanna do my proof for me? :p
20:01:54 <Vorpal> elliott, there is an image on the improbable dungeon inside the overhang too
20:02:57 <j-invariant> elliott: wait a second omega should handle Qle
20:03:11 <elliott> j-invariant: a second omega? it wouldn't even run once, because of that
20:03:25 <elliott> j-invariant: what upsets me is that it can't match (Is_true (... Zle_bool ...)) to Zle ...
20:03:32 <elliott> j-invariant: oh i have an idea
20:03:49 <elliott> D_decide : forall x, (D x) \/ ~(D x)
20:03:58 <elliott> that'd give less Is_true crap right?
20:07:28 <Vorpal> fizzie, you might want to check out http://sporksirc.net/~anmaster/minecraft/screenshots/overhang/ too
20:09:02 <elliott> ============================
20:09:03 <elliott> (1 ^ 3 <= 3 * Z_of_nat (nat_of_P 1) ^ 3)%Z
20:09:05 <elliott> that's decidable you stupid prover
20:11:50 <Vorpal> elliott, is this awesome: http://sporksirc.net/~anmaster/minecraft/screenshots/overhang/2010-12-23_19.59.38.png
20:11:57 <Vorpal> elliott, (see the dungeon!)
20:12:09 <elliott> darn now what's the thing to do ~(x=y)
20:12:38 <elliott> (Gt = Gt -> False) -> False
20:12:48 <elliott> j-invariant: look at that filthy coq, with its ~~p!
20:13:12 <elliott> congruence sorts it out though :P
20:13:19 <elliott> | p # q => (p ^ 3 <= 3 * Z_of_nat (nat_of_P q) ^ 3)%Z
20:13:20 <elliott> | p # q => (p ^ 3 <= 3 * Z_of_nat (nat_of_P q) ^ 3)%Z
20:13:33 <elliott> Error: Omega: Unrecognized predicate or connective: Qlt
20:14:01 <elliott> Vorpal: a creeper exploded one block from my house :)
20:14:36 <Vorpal> elliott, was it obsidian?
20:14:44 <elliott> no, wood, but as i said, one block outside of blast radius
20:14:47 <elliott> not the creeper one block away
20:14:50 <elliott> to be fair, i hit it with my sword
20:15:08 <Vorpal> elliott, scared to death
20:15:27 <Vorpal> elliott, yes, it dies when it explodes
20:15:55 <elliott> j-invariant: compute on this was a bad idea
20:16:23 -!- augur has quit (Remote host closed the connection).
20:16:37 <elliott> j-invariant: are you sure omega is meant to do Q?
20:19:05 <j-invariant> Coq version 8.3 is before all a transition version with refinements or extensions of the existing features and libraries and a new tactic nsatz based on Hilbert’s Nullstellensatz for deciding systems of equations over rings. <--- interesting
20:19:15 <elliott> ============================
20:19:15 <elliott> (p1 ^ 3 <= 3 * Z_of_nat (nat_of_P q1) ^ 3)%Z /\
20:19:17 <elliott> ~ (p2 ^ 3 <= 3 * Z_of_nat (nat_of_P q2) ^ 3)%Z ->
20:19:21 <elliott> induction here doesn't look promising
20:19:28 <elliott> the conversions make things more painful here :/
20:19:55 <elliott> wish it was Q := Z*(q:Z)*(q>0)
20:20:10 <elliott> j-invariant: is there a premade lemma for that? :P
20:20:23 <j-invariant> then you just need p1 * q2 < p2 * q1 iff p1 # q1 < p2
20:21:16 <elliott> Notation QDen p := (Zpos (Qden p)).
20:22:12 <elliott> H0 : (p1 ^ 3 <= 3 * ' q1 ^ 3)%Z
20:22:12 <elliott> H1 : ~ (p2 ^ 3 <= 3 * ' q2 ^ 3)%Z
20:22:12 <elliott> ============================
20:22:15 <elliott> looks less frightening now
20:23:16 <elliott> j-invariant: lame, stdlib doesn't have that theorem
20:23:42 -!- augur has joined.
20:26:15 <elliott> forall (p1 p2 : Z) (q1 q2 : positive),
20:26:15 <elliott> (p1*'q2 < p2*'q1)%Z -> (p1#q1 < p2#q2)%Q.
20:26:18 <elliott> that was surprisingly painless
20:26:59 <elliott> H1 : ~ (p2 ^ 3 <= 3 * ' q2 ^ 3)%Z
20:27:01 <elliott> just need to get this to be >
20:28:28 <elliott> Lemma Znot_ge_lt : forall n m:Z, ~ n >= m -> n < m.
20:29:35 -!- impomatic has joined.
20:29:47 <elliott> set (H2 := Znot_le_gt (p2^3) (3*'q2^3) H1).
20:29:50 <elliott> j-invariant: is there a nicer way to do this?
20:30:14 <elliott> in fact that doesn't even work
20:30:18 <j-invariant> dunno, I tend to build everything from scratch
20:30:41 <elliott> j-invariant: I thought I wanted "rewrite Znot_le_gt in H1" but that doesn't work
20:30:47 <elliott> H1 : ~ (p2 ^ 3 <= 3 * ' q2 ^ 3)%Z
20:30:53 <elliott> H1 : (p2 ^ 3 > 3 * ' q2 ^ 3)%Z
20:30:58 <elliott> Lemma Znot_le_gt : forall n m:Z, ~ n <= m -> n > m.
20:31:00 <j-invariant> see if you can rwewrite all the Q stuff in Z, then use omega for it
20:31:04 <elliott> rewrite Znot_le_gt in H1 doesn't work
20:31:07 <elliott> j-invariant: yeah that's what i've done
20:31:18 <elliott> so I still need to get this rewrite out of the way
20:32:20 <elliott> j-invariant: want the file
20:32:49 <elliott> j-invariant: http://sprunge.us/BbLA
20:33:47 <elliott> j-invariant: I think Coq would be better if all the propositions weren't named... so instead of H1 and the like, there were just the types
20:33:57 <elliott> and also all the proofs in Cut would just be the type with nothing before :
20:34:03 <elliott> it's hard to figure out names :P
20:37:38 <elliott> j-invariant: well that's a forall :D
20:37:40 <elliott> j-invariant: just intro it
20:38:42 <elliott> j-invariant: um i should hope not, since it's the example on wikipedia
20:38:46 <elliott> for a computable dedekind cut
20:39:08 <elliott> j-invariant: so it must be true, by wikipedian infallibility
20:39:18 <elliott> j-invariant: i think ic an hack the proof if youc an make "rewrite Znot_le_gt in H1" work :P
20:42:39 <j-invariant> tested it in haskell and found no counter example
20:43:24 <elliott> lol, when coq users can't prove something they do it in haskell instead
20:44:29 <Quadrescence> i was writing an article once on near-integers, modular forms, and the j-invariant
20:45:40 <elliott> D_true_ex_gt : forall r, D r -> (exists s, s > r /\ D s)
20:45:46 <elliott> j-invariant: this looks unfun
20:45:53 <elliott> (I "admit"ted the previous one and am moving on for now)
20:45:58 <elliott> | p # q => (p ^ 3 <= 3 * ' q ^ 3)%Z
20:45:58 <elliott> ============================
20:46:02 <elliott> | p # q => (p ^ 3 <= 3 * ' q ^ 3)%Z
20:46:11 -!- oerjan has quit (Quit: Good night).
20:46:25 <impomatic> Weather.Outside="frightful"; Fire.Delightful=true; Lights.Luminosity=WayDownLow; for (int i=1; i<=3; i++) { LetItSnow(); }
20:46:47 <Vorpal> elliott, can't you encode the statement in SAT or SMT and try an automated prover (iirc coq is interactive?)
20:47:05 <elliott> Vorpal: that's a rather ... generic statement
20:47:16 <elliott> coq has a ton of automatic solvers, none of them want to touch this :)
20:47:43 <Vorpal> elliott, I was thinking about using something like alt-ergo or cvc3
20:47:45 <elliott> step one is finding an s that satisfies
20:48:07 <elliott> j-invariant: any ideas about what s to try? :p
20:48:42 <Vorpal> elliott, what exactly are you trying to prove? I don't know co
20:49:15 <elliott> Vorpal: i'm trying to prove that D(p/q) = p^3 <= 3*(q^3) is a (computable) dedekind cut
20:49:24 <elliott> proving that it's the cube root of three is for after i have arithmetic. :p
20:49:39 <elliott> Vorpal: i.e. Yet Another approach to the computable reals
20:49:50 <Vorpal> elliott, I guess I can't help until I first check what the heck dedekind cut is
20:50:01 <elliott> Vorpal: you ... don't know?
20:50:07 <Vorpal> elliott, can't say I do no
20:50:24 <elliott> Vorpal: http://en.wikipedia.org/wiki/Dedekind_cut
20:50:28 <elliott> they're the "usual" definition of the reals
20:50:45 <elliott> Vorpal: the computable version is described in http://en.wikipedia.org/wiki/Computable_number#Formal_definition starting "There is another equivalent definition"
20:50:48 <Vorpal> oh right, the definition is familiar
20:50:55 <Vorpal> elliott, but I don't think I ever heard it's name
20:51:21 <elliott> Deewiant: well, i'd say so
20:51:23 <Vorpal> Deewiant, well not sure about that.
20:51:32 <elliott> Deewiant: dedekind cuts and cauchy sequences are the main ways
20:51:32 <Quadrescence> elliott: that's not an approach to computable reals
20:51:39 <elliott> Quadrescence: what aren't?
20:51:45 <Deewiant> I'd say the Cauchy sequence definition is far more common
20:51:54 <Vorpal> elliott, the second one really cauched on!
20:52:06 <elliott> Quadrescence: that's both unhelpful and unjustified
20:52:06 <Vorpal> (sorry, but oerjan wasn't here)
20:52:25 <Quadrescence> elliott: if you want computable reals, you have to start talking about bounds and turing machines
20:52:40 <elliott> Quadrescence: I know what computable reals are. There are several approaches.
20:52:56 <elliott> e.g. approximation functions, continued fractions, and this one, computable dedekind cuts
20:53:25 <Quadrescence> "computable dedekind cuts" sounds like a terrible idea
20:54:04 <elliott> Quadrescence: on what grounds?
20:54:11 <elliott> j-invariant: reals that are computable :-)
20:54:21 <elliott> Vorpal: clever use of shadows to give 32-colour images in minecraft: http://www.minecraftwiki.net/images/c/c4/Dither2.jpg
20:54:33 <Quadrescence> j-invariant: a number which there is a finite process for computing the digits of to arbitrary precision
20:54:35 <Slereah> Markov made a nice article on that
20:54:37 <elliott> "In mathematics, particularly theoretical computer science and mathematical logic, the computable numbers, also known as the recursive numbers or the computable reals, are the real numbers that can be computed to within any desired precision by a finite, terminating algorithm. Equivalent definitions can be given using μ-recursive functions, Turing machines or λ-calculus as the formal representation of algorithms."
20:54:42 <elliott> there are tons of definitions i'd say
20:54:46 <elliott> but they're all equivalent. probably
20:55:08 <Slereah> A computable number was basically just defined as (f(x) - g(y))/h(z)
20:55:18 <Slereah> Where all three functions are computable processes
20:56:57 -!- augur has changed nick to YeshuaLord.
20:57:19 -!- YeshuaLord has changed nick to augur.
20:57:55 <elliott> ineiros: Can I have some bacon?
21:00:34 <Quadrescence> also Deewiant is certainly right that the cauchy seq. defn is the usual defn of the reals
21:00:45 <elliott> sure, but dedekind cuts aren't actually niche.
21:01:14 <j-invariant> I always thought I was alone with my cauchy sequences
21:01:44 <elliott> http://en.wikipedia.org/wiki/Construction_of_the_real_numbers damn, wp lists dedekind second
21:01:47 <Slereah> The whole set of reals is useless!
21:01:51 <elliott> quick, let's edit it to prove Quadrescence wrong
21:02:05 <Slereah> Aleph one of it is basically not found in nature!
21:02:26 <Quadrescence> elliott: No, my evidence is from reading books
21:02:36 <elliott> wikipedia disproves all books, obviously
21:03:27 <Slereah> Cauchy sequence may be first because it's the most intuitive?
21:05:08 <pikhq> 3 math classes. :D
21:14:40 <elliott> Vorpal: notch should reintroduce holes in the bedrock, except make them correspond to holes in the nether ceiling
21:20:41 <elliott> Vorpal: :( he's going to add degrade-on-hit
21:21:02 <elliott> Vorpal: there's a creeper jumping outside my door
21:25:21 <Vorpal> <elliott> Vorpal: :( he's going to add degrade-on-hit
21:25:25 <Vorpal> elliott, what do you mean
21:25:39 <elliott> Vorpal: as in, your pick can break midway through mining obsidian
21:25:42 <elliott> because it's every hit that degrades it
21:26:09 <Vorpal> elliott, so it won't last as long as before?
21:26:25 <elliott> also, apparently block damage will last for like 10 seconds after you stop mining
21:26:30 <elliott> so you have time to change to a new pick
21:26:44 <Vorpal> elliott, not as bad then as it could be
21:27:11 <elliott> CREEPER JUST TO THE RIGHT OF MY DOOR HOLY FUCKING SHIT
21:27:21 <Vorpal> elliott, change to peaceful?
21:27:55 <elliott> waiting for someone to log in, kill some pigs, and share the bacon with me; I'll provide lodging in my house in return
21:27:58 <Vorpal> elliott, well I'm waiting for the old world. Not really worth building anything in the current world
21:27:59 -!- impomatic has left (?).
21:28:12 <elliott> Vorpal: no reason this world can't be brought out occasionally
21:28:19 <elliott> it's fun, i've already built a tall house
21:28:28 <elliott> and will mine once i can get coal for torches
21:28:31 <Vorpal> elliott, see, I like predictability
21:29:30 <Vorpal> elliott, "logging on thinking you will work on whatever your current project is. Huh, what, not the usual world?!"
21:29:43 <elliott> Vorpal: Or we could have it on predictable days ...
21:29:48 <elliott> Say every two weeks or something.
21:30:05 <elliott> If you're going to say "no just trash it permanently": I built a house I like too, you know.
21:30:14 <elliott> Except mine's in the survival world.
21:30:35 <Vorpal> elliott, well, donate to ineiros so he can get a good VPS to run all the servers on side by side
21:30:44 <elliott> Or just have the survival world come out at predictable times.
21:30:59 <Vorpal> elliott, but yes I think we should switch world whenever you feel like working on glass cube
21:31:12 <Vorpal> that sounds good, switch away from the glass cube one then
21:31:48 <elliott> Vorpal: Are you just trying to be a dick, or are you really so dense that you'll demand that the peaceful world is kept constantly because of ~YOUR BEAUTIFUL HOUSE~ while completely disregarding the fact that I've built things in the survival world, too, and plan to continue doing so?
21:32:22 <Vorpal> elliott, but you are the only one who built stuff there
21:32:26 <Vorpal> elliott, you against all the other ones?
21:32:47 <elliott> Vorpal: I didn't say "trash the peaceful world".
21:32:50 <elliott> I said "have the survival world occasionally".
21:33:05 <elliott> Besides, the beta update for hMod won't be out for a little while, and this is only the first day of the survival server.
21:33:14 <Vorpal> elliott, yeah, maybe one day per moth
21:33:24 <Vorpal> elliott, or whenever hmod breaks due to upgrade
21:33:41 <elliott> Vorpal: I agree, let's only have your house when hMod breaks.
21:34:25 <Vorpal> elliott, ... the other way around
21:34:42 <Vorpal> elliott, but I predict it will break fairly often
21:34:43 <elliott> Ahh, I see; your house is worth more because you made it.
21:34:45 <elliott> Your arguments, they are stellar.
21:34:51 <Vorpal> elliott, no it is the hmod world.
21:35:05 <Vorpal> elliott, further, it also happens to contain a LOT more
21:35:19 <Vorpal> elliott, fizzie, nailor, ineiros, PH and so on
21:35:36 <Vorpal> elliott, so you hate the "wonder's of the world" thing.
21:35:58 <Vorpal> (if you strawman me, then you get the same shit back)
21:39:25 -!- zzo38 has joined.
21:39:48 <Vorpal> elliott, or you could just switch topic...
21:40:22 <elliott> oh, not my dedekind cuts :P
21:42:08 <elliott> j-invariant: have you seen http://coq.inria.fr/stdlib/Coq.Logic.IndefiniteDescription.html?
21:43:21 <elliott> ineiros: I'll bribe you for bacon. :p
21:43:46 <Gregor> elliott: In retrospect, cunionfs is awesome :P
21:44:25 <elliott> Gregor: What's so awesome about it? :p
21:44:44 <zzo38> I should make a program to copy a CD/DVD with error emulation and other options. In case of error, it makes a error emulation file that the driver will emulate errors when reading the copy.
21:45:57 <elliott> Gregor: I'm considering writing a package manager based on it, please tell me it actually works well >_>
21:46:26 <Gregor> elliott: Oh it works fine, that's never been the issue, it just doesn't do much (it's just a per-process union FS)
21:46:32 <zzo38> Also a function to tell it to copy the entire disc even if it says only part of it has data, try copying everything anyways even if it says nothing there.
21:47:30 <Gregor> elliott: But what I imagine it's most useful for is building packages in a constrained environment where you don't want "smart" configuration dragging in dev packages you didn't want.
21:48:28 <Gregor> elliott: As well as, of course, "enterprise"y environments like I said before.
21:48:32 <elliott> <Gregor> elliott: But what I imagine it's most useful for is building packages in a constrained environment where you don't want "smart" configuration dragging in dev packages you didn't want.
21:48:36 <elliott> ITT: Debian and Fedora both have this
21:49:10 <Gregor> Yeah, but they do that in a fairly-bizarre way, this is just "let me see this, OK now they're in /usr"
21:49:24 <elliott> Gregor: So can you mount / as a cuneiformfs? :P
21:50:00 <Gregor> I was thinking about that, and my answer is yes-and-no. You could, if you're willing to do that first-thing then do (most) everything else under chroot.
21:50:46 <Gregor> The only real problem is that the directories you mount under it need to exist, but for them to exist they need to be in another FS (to be unioned in)
21:52:18 <elliott> <Gregor> I was thinking about that, and my answer is yes-and-no. You could, if you're willing to do that first-thing then do (most) everything else under chroot.
21:52:27 <elliott> it's what initramfses do IIRC
21:52:38 <elliott> and it pulverises your existing / :P
21:52:45 <Gregor> That doesn't resolve the issue I mentioned later.
21:53:06 <Gregor> The directories you're going to mount /other/ filesystems to need to exist, but for them to exist, they need to ... already exist :P
21:54:22 <pikhq> To have two degrees instead of 1, I need: 1 extra class.
21:56:18 <j-invariant> elliott: http://coq.pastebin.com/xnGjabR6 line 481
21:57:26 <j-invariant> that stuff is awful but I don't know a better way
21:57:49 <elliott> j-invariant: i like it, i also think that it will never work as an stdlib :D
21:58:34 <j-invariant> elliott: I'm going to try to define universal cones so I can get things like products
21:59:45 <elliott> Gregor: For a stow-alike, are there any advantages to cunionfs over unionfs? :p
22:00:08 <elliott> j-invariant: well it might with enough layering
22:00:16 <elliott> j-invariant: i just mean, if you have to build categories like that all the time
22:00:48 <j-invariant> elliott: I think (but I have to study this first) you can define a 'type theory' category, and just use that for everything
22:00:50 <Gregor> elliott: Yes, every process can see its own stowish environment, giving it the build advantage I mentioned before, as well as the ability to have conflicting packages installed simultaneously, etc. Whether that's an advantage to you depends on what you want *shrugs*
22:01:03 <elliott> Gregor: That's an sps-alike, not a stow-alike :P
22:01:15 <elliott> Gregor: But yeah, compelling.
22:01:22 <j-invariant> elliott: but that might be like defining a self interpreter for coq so..
22:01:25 <elliott> Gregor: Now guarantee that cunionfs is stable enough to use in Kitten :P
22:01:26 <Gregor> elliott: For a stow-alike, are there any advantages to unionfs over ... stow?
22:01:34 <elliott> j-invariant: also it'd be rather abstract
22:01:37 <Gregor> elliott: I can't guarantee anything, it's F/OSS :P
22:01:38 <elliott> j-invariant: like coding ASTs manually
22:01:43 <elliott> Gregor: stow uses ... symlinks.
22:01:46 <elliott> Ask pikhq for the gory details.
22:01:57 <elliott> Note: Symlinks are pretty much the devil.
22:02:15 <elliott> Gregor: Well, half-guarantee then :P
22:02:15 <j-invariant> I would like to implement something like Knuth Bendix completion to automatically prove theorems
22:02:35 <pikhq> Gregor: GNU stow does nothing more than looking in the stow directory and checking for files that aren't symlinked into the path.
22:02:55 <Gregor> elliott: I have fair confidence that cunionfs is sufficiently stable, and that any stability issues are sufficiently minor that they could be fixed in short order.
22:03:10 <pikhq> It's not *terrible*, but it's only one step removed from Slackware's "untar things to root".
22:03:41 <Gregor> pikhq: Well, it claims to be a package manager for people who don't want package managers, right? :P
22:03:53 <Gregor> elliott: By me if they're interesting :P
22:04:04 <elliott> Stow is the official package manager of the GNU Operating System.
22:04:13 <elliott> And cpio the official archiver :P
22:04:17 <Gregor> elliott: Frankly, I'd very much like to see a package manager properly integrated with cunionfs, so it's in my best interest to be helpful.
22:04:40 <elliott> Gregor: Note: It would not plug into an existing package manager, it'd be an entirely new one based on cunionfs :P
22:04:44 <elliott> If you don't like it, SUX2BEU.
22:04:44 * pikhq is definitely going to end up with a dual major, then.
22:05:06 <pikhq> 3 credit hours extra for the freaking second major? Hells yes.
22:05:10 <elliott> pikhq: Mathematical Knitting and Advanced Haberdashery?
22:05:11 <Gregor> elliott: My integration with dpkg was ... spotty at best, mainly because dpkg (as with all other package managers) is wildly unsuited to union-based "transient" packages.
22:05:22 <pikhq> elliott: CS and mathematics.
22:05:45 <elliott> I would pay endless amounts of money for a degree in Advanced Haberdashery.
22:06:07 <elliott> Gregor: To be honest, the actual package manager part looks suspiciously close to "untar into /pkg/x" the way this is looking :P
22:06:08 * pikhq bestows upon elliott a B.A. in Advanced Haberdashery.
22:06:30 <Gregor> elliott: If it has SOME dependency management beyond that, that'd be nice :P
22:06:46 <Gregor> pikhq: I increasingly regret not getting a minor in archaeology when I had the chance :(
22:06:57 <elliott> Gregor: Yes, but the actual install part :P
22:07:09 <pikhq> Gregor: I could also pick up a minor in just about anything.
22:07:15 <pikhq> Gregor: Er, anything liberal arts-y.
22:07:17 <elliott> Sidenote on archaeology: Indiana Jones films -- 10x better if he sat there doing actual archaeology when shit was doing down? Answer: yes.
22:07:30 <pikhq> Gregor: You see, the liberal arts requirements are completely undefined.
22:07:41 <pikhq> Except in terms of credit hours needed.
22:07:48 <Gregor> elliott: I see no reason why the /install/ would be anything else, it's just the runtime choice of packages to union in that's interesting.
22:07:56 <Gregor> pikhq: Awesome. Awesome to the max.
22:08:19 <zzo38> How I would do it, is package manager operating by pipes. I would do it other programs are also operating by pipes.
22:09:58 <elliott> Gregor: Wait, you mean when you do "emacs" it should create a new union with ONLY THE THINGS EMACS NEEDS?
22:10:14 <elliott> Gregor: That has the major flaw that looking at /usr/bin with emacs would make me go wtf and get angry at my computer for being too smart :P
22:12:01 <elliott> Gregor: Or is that not what you meant.
22:15:33 <elliott> I hypothesise that Gregor cannot see his IRC client from inside his current process.
22:16:37 <j-invariant> frustrating: Terminal objects are defined by universal cones on the empty diagram... but universal cones are defined in terms of terminal object. So i have to define terminal objects twice
22:18:44 -!- zzo38 has quit (Remote host closed the connection).
22:21:45 <elliott> Vorpal: I have pioneered a new type of mining.
22:22:00 <elliott> j-invariant: mutually recursive objects!
22:25:26 -!- sshc has quit (Quit: leaving).
22:29:45 <fizzie> Don't you mean: A New Kind of Mining.
22:30:45 <elliott> fizzie: Yes. I call it: Creeper mining.
22:30:50 <elliott> fizzie: It also works for cutting down trees.
22:31:30 <elliott> fizzie: Basically, you keep all your possessions in a chest. Then at night you walk out, see somewhere you'd like to mine/chop, and get a creeper there.
22:32:43 <Gregor> elliott: No, I do not mean that when you type "emacs", it should create a new union.
22:33:06 <elliott> Gregor: But that would be awesometerrible :P
22:33:47 <Gregor> elliott: The unions are, at the minimal, by process-level, but most users would probably opt to just have their own master list, and the system would presumably have a semi-compulsory list. Users could of course opt to make more/less restricted lists for individual processes if they so desired, but the default (for sanity) would be to /behave/ as a per-user package system, while providing sufficient abstraction for per-process.
22:34:45 <Gregor> Think of it like open fds: If you keep on opening new shit, it's going to have the same std{in,out,error}, unless you or it opt not to.
22:34:49 <elliott> Gregor: How does one actually add a new union?
22:35:40 <Gregor> elliott: CUNIONFS_DIRS environment variable, which (originally, Idonno if I reimplemented this in cunionfs (yet) or not) was backed by a per-user configuration file, backed by a per-system configuration file.
22:36:11 <Gregor> But since environment variables are passed exactly like fds, that abstraction fits best.
22:37:07 <elliott> Gregor: Woo me with its features
22:37:17 <elliott> fizzie: The saddest creeper: standing in a tree, facing the leaves, so that it will never see anything.
22:37:54 <Gregor> elliott: I don't have sufficient motivation to try so hard; if you want each process to have its own unique and easily-malleable view of the FS, you want cunionfs. Otherwise, you don't.
22:38:17 <elliott> Gregor: I don't, but I do want semi-sane builds and stow-like package management :P
22:39:31 <Gregor> In my opinion "stow-like" is a bad compromise. So long as you have everything separated, there's no need to use something so clunky as a whole-system union to view it.
22:40:03 -!- zzo38 has joined.
22:40:36 -!- zzo38 has set topic: The sillier you are to the batsman, the closer you are. | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
22:41:29 <Gregor> But then, once again, it comes down to needs/wants: If you want more flexibility in what individual users/processes/whatever sees, then you really don't have all that many options .. it's either a smarter unionfs or something that clunks together a bunch of PATH, LD_LIBRARY_PATH, CFLAGS, LDFLAGS, etc etc environment variables.
22:41:40 <Gregor> If you don't, don't use it :P
22:41:43 <elliott> <Gregor> In my opinion "stow-like" is a bad compromise. So long as you have everything separated, there's no need to use something so clunky as a whole-system union to view it.
22:41:50 <elliott> What would you have as the usual configuration, then, if not "all packages exposed"?
22:41:52 <zzo38> Do you have time now to review my TeXnicard program so far? http://sprunge.us/RgEZ
22:42:50 <Gregor> elliott: "All packages" could very well have conflicting packages in it. The default would be defined by the package manager. It would probably be what most distros consider a "task", or just whatever the system maintainer cared to make default.
22:43:44 <elliott> Gregor: I'm just saying that in general, 99.9% of the packages would be available, since two conflicting packages are rather rare to install
22:43:55 <Gregor> Remember, packages have versions too.
22:44:09 <elliott> Gregor: OK then, the default would be the largest set of non-conflicting packages or whatever.
22:44:12 <elliott> Gregor: But that's still stow-like
22:44:35 <Gregor> e.g. with stow, macports, etc you can have multiple versions of programs/libraries installed. Then you can swap them out when there are incompatibilities, bugs, whatever. A cunionfs-based system would let you work around such bugs on a per-process basis without having to do anything particularly crazy.
22:45:11 <Gregor> But once again, all the cases I've listed where cunionfs gives you advantages over something whole-system are sort of corner-cases, I just argue that there are sufficiently many such corner cases that giving the flexibility /might/ be worthwhile *shrugs*
22:45:31 <elliott> Gregor: To be honest, I'd go straight to unionfs if not for the nice-builds thing. That seems compelling to me.
22:45:43 <elliott> Hmm, well, how easy is it to change the union of a unionfs/aufs mount without umounting?
22:45:46 <elliott> If not easy: CUNIONFS IT IS
22:46:01 <Gregor> elliott: I have no idea, probably not difficult.
22:46:04 <elliott> ineiros: Well. Uh. I sorta died enough that I have full health now. But you can come see my house!
22:46:34 <zzo38> ineiros: Why do you ask?
22:48:23 <pikhq> elliott: Not very hard; mount -o remount
22:48:34 <elliott> pikhq: Is there any window of not-workingness for the mount?
22:48:56 <elliott> Gregor: Still, I could avoid chroots entirely with cunionfs... right? For building, I mean.
22:49:24 <Gregor> You'd just say "give me these dev packages, OK now build"
22:50:06 <pikhq> elliott: The inodes aren't guaranteed to remain the same across a remount, but otherwise everything continues to work.
22:50:44 <j-invariant> elliott: I need to design a language for category theory (so I can automatically compute things like duals)
22:51:15 <zzo38> j-invariant: Do you have ideas how you can do that?
22:51:34 <j-invariant> zzo38: not yet but I am wondering if it could be done in terms of category theory
22:51:51 <zzo38> j-invariant: How much do you know of category theory?
22:53:04 <j-invariant> zzo38: I think I finally have the basic definitions down, but that took a long time
22:59:54 <j-invariant> but I think the mindset can take a long time to learn because a lot of things can be descrivbed in this language in a very conscice way which you would not expect
23:01:08 <Vorpal> elliott, like TNT, creepers destroy a percentage of the blocsk
23:02:28 <zzo38> Vorpal: What is the percentage?
23:06:55 <zzo38> Someone told me that Wikipedia will be removed. Is that true?
23:07:35 <j-invariant> zzo38: I doubt it unless there is a good reason
23:07:51 <j-invariant> without an explanation there is no reason to beleive it
23:08:06 <zzo38> j-invariant: They told me it is because they have no more money.
23:08:42 <j-invariant> I went to wikipedia to click on the advert bar.. now its gone
23:09:52 <zzo38> j-invariant: I look too and it is gone. I also looked in the meta and in the preferences and did not find information about it.
23:12:14 <Vorpal> j-invariant, the bar at the top of http://wikimediafoundation.org/wiki/Donate looks rather bad though
23:12:43 <zzo38> Vorpal: O, that is where they moved it to.
23:12:53 <Vorpal> j-invariant, oh wait, it needs js
23:12:53 <j-invariant> thanks Vorpal that's what I was looking for
23:13:03 <Vorpal> j-invariant, that looks very very different when I allow it in no-script
23:13:30 <elliott> oh i thought you were mentioning Vorpal for no reason XD
23:14:13 <Vorpal> (who was mentioning I mean)
23:15:03 <Vorpal> elliott, hm, I suspect you have me on ignore.
23:15:46 <Vorpal> he can't take being wrong I guess.
23:16:41 <zzo38> Vorpal: But what if you are being unsure instead of just plainly wrong?
23:17:02 <Vorpal> zzo38, what? I'm not wrong. elliott is
23:17:17 <Vorpal> zzo38, please read what I actually wrote
23:17:22 <zzo38> Vorpal: I didn't mean you personally. I meant in general.
23:18:52 <zzo38> Vorpal: Maybe I am not writing clearly.
23:23:38 <zzo38> Now review TeXnicard program! Tell me if there is any mistake, thing I forgot, opinion, question, etc. http://sprunge.us/RgEZ
23:23:52 <zzo38> And did you notice I changed a few words around in the topic message for this channel, today?
23:26:01 <zzo38> Tell me if (in your opinion) I did the random numbers correctly.
23:29:22 <ineiros> Vorpal, fizzie, elliott and others who might care: I put the normal world back up, without hMod. The backups are now unattended and done without turning the level saving off, so there's a good chance they won't all go as planned.
23:31:51 -!- MigoMipo has quit (Read error: Connection reset by peer).
23:42:40 <Vorpal> ineiros, is it without monsters?
23:58:07 <j-invariant> elliott: "The concrete details of a given construction may be messy, but if the construction satisfies a universal property, one can forget all those details: all there is to know about the construct is already contained in the universal property" - quoted from Dear Learder Wikipedia
23:58:50 <j-invariant> so it's sort of like proof irrelevance, but for general constructions
23:58:51 <elliott> j-invariant: it's like a philosophy of life!
23:59:21 <elliott> j-invariant: hmm, that reminds me of how in non-dependent functional langs, you can encode values as their type's fold combinator