00:00:41 <elliott> "Helped to port scripts on Linux."
00:12:05 <elliott> "[...] I usually don't use my linux machine for development. For me using linux is commandline, I don't need no fancy gui stuff :) That's what windows is for..."
00:12:13 <elliott> i have to read these horrible forums to get this working ;_;
00:25:25 -!- GreaseMonkey has joined.
00:26:41 <Gregor> That sentence makes my brain hurt :P
00:26:46 <Gregor> Mainly the implication that development -> fancy GUI
00:30:06 <elliott> Gregor: (This was in response to someone offering to do Linux support)
00:30:19 <elliott> For a bunch of batch files.
00:30:31 <elliott> Psht, yeah, if you want development tools on Linux I guess. Personally I prefer to keep such fancy GUI things that development requires, like my batch files, on Windows. Silly Linux user.
00:30:50 <elliott> When I'm using Linux, I'm hardcore. Accordingly, I therefore do no development. (Where's vi's refactoring function, I ask you?!)
00:31:49 <elliott> Gregor: (I fabricated the last lines myself.)
00:32:38 -!- sftp has quit (Remote host closed the connection).
00:33:19 <Gregor> I figured since people like that don't use the word "Accordingly"
00:37:12 <pikhq> ... Fancy. GUI. For. Development.
00:37:21 <pikhq> Heck, I'm about to shank at the term "fancy GUI".
00:43:21 <elliott> Gregor: In which Sony bases their fancy new platform on
00:43:22 <elliott> http://snap.sonydeveloper.com/pages/about/
00:43:37 <elliott> If you guessed it before you read it, ... that's implausible!
00:48:04 -!- perdito has quit (Ping timeout: 245 seconds).
00:50:01 -!- FireFly has quit (Quit: swatted to death).
00:55:06 -!- augur has quit (Remote host closed the connection).
00:56:33 <pikhq> elliott: ... They're *using GNUstep*‽
00:56:43 <elliott> pikhq: SONY ADVANCED TECHNOLOGY
00:59:41 <Sgeo> elliott, help remind me that it's ok to kill norns
00:59:58 <elliott> Sgeo: it's not you're a horrible person, they are the only thing that's real
01:00:01 <elliott> them and activeworlds avatars
01:00:17 <Sgeo> Well, no. Today it's not ok. I have homework I need to do.
01:01:01 <Sgeo> But figuring out why cyanide killed my genetically-modified norns is on my list of things to do
01:01:58 <elliott> Sgeo will still have a similar, but not identical, todo list when he is 30. Ponder that, channel, if you will.
01:14:26 * Goose124 can't figure out how to do multiple digit addition
01:15:48 <Gregor> Multiple-digit addition? You have one digit in each cell or something?
01:17:17 <Goose124> but I don't know how to implement
01:22:08 <elliott> Note to self: Tangible values, Haskell?, Manatee-esque, Emacs; combine. Sweet sugar.
01:22:17 <elliott> Goose124: you do realise cells can store up to 255 in a regular implementation?
01:22:24 <elliott> so once you have the digits loaded in what the hell is the problem
01:22:43 <elliott> Goose124: here's how to do b = a + b: [>+<-]
01:22:48 <elliott> where a is cell at pointer and b is one forwards
01:22:50 -!- elliott has quit (Quit: Leaving).
01:23:03 <Goose124> but I want the output to be multiple xD
01:23:22 <Sgeo> Storing as multiple base-256 digits sounds easily doable but... tedious unless you actually use your brain to loop a fixed am... hm
01:23:26 <Sgeo> Actually, what
01:23:33 <Sgeo> I have no idea what I'm talking about
01:24:40 <Sgeo> Please don't get brainfuck advice from me, especially if I mention a 4-letter... it's not an ancronym, WTF is it?
01:25:13 -!- elliott has joined.
01:25:21 -!- elliott has quit (Remote host closed the connection).
01:26:33 <Sgeo> elliott claims to not be here when he really is here in spirit!
01:32:07 <Goose124> That's funny, because I'm pretty sure he has no soul :D
01:38:58 <Sgeo> Goose124, I'm sure he has no soul too
01:39:07 <Sgeo> I'm pretty sure I have no soul either
01:47:50 -!- Hiant has joined.
01:48:32 -!- Hiant has quit (Client Quit).
02:20:35 -!- perdito has joined.
02:40:31 -!- digger has quit (Ping timeout: 276 seconds).
03:13:33 -!- GreaseMonkey has quit (Ping timeout: 240 seconds).
03:15:31 -!- GreaseMonkey has joined.
03:15:31 -!- GreaseMonkey has quit (Changing host).
03:15:31 -!- GreaseMonkey has joined.
03:29:13 -!- augur has joined.
03:38:02 -!- wareya_ has quit (Read error: Connection reset by peer).
03:38:58 -!- wareya has joined.
03:40:37 <Gregor> Yeah, having android and X11 compete for a framebuffer = madness.
03:41:01 <Goose124> ,[>+>+<<-]>>>,[>+>+<<-]<<[>>>[>>+<<-]>>-<<<<<-]>>>>>+<[-]<[-]<[-]<[-]<[-]>>>>>[<<<<+>>>>-]
03:41:34 <Gregor> What does the input do here?
03:41:56 <Goose124> if A is True then B should be True
03:42:07 <Goose124> If A is False, B doesn't matter
03:42:23 <Goose124> output is True unless B is true and A is false
03:45:53 <Goose124> I have OR, AND, NOT, XOR, and IMPLY so far :]
03:46:24 <Gregor> elliott via log: Wow, running X11 + Android at the same time = MADNESS.
04:22:29 -!- digger has joined.
05:05:06 <Gregor> That shooting pain up and down your wrist? That's how you know you're a REAL computer scientist.
05:09:40 <Gregor> pikhq: DOOD, totes join my alternative cover-art contest! http://spamusers.com/forums/viewtopic.php?f=7&t=13483
05:13:53 -!- augur has quit (Remote host closed the connection).
05:24:22 -!- augur has joined.
05:24:51 -!- tee_bot has joined.
06:00:57 -!- sftp has joined.
06:07:35 -!- sftp has quit (Remote host closed the connection).
06:08:42 <Gregor> Well, I think I officially have the ability to grow a douchebag-worth of goatee.
06:08:55 <Gregor> That is to say, a patchy bit of scruff on my chin that probably makes me look like a douchebag.
06:09:04 <Gregor> Or would if not combined with the long hair n' chops.
06:34:25 -!- Sasha has quit (Read error: Connection reset by peer).
06:44:37 -!- FireFly has joined.
06:44:41 -!- FireFly has quit (Changing host).
06:44:41 -!- FireFly has joined.
06:47:13 -!- Sgeo_ has joined.
06:48:29 -!- Sgeo has quit (Ping timeout: 245 seconds).
06:48:54 -!- Mathnerd314 has quit (Disconnected by services).
06:49:13 -!- Mathnerd314_ has joined.
06:50:23 -!- Mathnerd314_ has changed nick to Mathnerd314.
07:50:51 -!- tee_bot has quit (Remote host closed the connection).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:06:18 -!- Mathnerd314 has quit (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.12/20101026210630]).
08:17:03 -!- Sgeo_ has quit (Read error: Connection reset by peer).
08:20:42 -!- Ugo has joined.
08:30:45 -!- FireFly has quit (Quit: swatted to death).
08:36:40 -!- atrapado has joined.
08:43:29 -!- GreaseMonkey has quit (Quit: ilua).
08:55:57 -!- ais523 has joined.
09:24:44 -!- ais523 has quit (Read error: Connection reset by peer).
09:29:51 -!- ais523 has joined.
09:31:54 <ais523> second most unexpected news of the week: Sony decide to use GNUStep as the framework for their next-generation products
09:33:44 <ais523> (I had to look it up; it's a GNU implementation of what evolved into the Apple desktop stack, so it's basically Objective-C and Cocoa, although it actually predates Cocoa)
09:36:12 <ais523> (if only Novell hadn't been bought by Attachmate, it'd definitely have been first)
09:42:41 -!- ais523 has quit (Read error: Connection reset by peer).
09:46:03 -!- ais523 has joined.
10:12:27 -!- ais523 has quit (Read error: Connection reset by peer).
10:35:21 -!- digger has quit (Ping timeout: 250 seconds).
11:07:31 -!- nooga has quit (Ping timeout: 276 seconds).
11:26:03 -!- perdito has quit (Ping timeout: 252 seconds).
11:29:23 -!- ais523 has joined.
11:35:46 -!- ais523 has quit (Ping timeout: 245 seconds).
11:36:39 -!- ais523 has joined.
11:46:14 -!- FireFly has joined.
11:53:42 -!- perdito has joined.
11:58:46 -!- kaasen has joined.
11:59:43 <fizzie> We have not seen him in quite a while.
12:00:25 <perdito> hehe..always missing oerjan
12:00:46 <perdito> whoever he is..it gets kinda mystic
12:09:27 <fizzie> I am a very unstalky person, but I think the other regulars have been more snoopingy about why/where he's gone; not sure about any results, though.
12:10:51 -!- FireFly has quit (Quit: swatted to death).
12:12:24 -!- FireFly has joined.
12:14:03 <fizzie> There is some sort of snoopery talk in the logs at http://tunes.org/~nef/logs/esoteric/10.11.17 but of course it's mostly speculation.
12:20:49 -!- oerjan has joined.
12:28:25 -!- oerjan has quit (Quit: Trying to fix utf8 again).
12:28:42 <ais523> nice to see you here again; I was worried something had happened, but when I saw you editing the wiki I realised you were OK again and stopped wondering
12:30:02 -!- oerjan has joined.
12:30:21 <ais523> <ais523> nice to see you here again; I was worried something had happened, but when I saw you editing the wiki I realised you were OK again and stopped wondering
12:31:25 <oerjan> it would appear one of you got kaasen to reopen my account :D
12:31:46 <oerjan> OK _would_ be overstating it.
12:34:06 -!- coppro has set topic: Hip Hip Hurrah! | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
12:34:09 <oerjan> now i need to find out why irssi doesn't accept my utf-8 keyboard input
12:34:30 <oerjan> could someone paste a utf-8 character of some sort?
12:36:04 <oerjan> that recent wiki spam was awful.
12:37:22 <oerjan> i wish it could have been rolled back instead, how the heck do i find out if there were any _legitimate_ edits?
12:40:16 <ais523> <namekuseijin> it is about programming in the sense that in the near future your only programming options will be coding in either oraclescript or brainfuck, the last bastion of freedom!
12:40:39 <ais523> oerjan: if I'd got to it first (which I didn't), I'd have used bot rollback, which hides both the revert and the original edit from recent changes by default
12:40:52 <ais523> I'm annoyed, because it pushed some vandalism off the bottom of recent changes and so I couldn't find it
12:41:22 <oerjan> well there _is_ an option to show 500 at a time, and i think a previous button as well
12:41:46 <ais523> there's the 500 at a time, but I don't think there's a previous button
12:43:33 <ais523> I fear I'll have to check every page individually to see if it was vandalised (Talk:Befunge had vandalism nobody caught, I only noticed when the page was revandalised)
12:43:37 <oerjan> there should be an option to rollback pairs of edits that revert each other
12:43:56 <ais523> there's an admin-only option to hide both a rollback and the original from recent changes
12:44:10 <ais523> but it's hidden (a GET parameter), and not generally known
12:44:19 <oerjan> yes but i mean you should be able to roll back even if _someone_ else did the revert
12:44:19 <ais523> I've used it before, but have to look it up every time
12:45:22 <oerjan> and of course at the recent spam volume it's a pain regardless (note i didn't even bother to try)
12:49:47 <oerjan> kaasen: kan du skrive noen norske tegn (utf8) her?
12:50:02 <oerjan> har problemer med utf8 i irssi
12:50:22 -!- Slereah has quit (Ping timeout: 240 seconds).
12:50:55 -!- Slereah has joined.
12:51:24 <kaasen> oerjan: kjører du screen(1)?
12:51:36 <oerjan> de siste tre bokstavene funket ikke
12:51:47 <oerjan> de tre foerste saa ok ut
12:52:20 <kaasen> Har du satt $LANG og $LC_... til ....UTF-8?
12:52:33 <oerjan> bare irssi direkte, men jeg har brukt iso-8859-1 lenge og klarer ikke finne ut hvilken opsjon some ikke virker
12:52:54 <oerjan> vel jeg fjernet alle mine personlige LANG/LC_ settinger
12:53:24 <oerjan> og utf8 funker naa fint i pine og paa kommandolinja
12:53:31 <kaasen> Mulig du må ha utf8 i de miljøvariablene. De blir typisk satt av ssh når du logger inn.
12:55:06 <oerjan> er logget inn paa tyrell
12:55:08 <kaasen> Prøv "/set term_charset "utf-8"" i irssi
12:55:30 <oerjan> hm har utf8 uten bindestrek...
12:56:10 <oerjan> nå virker _input_ men ikke output
12:56:52 <kaasen> Ett skritt i riktig retning. :-)
12:56:59 <oerjan> dvs de store bokstavene vises fremdeles feil
12:58:10 <oerjan> det ser rett ut mens jeg skriver det men blir feil vist etter jeg trykker return
13:00:00 <kaasen> Du har valgt UTF-8 i putty, da?
13:00:44 <kaasen> Kanskje du må sette UTF-8 i LC_CTYPE også?
13:01:37 -!- oerjan_ has joined.
13:01:53 -!- oerjan_ has quit (Client Quit).
13:02:03 <kaasen> Hvis du suspenderer irssi med ^Z, hva sier output fra kmd. locale?
13:02:22 -!- oerjan_ has joined.
13:02:45 -!- oerjan_ has quit (Client Quit).
13:03:40 -!- oerjan has quit (Quit: leaving).
13:03:48 -!- oerjan has joined.
13:04:12 <oerjan> nå virker det, LC_CTYPE=en_UTF-8 fikset det
13:05:03 <kaasen> Tegnsett kan være litt kranglete. :-p
13:05:37 <kaasen> Her kjører jeg screen oppå i tillegg. Tør ikke røre oppsettet, for det virker nå. :-p
13:05:46 <oerjan> jeg hadde nå ellers håpet at å fjerne alle mine personlige innstillinger ville gi _ett_ eller annet konsistent resultat ;D
13:05:58 -!- Ugo has quit (Quit: Ugo).
13:06:34 <kaasen> Nja, kommer an på så mangt. ssh f.eks. kan sette den, ~/.profile kan sette dem, osv.
13:07:05 <oerjan> vel min .profile er bare en source av den globale
13:07:50 <kaasen> Det hadde vært så meget enklere om TERMCAP kunne si om en terminal støttet UTF-8 eller ikke, så kunne man bare sette $TERM=xterm-utf8 f.eks.
13:25:40 -!- sftp has joined.
14:13:53 <Vorpal> oerjan, long time since I last saw you
14:14:08 <oerjan> a bit more than a month
14:14:20 <Vorpal> something like that yes
14:23:10 -!- ais523_ has joined.
14:25:27 -!- ais523 has quit (Ping timeout: 250 seconds).
14:27:51 -!- ais523_ has changed nick to ais523.
14:59:32 -!- Mathnerd314 has joined.
15:00:24 <oerjan> wuzzup the merciless, emir of waziristan
15:14:21 -!- MigoMipo has joined.
15:46:52 -!- MigoMipo has quit (Read error: Connection reset by peer).
15:47:26 -!- MigoMipo has joined.
16:01:25 -!- Phantom_Hoover has joined.
16:09:51 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
16:14:52 -!- ais523 has quit (Ping timeout: 264 seconds).
16:20:56 -!- ais523 has joined.
16:21:40 -!- Phantom_Hoover has joined.
16:30:45 -!- Mathnerd314 has quit (Ping timeout: 252 seconds).
16:32:15 -!- elliott has joined.
16:34:19 <elliott> 19:40:37 <Gregor> Yeah, having android and X11 compete for a framebuffer = madness.
16:34:25 <elliott> Gregor: You don't have to call my name to make me read a log line :P
16:34:32 <elliott> Gregor: And lolz at that. How about disabling Android's UI?
16:34:51 <elliott> 21:09:40 <Gregor> pikhq: DOOD, totes join my alternative cover-art contest! http://spamusers.com/forums/viewtopic.php?f=7&t=13483
16:34:52 <Gregor> A) X11 isn't reading the touchscreen (yet!)
16:35:16 <Gregor> B) I haven't yet figured out how to kill the Android UI ...
16:36:15 <elliott> Gregor: does it show up as a unix process?
16:36:29 <elliott> Android is the only Linux implementation that isn't even vaguely Unix-like :P
16:36:35 <Gregor> Yeah, but that process has a watcher process to revive it, and that watcher process is init :P
16:37:11 <elliott> Gregor: ITT: replace /sbin/init
16:37:29 <Gregor> I don't want to break Android, since if X11 doesn't work then I'm kinda fucked :P
16:37:45 <elliott> Gregor: Just make /sbin/init start /bin/sh. (Does it have /dev/console? Let's assume: yes.)
16:37:47 <elliott> 01:33:44 <ais523> (I had to look it up; it's a GNU implementation of what evolved into the Apple desktop stack, so it's basically Objective-C and Cocoa, although it actually predates Cocoa)
16:37:50 <elliott> lolz @ not knowing what gnustep is
16:37:56 <elliott> 04:09:27 <fizzie> I am a very unstalky person, but I think the other regulars have been more snoopingy about why/where he's gone; not sure about any results, though.
16:38:13 <elliott> fizzie: he's alive, he's still editing the wiki, he's still posting blog comments, and he's in some kind of crisis.
16:38:42 <elliott> oerjan: wait, was all of this just a set-up to scare us when you came back?!
16:39:25 <elliott> oerjan: now we just need cpressey back
16:39:34 <elliott> and maybe a form of fax that isn't insane
16:40:02 <elliott> oerjan: well i also want unicorns in the channel, can't a man wish for nonexistent things?
16:40:15 <Gregor> Dinosaur Comics has ties!!!
16:40:18 <elliott> oerjan: speaking of which, get some unicorns
16:41:36 <elliott> oerjan: Phantom_Hoover was looking for you; prepare to be bothered with trivial mathematics!
16:41:41 <oerjan> Gregor now has ties to Dinosaur Comics
16:41:55 <oklopol> i just bored him with trivial maths in pm
16:42:03 <oklopol> so Phantom_Hoover might be too much for one day
16:42:08 <oerjan> elliott: unless i ban him first for stalking me on Godel's Letter
16:42:17 <elliott> oerjan: to be fair, i did that too, i just didn't reply
16:42:27 <elliott> oerjan: i also emailed NVG, but that was before i had any idea you were even alive :D
16:42:35 <elliott> (and after your email bounced)
16:42:55 <oerjan> elliott: in that case you are partially responsible for me coming back today
16:43:11 <elliott> oerjan: how? (obviously 'cuz i'm magic)
16:43:28 <oerjan> elliott: you got kaasen to reinstate my account
16:43:35 * kaasen runs Debian in a chroot on his Android phone.
16:43:38 <elliott> oerjan: by /asking who you are/? :D
16:43:41 <oerjan> ...and phone me about it
16:44:06 <elliott> oerjan: did they kick you out for being way too old?
16:44:19 <oerjan> no, for not paying my membership fee
16:45:08 <elliott> kaasen: is that in reply to Gregor's insane Android tablet muckery?
16:45:12 <Gregor> I can't decide between the "T-Rex Stealth Tie" and the "T-Rex's Busy Day Tie"
16:45:34 <Gregor> kaasen: I have that too, but X11 only through VNC, which is lamesauce :P
16:45:36 <oerjan> oklopol: hey your math may be boring but it isn't trivial! er, wait...
16:46:17 -!- pikhq has quit (Ping timeout: 250 seconds).
16:46:46 <oklopol> it's trivial, you're just so stupid you can't tell the difference between not understanding and being bored by triviality.
16:47:07 <kaasen> elliott: yes, I think so.
16:47:08 <elliott> yep, oerjan's is pretty stupid, gotta say
16:47:20 <elliott> i claim that the double is is a linguistic innovation
16:47:24 <kaasen> Gregor: agree, it's a pain.
16:48:24 <elliott> Gregor: http://www.topatoco.com/merchant.mvc?Screen=PROD&Store_Code=TO&Product_Code=QW-WHITEBOARD&Category_Code=QW do you own this? i am not sure why i do not own this
16:48:25 <Gregor> Somebody who isn't me should write a native Android X11 server.
16:48:31 <fizzie> I run Debian in a chroot on my Maemo phone.
16:48:43 <fizzie> But that's not anything clever, it's right there in the repositories.
16:48:45 <Gregor> elliott: I don't because that's new and already sold out :P
16:48:50 <fizzie> It's even called "easy-debian".
16:49:10 <elliott> fizzie, linux-based phone OS snob
16:49:31 <elliott> "This package, when installed on your tablet or phone, will give you OpenOffice.org, Gimp, the LXDE Desktop Environment, Evince, Firefox (called Iceweasel in Debian) with Java and Flash support, printing support, and access to thousands of precompiled applications from Debian that can be easily browsed and downloaded to your hand-held device."
16:49:35 <kaasen> I thought Maemo _was_ Debian?
16:49:41 <elliott> OpenOffice, GIMP, and LXDE on my phone?
16:49:44 <elliott> MY DREAMS, THEY HAVE COME TRUE
16:50:11 <fizzie> Well, it uses Debian's package management system and such, but the repository is separate and doesn't contain the Debian packages.
16:50:30 <elliott> 04:41:22 <oerjan> well there _is_ an option to show 500 at a time, and i think a previous button as well
16:50:30 <elliott> 04:41:46 <ais523> there's the 500 at a time, but I don't think there's a previous button
16:50:36 <elliott> oerjan: you can hack the URL to make it 50000 instead :p
16:50:42 <elliott> there's some built-in limit i think, but it works
16:50:51 <fizzie> OpenOffice.org on the phone wasn't exactly what I'd call pleasant; I tried it out, of course.
16:51:50 <elliott> oerjan: i like how i can understand your conversation about charsets in norwegian without actually reading any of the words
16:51:59 <elliott> fizzie: what about GIMP :Dm
16:52:03 <elliott> multi-window mobile heaven!
16:52:15 <elliott> oerjan: isn't it kind of breaking tradition to get utf8 working?
16:52:19 <fizzie> elliott: That I didn't even try.
16:55:18 <Gregor> When you can't decide, ya gotta buy both :P
16:56:13 <elliott> "America is warning allies that they may hate the US after the latest Wikileaks release." Best headline ever :P
16:56:33 <elliott> [[The U.S. government has notified Ottawa that the WikiLeaks website is preparing to release sensitive U.S. diplomatic files that could damage U.S. relations with allies around the world.
16:56:33 <elliott> U.S. officials say the documents may contain accounts of compromising conversations with political dissidents and friendly politicians and could result in the expulsion of U.S. diplomats from foreign postings.]]
16:56:56 <elliott> tl;dr "Hi! Uh, you know that WikiLeaks release they say they're gonna make? Yeah, yeah, that one. You're gonna hate that one, I tells ya!"
16:57:31 -!- p_q has joined.
16:57:32 <coppro> because that will lend credibility to the contents
16:57:46 <fizzie> You *could* also use the recentchanges API -- it has rcstart/rcend parameters to specify a time range, and can list oldest-first if you prefer -- but apparently http://esolangs.org/w/api.php just says "MediaWiki API is not enabled for this site."
16:58:11 <elliott> coppro: The US are not known for being non-dumb.
16:58:24 <elliott> New Minecraft update! I wonder if he fixed ALL THE HUNDREDS OF BUGS he just introduced.
16:58:30 <elliott> "Bug update #1. More coming tomorrow"
16:58:50 <elliott> [[Speaking of which, I will make the client updater ask the player before applying new updates. Automatic forced updates are scary, and they were only in because they were convenient and I never thought about it.]]
16:58:57 <elliott> That's good... now let us connect to older servers.
16:59:19 -!- poiuy_qwert has quit (Ping timeout: 245 seconds).
16:59:48 <Phantom_Hoover> oerjan, elliott, what trivial mathematics do I ask about?
16:59:54 -!- ais523 has quit (Read error: Connection reset by peer).
17:02:28 -!- ais523 has joined.
17:02:43 <elliott> Phantom_Hoover: You said you wanted oerjan
17:02:53 <elliott> I assume it wasn't to discuss your Ph.D. thesis :P
17:05:31 <elliott> oerjan: btw we all got addicted to minecraft while you were gone
17:05:48 <Vorpal> elliott, lost connection
17:05:58 <elliott> Vorpal: i can't log in now
17:06:00 <oklopol> oerjan: also we watched a lot of teletubbies, wanna join the fun?
17:06:06 <elliott> maybe it's being upgraded :P
17:06:11 <Vorpal> elliott, nor me. Wait a bit
17:06:14 <Vorpal> elliott, I don't think so
17:06:20 <elliott> Notch i hope you read this. but you need to make it so that difficulty levels are forced. if the host is on normal and another player is on peaceful, he wont see all the monsters the host does. host difficulty needs to be forced to the other players or players will die randomly.]]
17:06:25 <Vorpal> elliott, he said he was going to make food just about 1-2 minutes before
17:07:28 <Phantom_Hoover> elliott, I wanted oerjan because the swatter is making me feel guilty!
17:10:00 -!- Sasha has joined.
17:12:33 <oerjan> <elliott> oerjan: isn't it kind of breaking tradition to get utf8 working? <-- shush
17:13:12 <elliott> oerjan: you need to fashion yourself a unicode swatter now
17:13:27 <Vorpal> ineiros, tell me when it is done
17:14:59 <elliott> Vorpal: how about just trying to connect...
17:15:29 <Vorpal> elliott, then I don't know if it is pre or post upgrade
17:15:50 <elliott> Vorpal: if it works, it's post.
17:15:53 <elliott> connection currently fails
17:20:33 <elliott> [X] Migrate legacy boot sequencing to dependency-based sequencing?
17:21:19 <fizzie> You could use U+25A6 "square with orthogonal crosshatch fill" ▦ for swatting, it's more dense than the #.
17:21:20 <ineiros> Phantom_Hoover: Feel free to come back now.
17:21:44 <fizzie> (There is also a diagonal crosshatch variant.)
17:22:20 <oerjan> elliott: actually the reason i finally got utf8 working was that today when i logged on my _email_ wasn't showing characters properly with my old setup, so i thought wtf and reconfigured.
17:22:55 <oerjan> and then of course i had to reconfigure irssi to work with the rest.
17:24:27 <elliott> fizzie: Took off one of the torches in the tunnel-to-skyway near spawnpoint; it's in the spawnpoint chest, bottom-left, if you want to fix it sometime.
17:33:00 <fizzie> Was something wrong with it or what?
17:35:21 -!- oklopol has quit (Ping timeout: 245 seconds).
17:35:26 <Goose124> why would someone write if x == 0 run code like this: ,[<<+>+>-]<<[>>+<<-]+>[<->-]<[code-
17:36:06 -!- oklopol has joined.
17:37:11 <elliott> Goose124: although if x==0 is never short
17:37:48 <Goose124> and based on how much code there is to run
17:37:52 <elliott> >[-]>[-]<<[>+>+<<-]>[<+>-]+>[<->[-]]<[code (go to temp0)-]
17:38:08 <elliott> where temp0 is one to the right of the cell you start on
17:38:12 <elliott> that requires the first two cells to be free
17:38:17 <elliott> that's the code from http://esolangs.org/wiki/Brainfuck_algorithms#if_.28x_.3D.3D_0.29_.7B_code_.7D
17:38:21 <elliott> Goose124: er are you sure that works?
17:38:41 <Goose124> only works if there is wrapping
17:38:44 <elliott> you could always just do http://esolangs.org/wiki/Brainfuck_algorithms#x_.3D_not_x_.28boolean.2C_logical.29 and then put a loop after it, of course
17:39:30 <Goose124> mine is for boolean values only
17:39:33 <oerjan> that second and third loops in the long one above are stupid, at least. you shouldn't need to move a value like that if you put them in the right spot to begin with.
17:39:38 <Goose124> so if the vaule was 2 it would break
17:40:03 <Goose124> but I intend to use it with all my logic operations
17:41:47 <Goose124> I couldn't get equiv to work exactly
17:41:59 <Goose124> sure it gives a non 0 value if they are equivalent
17:42:10 <oerjan> it's easy to turn a non-boolean value to a boolean one, anyway
17:42:26 -!- Sean13 has joined.
17:43:28 <Goose124> the only operations I somehwat fucked up though is IMPLY
17:43:35 <Goose124> at the end there is a lot of trash
17:43:36 <oerjan> that does a not but you can fix that
17:43:49 <Goose124> and I had to move the answer to 0002 also
17:43:55 <Goose124> ,[>+>+<<-]>>>,[>+>+<<-]<<[>>>[>>+<<-]>>-<<<<<-]>>>>>+<[-]<[-]<[-]<[-]<[-]>>>>>[<<<<+>>>>-]
17:44:42 <Goose124> but is there any good way to implement multiple operations without overlapping cells and getting confused?
17:46:44 -!- Mathnerd314 has joined.
17:48:39 -!- ais523 has quit (Remote host closed the connection).
17:48:54 -!- ais523 has joined.
17:49:18 -!- perdito has quit (Quit: perdito).
17:55:34 -!- wareya_ has joined.
17:58:46 -!- wareya has quit (Ping timeout: 260 seconds).
17:59:26 <elliott> oerjan: hmm is [[-]-]+ not?
17:59:48 <elliott> oerjan: [[-]-]+ does x = !x if x is either 0 or 1
17:59:54 <elliott> which is what Goose124 wants, i think
18:00:08 -!- ais523 has quit (Read error: Connection reset by peer).
18:00:20 -!- atrapado has quit (Quit: Abandonando).
18:02:39 <Goose124> I did not like this...,>+<[>-] :/
18:02:48 -!- augur has quit (Remote host closed the connection).
18:03:31 <oerjan> elliott: erm i was trying to convert from a non-boolean to a boolean
18:03:44 <elliott> oerjan: in response to Goose124
18:03:47 <elliott> who wanted x = !x for boolean x
18:04:04 <elliott> Goose124: current cell must be 0 or 1.
18:04:13 -!- ais523 has joined.
18:04:33 <oerjan> [[-]-]+ most definitely doesn't work
18:04:47 <fizzie> Can [[-]-] ever terminate? After [-] it's 0, after the - it's not.
18:04:56 <oerjan> because at the end of a loop, the value is _always_ 0
18:05:01 <elliott> stupid brainfuck having while instead of if :)
18:05:13 <elliott> ok, i think that's easy enough to fix. hmm
18:05:43 <elliott> -[...it was 0, now 255...]...it was 1, now 0...
18:06:11 <oerjan> i don't think there's any chance of doing it with just one cell
18:06:23 <elliott> oerjan: wait, er, of course there is
18:06:33 <elliott> oerjan: that's "if (!x) foo" for boolean x
18:07:06 <oerjan> i mean if foo doesn't touch any other cells
18:07:57 <oerjan> i meant you cannot do not with just one cell
18:07:58 -!- augur has joined.
18:08:36 <fizzie> So removing the double negative, you can do with just one cell.
18:08:50 <elliott> oerjan: boolean not is just >[-]<-[[-]+>] of course
18:08:52 <elliott> with one cell to the right
18:09:06 <elliott> oerjan: in fact you almost wrote that but not really
18:09:22 <Goose124> I guess my not is a bad implementation xD
18:09:34 <elliott> oerjan: forget i even spoke.
18:10:32 -!- augur has quit (Remote host closed the connection).
18:10:43 <elliott> ais523: thought: mediawiki needs the ability to (1) mark all of a user's edits as bot edits, and (2) roll them all back
18:11:58 <Goose124> what are your thoughts on pbrain?
18:13:41 -!- ais523 has quit (Read error: Connection reset by peer).
18:14:05 -!- ais523 has joined.
18:17:49 <elliott> Goose124: "gluing parts onto a skateboard
18:18:39 <elliott> "Extending the brainfuck language is like bolting parts to a skateboard in an attempt to build a pickup truck. Nonetheless it is a popular activity and will probably remain so. Some more interesting projects use brainfuck as a testbed for programming constructs. Others remove or combine commands, sacrificing symmetry or simplicity. At a minimum, all these efforts should be given names clearly distinct from 'brainfuck' or any euphemism for it."
18:24:41 <Vorpal> elliott, btw (mc): I'll start rebuilding the rest of that cobblestone road when start replacing the cobble in your stairs. And I'll aim for a similar completion date as you
18:25:31 <elliott> Vorpal: My stairs are a piece of public art whose construction was supported by Phantom_Hoover and fizzie. Your stairs, on the other hand, are just ugly, and I don't know of anyone who disagrees.
18:25:38 <elliott> *Vorpal: My stairs are a piece of public art whose construction was supported by Phantom_Hoover and fizzie. Your road, on the other hand, is just ugly, and I don't know of anyone who disagrees.
18:26:03 <Vorpal> elliott, I don't know about anyone agreeing. Also the subtree tower is made of cobble. So is the underwater library
18:26:08 <Vorpal> elliott, and many other things
18:28:23 <elliott> Phantom_Hoover: Support what, the stairs?
18:28:40 <Vorpal> indeed. Neither did anyone support when you *started* those stairs
18:28:46 <Phantom_Hoover> (As an aside, /real/ cobble roads are nice, but they're ugly in MC.)
18:28:54 <elliott> Vorpal: Nobody explicitly *un*-supported, they just didn't help.
18:28:55 <Vorpal> Phantom_Hoover, so use a custom texture pack
18:29:05 <Vorpal> elliott, um I unsupported it :P
18:29:16 <elliott> Vorpal: Yes, but you're just stupid.
18:29:24 <elliott> I'm going to build a gigantic floating penis right outside mount Vorpal and tell you to change texture pack if it's too glaring.
18:29:28 <Vorpal> elliott, personal insult. How fun.
18:29:46 <Vorpal> elliott, anyway. you two doesn't make a majority.
18:29:59 <Vorpal> elliott, but really what about the subtree tower
18:30:00 <elliott> Vorpal: No, but it does make a majority out of everyone who's expressed an opinion.
18:30:10 <Vorpal> elliott, so why aren't you angry at it
18:30:19 <elliott> I'm not angry, your road is ugly, that is all.
18:30:22 <Vorpal> elliott, nor is the road everywhere
18:30:30 <elliott> Vorpal: Also, I would not have brought it up if you didn't say <Vorpal> elliott, btw (mc): I'll start rebuilding the rest of that cobblestone road when start replacing the cobble in your stairs. And I'll aim for a similar completion date as you
18:30:33 <elliott> Vorpal: which was just stupid and petty.
18:30:54 <Vorpal> elliott, well it is just that you called it ugly in the game :P
18:31:04 <Vorpal> sure it could look better
18:31:15 <Vorpal> but so could those stairs
18:32:35 <elliott> Vorpal: My stairs look like that on purpose.
18:33:09 <elliott> Vorpal: Also, you can't walk up stairs of cloth, don't be silly. Physics doesn't work like that.
18:33:13 <elliott> Vorpal: Besides, can you make cloth stairs? I doubt it.
18:33:20 <Vorpal> elliott, well in minecraft it does :P
18:33:38 <Vorpal> elliott, you could use stone
18:34:01 <elliott> Goose124: yeah yeah i'll play DF when i get a computer that wouldn't fall over and cry.
18:34:09 <Vorpal> elliott, seriously. Physically speaking cobble wouldn't hold together by it
18:34:09 <elliott> Goose124: also: totally different, really, despite their similarities.
18:34:12 -!- ais523 has quit (Read error: Connection reset by peer).
18:34:12 <elliott> like saying "doom > nethack"
18:34:17 <elliott> no reason you can't enjoy both
18:34:26 <elliott> Vorpal: It's a very carefully-constructed stairway.
18:34:48 <elliott> Anyway, mining is probably the most boring part of Minecraft.
18:34:51 <Vorpal> elliott, riight. So is the cloth one. Maybe it is reinforced cloth?
18:35:00 <elliott> So your advert is unlikely to work. :p
18:35:07 <elliott> Vorpal: Pics or it didn't happen.
18:35:22 -!- ais523 has joined.
18:35:36 <Vorpal> elliott, anyway it would be possible to make one with enough blocks
18:37:36 <elliott> I'm not going to punch sheep all day.
18:38:08 <Goose124> I just like putting 3500 blocks of TNT underground...:/
18:38:16 -!- poiuy_qwert has joined.
18:38:54 -!- p_q has quit (Ping timeout: 252 seconds).
18:51:59 -!- p_q has joined.
18:52:26 -!- poiuy_qwert has quit (Read error: Connection reset by peer).
19:02:56 -!- Sasha2 has joined.
19:03:04 -!- cheater99 has quit (Ping timeout: 245 seconds).
19:03:30 -!- Sasha has quit (Ping timeout: 240 seconds).
19:06:44 <elliott> Where in the Linux menuconfig does one enable the display of advanced options, again? I can't find it.
19:06:56 -!- perdito has joined.
19:08:36 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
19:11:37 <oerjan> why, obviously it's an advanced option
19:12:41 <fizzie> If you mean the "enable experimental code and/or drivers", I think it was among the very first options in the first group.
19:13:07 <fizzie> Or rather, "enable asking about experimental code"; it doesn't enable anything per se.
19:13:12 -!- ais523 has quit (Remote host closed the connection).
19:13:59 <Goose124> so i gafs it an interface sorta
19:14:28 -!- ais523 has joined.
19:16:21 <elliott> Goose124: okay, grayson miler
19:16:47 <elliott> fizzie: i think it was the "enable configuration of standard stuff (for small systems)" -- paraphrased -- just checked that, and nothing i want to configure
19:16:54 <elliott> (Yes, naysayers, this is for Kitten.)
19:17:29 <elliott> Vorpal: you've probably used oprofile. it's shitty yes?
19:17:32 -!- cheater99 has joined.
19:20:34 <Vorpal> elliott, depends on what you are doing
19:20:48 <Vorpal> elliott, it takes a bit of work. But for some stuff it excels
19:20:49 <elliott> Vorpal: what i'm saying is: convince me not to disable kernel support for me.
19:20:54 <elliott> Vorpal: what i'm saying is: convince me not to disable kernel support for it.
19:21:07 <Vorpal> elliott, who cares. It is your kernel
19:21:23 <Goose124> does anyone know where I can get my hands on a wierd interpreter
19:21:43 <elliott> http://catseye.tc/projects/wierd/
19:21:49 <elliott> http://esoteric.voxelperfect.net/files/wierd/impl/
19:21:55 -!- MigoMipo_ has joined.
19:22:53 <elliott> Goose124: those links are from the wiki page [[Wierd]] :P
19:24:17 <elliott> Goose124: if you know chris pressey was involved, or when in doubt, just check http://catseye.tc/
19:24:31 -!- MigoMipo has quit (Ping timeout: 260 seconds).
19:26:07 * elliott applies patch-2.6.36-ck2.bz2
19:28:41 -!- ais523 has quit (Read error: Connection reset by peer).
19:28:41 * Goose124 thanks elliot for reminding him
19:29:20 -!- ais523 has joined.
19:29:37 -!- Phantom_Hoover has joined.
19:29:55 -!- Sasha2_ has joined.
19:30:31 -!- Sasha2 has quit (Ping timeout: 245 seconds).
19:30:50 <Goose124> gonna get rid of windows for good this time :P
19:31:36 -!- kar8nga has joined.
19:35:17 -!- Goose124 has quit (Quit: A day without sunshine is like .... night).
19:39:16 <elliott> and he even misspelled my name (how do people do that??)
19:39:25 <elliott> Elliot Toodledoo or something
19:39:31 <elliott> ugh even typing that makes me pain
19:43:41 <Ilari> Mispell someone's name... I have whole list of mispellings of my name I have seen (ranging from understandable sans-serif typos to just plain weird).
19:45:20 <Ilari> Well, most are from IRC...
19:45:43 <fizzie> My surname's been once mangled to Mallasjoki (lit. "malt river").
19:47:01 <Ilari> Perhaps the most bizarre is "Elarry".
19:47:54 <ineiros> I think a few times I've ordered from the States, my name is written as "LLARI" on the package.
19:48:54 <Ilari> Sans-serif fonts can produce that.
19:49:53 <Ilari> Because in those fonts, Capital I and small L look exactly alike. I have also seen IIari.
19:50:11 <Ilari> (that's the other way to confuse the characters).
19:51:06 <elliott> Not exactly in this one at least.
19:51:22 <elliott> that is, little L is taller than big i
19:52:14 -!- perdito has changed nick to perdito|afk.
19:52:24 <Ilari> But I haven't seen lIari yet...
19:56:19 <elliott> Vorpal: you've done insane wakeups optimisation yes?
19:57:35 <olsner> wakeups optimisation yes?
20:00:37 <elliott> │ Use early_res directly instead of bootmem before slab is ready. │
20:00:37 <elliott> │ - allocator (buddy) [generic] │
20:00:37 <elliott> │ - early allocator (bootmem) [generic] │
20:00:37 <elliott> │ - very early allocator (reserve_early*()) [x86] │
20:00:37 <elliott> │ - very very early allocator (early brk model) [x86] │
20:00:38 <elliott> │ So reduce one layer between early allocator to final allocator │
20:00:41 <elliott> best (worst) config description ever
20:02:00 <olsner> I'm still only at my hardcoded page numbers allocator
20:04:56 <olsner> I was thinking about implementing *the* allocator in assembly as probably the next step of my kernel
20:05:02 -!- Sgeo has joined.
20:05:16 <Sgeo> Illiad forgot to put the "This 'toon is a repeat" thing up
20:08:05 <elliott> That's okay because nobody with a brain reads UserFriendly.
20:08:16 <elliott> And it's not very popular among the decerebrates, either.
20:08:36 <elliott> olsner: put a GC system in the kernel
20:08:52 <elliott> olsner: (repeat after me: Awesomeness takes precedence over supporting C. Awesomeness takes precedence over supporting C.)
20:09:21 <olsner> supporting C? I'm going to support assembly
20:09:55 -!- Phantom_Hoover has changed nick to lIari.
20:09:56 <olsner> everything else is up to whoever cares about making a C implementation
20:10:18 <elliott> olsner: but if you do a GC and implement the rest in $language, it'll be awesome! and I can steal your code again!
20:10:36 <olsner> ooh, I see what you're up to now
20:10:41 -!- lIari has changed nick to Phantom_Hoover.
20:12:36 <elliott> olsner: I *will* get you to implement ElliottOS for me!
20:17:19 <elliott> ┌──────────────── /dev/cpu/microcode - microcode support ─────────────────┐
20:17:30 <elliott> │ │ [*] Intel microcode patch loading support │ │
20:17:31 <elliott> │ │ [*] AMD microcode patch loading support │ │
20:27:26 <elliott> │ 250 HZ is a lousy compromise choice allowing server interactivity │
20:27:26 <elliott> │ while also showing desktop throughput and no extra power saving on │
20:27:26 <elliott> │ laptops. No good for anything. │
20:30:49 <fizzie> "Good for when you can't make up your mind."
20:32:00 <fizzie> It doesn't seem to be so written in this 2.6.34.1 kernel I have here, but I see it's some sort of hz-no_default_250.patch in someone's tree.
20:33:15 <elliott> fizzie: It's part of the infamous-and-now-revived -ck patchset.
20:33:33 <elliott> fizzie: (It recommends 300 Hz for the compromisers.)
20:34:33 <elliott> fizzie: I've applied -ck because it has the Brain Fuck Scheduler and the other stuff looks benign and/or obvious. :p
20:34:45 <elliott> I figure he's unlikely to do anything stupid.
20:35:02 <fizzie> Phantom_Hoover: Oh, I thought you were planning an ambitious build in you-know-where!
20:36:00 <olsner> hmm, thinking about cool ways to make VM features available to user-mode, like registering callbacks for your own page faults... looks like pretty much everything is also possible from signal handlers though
20:36:24 <elliott> olsner: why do you have kernel vs user mode, THINK OUTSIDE THE BAUX
20:36:35 <elliott> CAPABILITY PERMISSION VERILY ANTELOPWE
20:37:21 <Vorpal> <elliott> Vorpal: you've done insane wakeups optimisation yes? <-- to some extent.
20:37:28 <elliott> olsner: baux is how you spell box if you're classy.
20:37:33 <elliott> olsner: antewelope is an antelope with extra w.
20:37:48 <Vorpal> elliott, and the places I'm currently in mostly have power outlets. So no longer a major issue
20:37:54 <elliott> Vorpal: higher kernel tick Hz won't cause extra wakeups if dynamic ticks is on, yes? -ck option descriptions say so.
20:38:14 <elliott> right now I'm configuring it as 1000 Hz with dynamic ticks
20:38:20 <Vorpal> elliott, correct in theory at least. Though iirc some versions have been buggy wrt that
20:38:35 <Vorpal> (I have no clue about -ck)
20:38:43 <elliott> Vorpal: -ck replaces the whole scheduler, so.
20:38:48 <elliott> it has a good chance of fixing any bugs like that :P
20:38:55 <Phantom_Hoover> Bah, this paper is about computing the geodesic between two points!
20:38:55 * elliott considers turning off kvm... naw, i'm not that cruel
20:39:08 <elliott> I have disabled Xen support and "special stuff" for running *under* KVM.
20:39:12 <elliott> That can go in another kernel package.
20:39:33 <Vorpal> Phantom_Hoover, you could reverse it presumably
20:39:41 <elliott> "If unsure, say M." -- that's a new one.
20:39:54 <Vorpal> unless I'm totally confused about how geodesic works
20:40:08 <elliott> "Enable the block layer" Pointless bloat!
20:40:09 <Vorpal> elliott, did you see my minecart system?
20:40:16 -!- p_q has quit (Ping timeout: 272 seconds).
20:40:21 <Phantom_Hoover> Vorpal, the geodesic between two points is the shortest distance between them (taking topology into account).
20:40:23 <elliott> also, /me notes that this kernel is going to be like 7, 10 MiB... stupid overhead
20:40:33 <elliott> I suppose I could decrease max processors from 512; that'd save 4 MiB. :P
20:40:35 <Vorpal> elliott, between subtree/fizzie intersection and my entrance to the tunnel
20:40:39 <elliott> (Well, if I reduced it to... 0.)
20:40:42 <Vorpal> elliott, plan is to extend it to mt hover
20:40:48 <Vorpal> elliott, but I'm out of iron
20:40:48 <fizzie> elliott: Is "enabling the block layer" some sort of euphemism for an in-kernel minecraft?
20:40:49 <Phantom_Hoover> The geodesic from a point given a direction is not something I understand.
20:41:01 <elliott> Vorpal: Cool; route it to the spawn point too.
20:41:07 <Vorpal> elliott, you must bring your own minecart to the place
20:41:15 <Vorpal> elliott, must be underground. Or animals get in
20:41:30 <Vorpal> elliott, there are some minecarts at the subtree
20:41:30 <elliott> Vorpal: Yes really; I'd rather work the skyway than craft every time I want to go somewhere.
20:41:49 <elliott> Vorpal: That's not "my own".
20:42:09 <Vorpal> elliott, well I carted some of them in previous attempts. Feel free to take
20:42:15 <Vorpal> elliott, they are in the one of the chests there
20:42:34 <Vorpal> elliott, you could have a chest with them at each end and then put them back. If you go about as often in both directions you should have a constant supply in both places
20:42:48 -!- poiuy_qwert has joined.
20:42:52 <elliott> Vorpal: Can't you just have a system where they go around in a loop forever...?
20:42:52 <olsner> elliott: I quite like kernel/user mode separation, and I definitely don't intend to only support specific languages (like lisp)
20:42:55 <Vorpal> elliott, and /home isn't that much faster any more. It cuts travel time a LOT. Seems that minecarts are faster on MP than in single player
20:43:08 <Vorpal> elliott, only if all the involved chunks stay loaded
20:43:10 <elliott> Vorpal: "and /home isn't that much faster any more." Eh?
20:43:31 * elliott ponders momentarily whether to drop all EFI support.
20:43:35 <Vorpal> elliott, /home from mt. hoover is currently way faster than walking. The difference will be way smaller with the minecarts
20:43:49 <Vorpal> well currently they don't go all the way. But based on how far they go now...
20:43:53 <olsner> but I would like the kernel part to be as small as possible, and microkernelish... then everything is user mode :)
20:44:47 <elliott> olsner: just make it a picokernel; have some kind of IPC mechanism, and export a fake "kernel" process that has one message: task-switch-to(process-id). Have no automatic task-switching.
20:44:59 <elliott> olsner: Then you can implement the task switcher as the first user-mode process below that. :)
20:45:08 <elliott> (And have it deny IPC access to the kernel process for all its children.)
20:45:38 <Vorpal> elliott, between intersection and my tunnel entrance: about 7 seconds.
20:45:48 <elliott> Vorpal: Where is it again?
20:45:55 <Vorpal> elliott, where is what?
20:46:00 <Vorpal> elliott, the intersection?
20:46:18 <elliott> Vorpal: The Minecart system.
20:46:36 <Vorpal> elliott, go down to subtree and get minecart. then go to mt. hoover in the tunnel. You will see it fairly soon
20:47:11 <elliott> Vorpal: wait, it starts in the Nailor Memorial Mount Hoover Tunnel Passage?
20:47:29 <elliott> Vorpal: It needs a name so I've named it as of now.
20:47:35 <elliott> Vorpal: It is the tunnel to Mount Hoover that I refer to.
20:47:40 <olsner> elliott: yes, that's a pretty good start... you do need some way to manage pages though, and you definitely can't let processes access any physical page they want
20:49:09 <elliott> olsner: OK, have two fake processes: page-manager and task-switcher. :p
20:49:11 <Vorpal> elliott, anyway yes that is where it is. Just to my entrance to that tunnel so far
20:49:21 <elliott> olsner: Actually, just have one kernel one, and also expose a couple of page functions.
20:49:47 <elliott> olsner: You can do "security" i.e. not letting processes access certain pages by using whatever simple security thing you put in the IPC -- like "deny my children IPC access to process 0 (kernel)".
20:49:58 <elliott> And then implement the subset they're allowed to do -- the safe page operations -- as an interface of that process.
20:51:40 -!- oerjan has quit (Quit: leaving).
20:54:41 <olsner> yep, something like that
20:56:51 <olsner> was thinking about just having a flexible interface for transferring/sharing pages between processes and use that both for IPC and for memory allocation
20:57:57 <olsner> e.g. ask the page-process to give you a bunch of zero-initialized pages and it'll get back to you with a "hey you got pages" call
21:00:51 -!- GreaseMonkey has joined.
21:04:26 <elliott> olsner: I approve of this; all you'd need to do is expose a single process with a few paging functions and one task-switching call, and that'd be the entire kernel.
21:05:04 <olsner> there are a couple of details more though: I/O access and interrupt handlers, which will be required to implement drivers
21:05:05 <Vorpal> elliott, it broke didn't it?
21:07:59 <elliott> Vorpal: Question: Could you avoid having to push it by having the first section, up until the booster, be powered? So that as soon as you stepped on a plate just in front of the track, it'd power a circuit that, when you put the cart down, would start moving it, slow-ish, towards the booster; you would then right click it, without having to dash for it, and since you'd have stepped off the plate, the power would stop, and momentum would carry you
21:08:21 <Vorpal> elliott, not really. Plus you easily learn how to push
21:08:26 <elliott> Vorpal: Server is working now.
21:08:28 <Vorpal> elliott, a nudge then jump in
21:08:38 <Vorpal> elliott, how would you route the thing off
21:08:50 <elliott> Vorpal: Can you restate that sentence?
21:10:34 <olsner> been thinking about what would be the most efficient way to do IPC in a multi-cpu system, with a minimum of explicit inter-cpu communication and costly system calls... maybe something like processes being able to sleep in a wait-on-shared-page-access (to avoid the need for explicit wakeup signals from the sender)
21:12:06 <elliott> Vorpal: I don't understand.
21:12:22 <elliott> ineiros: Couldn't connect to the server once. Works now.
21:12:40 <Vorpal> elliott, the powered minecart would be hard to route off another place. Delays are not reliable on MP
21:12:53 <elliott> Vorpal: Only the very first section would be powered, up to the booster.
21:13:00 <Vorpal> elliott, and sometimes minecarts going over plates don't trigger the plates
21:13:02 <elliott> olsner: alloc-page() -> page
21:13:11 <elliott> olsner: free-page(page) -> void (?)
21:13:11 <Vorpal> elliott, yes you need a powered minecart for that
21:13:17 <Vorpal> elliott, which you need to route off
21:13:20 <Vorpal> elliott, which is tricky
21:13:32 <elliott> olsner: hmm, identify pages by SHA512 hash
21:13:44 <elliott> olsner: and take it as read that anyone with the hash has been granted access to the page :P
21:15:41 <olsner> I think free-page would be pretty much the same as when the page-process gives pages to other processes, only that the page-manager knows that any page it receives exclusive access to is one that has been "freed"
21:16:13 <elliott> ineiros: We could really use Nx64 minecart tracks. Also minecarts. :P
21:16:25 <olsner> i.e. there's some kind of flag for asking the real kernel to remove the page from your own page table when adding it to the other process' page table
21:16:35 <elliott> olsner: alloc-page() -> page
21:16:42 <elliott> olsner: grant-page(page, pid) -> void
21:16:49 <elliott> olsner: revoke-page(page, pid) -> void
21:16:54 <elliott> olsner: wait-until-write(page) -> void
21:17:35 <elliott> olsner: make grant-page instead
21:17:45 <elliott> olsner: grant-page(page, pid[, is_admin]) -> void
21:17:51 <elliott> olsner: only page admins can use the grant/revoke functions, obviously
21:18:04 <elliott> olsner: and trying to revoke a page from its only admin is an error
21:18:16 <elliott> olsner: (this lets you grant pages to people without necessarily letting them mess around with who they're granted to)
21:18:23 -!- kaasen has left (?).
21:18:35 <elliott> olsner: So all you need is that and switch-to(pid) -> void, and "theoretically" you can do the rest in userspace. :P
21:18:35 <olsner> hmm, but I want read-only sharing and private mappings and stuff like that
21:18:49 <elliott> olsner: (And of course, all those functions are exposed as page-based IPC.)
21:18:58 <elliott> olsner: get rid of revoke-page, and make grant-page be
21:19:01 <olsner> e.g. if someone read-write mmaps a file, and a bunch of process read-only/private map them, they'll all share the same actual frame
21:19:03 <elliott> grant-page(page, pid, permissions) -> void
21:19:09 <elliott> olsner: if permissions = 0, it's the same as revoke
21:19:14 <elliott> if permissions & 1, it's read
21:19:17 <elliott> if permissions & 2, it's write
21:19:24 <elliott> if permissions & 3, it's ability to use grant-page on this page
21:20:07 <elliott> olsner: so freeing a page is... grant-page(page, 0, 7)
21:20:15 <elliott> grant-page(page, mypid, 0)
21:20:21 <olsner> pretty much, still not sure about having "granting" control at all though
21:20:26 <elliott> olsner: well, that's how you'd share it, no?
21:20:41 <elliott> olsner: this thing implements both exposing IPC and sharing privileges in one interface, doesn't it?
21:20:46 <olsner> if you have access you can always give away that much access to someone else, but you can obviously also restrict them
21:21:31 <elliott> olsner: hmm, ok then: remove the admin bit
21:21:39 <elliott> olsner: and have the rule that you can only grant-page to a pid that is your child
21:21:47 <elliott> tada, suddenly you can't revoke your parent's page access any more
21:21:53 <elliott> and the process hierarchy enforces security
21:22:17 <elliott> olsner: so basically, saying to something "hey, IPC me up on an interface" would look something like... this i think (this is just me thinking, feel free to comment)
21:22:30 <elliott> grant-page(page, thepid, WRITE)
21:22:49 <olsner> I wouldn't have restricted it to only children though, am I missing an obvious risk there?
21:22:50 <elliott> olsner: you realise i'm trying to tarpit your kernel into near-nothingness :D
21:23:08 <elliott> grant-page(page, some_pid_that_can_currently_read_and_write_the_page, 0)
21:23:12 <elliott> then you revoke its access
21:23:15 <elliott> this is useful because e.g.
21:23:18 <elliott> grant-page(page, my_pid, 0)
21:23:29 <elliott> (well, only if you're the only one with access to it, but you see what i mean)
21:23:36 <elliott> grant-page is actually modify-permissions-for-page
21:23:37 -!- ais523_ has joined.
21:23:39 <elliott> but grant-page is a snappier name :P
21:23:49 -!- ais523 has quit (Disconnected by services).
21:23:51 -!- ais523_ has changed nick to ais523.
21:24:13 <olsner> oh, that's quite different from the api I had in mind though
21:24:35 <elliott> olsner: what did you have in mind? i'm interested
21:24:55 <elliott> olsner: otoh, my model currently has four functions representing the *entire* kernel interface, and they're all trivial, so it better be good :D
21:25:15 <elliott> (alloc-page, grant-page, block-until-page-write, switch-to)
21:25:18 <olsner> for one, you can never give away a page that you don't have access to yourself, and you can't revoke access
21:25:30 <elliott> (2) then how do you do free()?
21:25:40 <elliott> you said that if the kernel was the only one with access to a page, it'd know to free it
21:25:46 <olsner> and granting the same page to the same process twice just maps the same page in another place
21:25:46 <elliott> but how can you do that if you can never give up a page?
21:26:00 <elliott> olsner: hmm, i was thinking that you'd have the page address be the same for every process
21:26:06 <elliott> and you'd just communicate that address to the other processes
21:26:55 <olsner> not decided about that, but I was thinking that you give a page using your own virtual address and the other end receives the same physical page in its own address space :)
21:27:21 <elliott> olsner: or you could have the same address space for the whole OS, and just have every page access crash the process unless it has the privileges
21:27:25 <elliott> olsner: that's super-simple :P
21:28:19 <olsner> and when you give away a page "exclusively" you actually only give away your own reference to it
21:28:29 <olsner> and existing sharing still remains
21:29:00 <olsner> this causes some tricky stuff when freeing though, because the page-master needs to know when all other owners of pages have disappeared
21:30:20 <elliott> olsner: that sounds really complicated, with mine, all you have to do is store two bits (well, three if you have an execute permission or whatever) for each page
21:30:34 <elliott> olsner: two bits * processes given access to it
21:30:47 <elliott> olsner: and when those bits are zero, you can reuse them for another process that accesses it
21:30:52 <elliott> olsner: my system is beyond dirt simple :P
21:30:57 <elliott> your kernel would be like. 5 lines.
21:31:40 -!- ais523 has quit (Read error: Connection reset by peer).
21:34:02 -!- ais523 has joined.
21:36:45 -!- jcp has quit (Ping timeout: 240 seconds).
21:40:55 -!- jcp has joined.
22:10:19 -!- Sean13 has quit (Quit: If you think nobody cares, try missing a few payments).
22:26:07 <elliott> ineiros: apparently the new server is also hangy sometimes!
22:27:07 <ineiros> Might be related to the connection as well. I might accidentally use all of the INTERNETS now and then.
22:43:50 -!- MigoMipo_ has quit (Read error: Connection reset by peer).
22:44:34 -!- kar8nga has quit (Remote host closed the connection).
22:50:19 <elliott> ineiros: Can I have /tp ehird BCxVAhWQxi?
22:51:14 <perdito|afk> are different users..different process' in that view?
22:55:59 <perdito|afk> forget that question.. there's enough log to hand thath oneenoght
23:03:41 <Vorpal> ineiros, timeout again
23:06:38 <ineiros> I noticed, if this keeps going on, I'll complain to my ISP.
23:10:46 <fizzie> Ask the "what sort of minecraft discriminators are you!?"
23:13:05 -!- ais523 has quit (Remote host closed the connection).
23:19:09 <elliott> ineiros: Well, consumer ISPs generally aren't very good for servers. :p
23:19:21 <ineiros> Yeah, but it's affecting all my connections.
23:19:37 <elliott> ineiros: How much RAM does it use? When I get around to buying a VPS again I'd be happy to host it if it isn't too RAM hungry.
23:20:00 <elliott> (But I hear the server uses hueg amounts of RAM.)
23:20:27 <ineiros> elliott: As much as you want to give it. I think 1G is about the minimum what it will run on.
23:20:39 -!- FireFly has quit (Quit: swatted to death).
23:20:43 <elliott> ineiros: And it uses more for you, presumably.
23:20:54 <elliott> ineiros: A shame then; the VPS I'm probably getting is 1 GiB total.
23:21:18 <elliott> (Also in America, which is problematic for every single current player.)
23:21:24 <ineiros> elliott: Well, I run it with java -Xmx3072M -Xms3072M -jar minecraft_server.jar nogui
23:21:34 -!- Phantom_Hoover has quit (Remote host closed the connection).
23:22:32 <ineiros> elliott: Where are you planning to get a VPS?
23:23:14 <elliott> ineiros: http://prgmr.com/xen/; they're cheap, run by long-time experienced Xen guys, open about downtime and problems on their blog, and used and recommended by Gregor (after I second-hand recommended them).
23:23:35 <elliott> ineiros: (There was that disk thrashing issue, but that's resolved and he got a discounted price for the duration anyway.)
23:23:52 <elliott> ineiros: $20/mo for 1 GiB RAM, 24 GiB disk, 160 GiB network transfer seems to be the sweet spot for what it's worth.
23:26:04 <elliott> ineiros: (Previously I was with Slicehost who I would no longer recommend since they got bought out by The Man^W^Wrackspace. Besides, *they* don't let you run your own kernel, just a predefined distro list. I'll stop blabbing now.)
23:27:19 <Vorpal> elliott, 1 GB RAM is not enough
23:27:24 <Vorpal> elliott, it wants 2 GB no less
23:28:11 <elliott> Vorpal: yeah, I've heard that. (I tried to tell the same to two friends thinking they could set up a server on a machine with not much ram (256 or 512 MiB), but they decided to try anyway because of, I don't know, mental retardation or whatever.)
23:28:34 <Vorpal> elliott, and seems ineiros used 3 GB
23:28:45 <elliott> I can indeed convert MiB values to GiB.
23:29:18 <Vorpal> elliott, really? Sounds like an unique skill. Maybe you should enter some competition. Or a freak show.
23:29:47 <fizzie> You don't even have to, because he said "3G" in addition to the command line.
23:30:01 <Vorpal> oh didn't see that line
23:30:14 <Vorpal> fizzie, also that would explain the laggyness (hur hur)
23:31:26 <fizzie> Coincidentally, our gracious host was quite a while (month?) with only a 3G uplink at home. That would have made for some quality crafting.
23:32:39 <elliott> fizzie: Ditto, a while ago! It liked to downgrade to EDGE, too.
23:32:44 <Vorpal> fizzie, you can go all the way now
23:33:22 <fizzie> I can go all the way to sleep, you mean.
23:33:46 * Sgeo is reminded of Orisinal <3
23:34:24 <fizzie> Some big, hairy (well, most likely) dudes are going to barge in at 08am to put in more stuff to make the windows a bit less leaky.
23:34:36 <Vorpal> fizzie, no I mean to mt. hover
23:35:04 <Vorpal> fizzie, in several steps due to MP bugs
23:35:17 <fizzie> Mt. Hover sounds more floating.
23:36:20 <fizzie> I guess there should also be tracks from the intersection to subtree, optimally.
23:37:04 <elliott> fizzie: There should be tracks RIGHT TO YOUR DOOR, too!
23:37:17 <Vorpal> fizzie, sure. Go ahead and place them
23:37:24 <Vorpal> elliott, that is a very short walk
23:37:38 <Vorpal> elliott, but fizzie should fix his horrible stairs (annoying to walk up)
23:37:53 <elliott> also, it's his damn house :P
23:37:57 <Vorpal> elliott, that would take more time to walk up
23:38:18 <Vorpal> Than walk from intersection to fizzie
23:38:24 <Vorpal> good idea to subtree though
23:38:54 <fizzie> What's wrong with my stairs?
23:39:08 <Vorpal> fizzie, that they are not very user friendly?
23:39:13 <elliott> fizzie: ever tried walking up them? :p
23:39:30 <fizzie> I've walked up them at least a dozen times.
23:39:34 <Vorpal> it can be done but is annoyingly slow
23:39:47 <Vorpal> due to having to turn very carefully
23:39:52 <fizzie> You don't have to visit if the stairs are too much for you. :p
23:40:24 <elliott> "Forget that, MINECRAFT takes the edge off? I freely admit I'm a wuss, but I've played once since the Halloween update, and spent pretty much the entire time cowering as armies of creepers overran my ineffectual day 1 forts."
23:40:25 <Vorpal> anyway: tracks to near spawnpoint: good idea
23:40:49 <elliott> Vorpal: just do it to the closest area you can build on, and then make a hole with water. :p
23:40:53 <elliott> and have a chest of minecarts there
23:41:05 <elliott> it would replace the skyway as the easiest way to get around
23:41:15 <Vorpal> well no. I would not have a chest there :P
23:41:26 <Vorpal> elliott, after all the only reason to go there is /home
23:41:35 <Vorpal> elliott, and then you have it in inventory
23:41:40 <Vorpal> Oh? Dying? What dying?
23:41:40 <elliott> Vorpal: not if you just died.
23:41:57 <Vorpal> elliott, yes but the issue is that chest will never be filled up
23:42:04 <Vorpal> elliott, thus it will soon become empty
23:42:14 <Vorpal> elliott, thus the whole plan fails
23:43:06 <elliott> Vorpal: no, because you're expected to put the carts in the next chest
23:43:20 <elliott> and every now and then people should go to the chest and deposit some carts there, from the next chest
23:43:39 <Vorpal> elliott, would you do that? Or would "people" be "me"?
23:44:03 <elliott> Vorpal: I mean, it'd consist of getting some carts, typing /home, walking a few places, and putting stuff in a chest.
23:44:20 <Vorpal> elliott, okay you can build that. Personally I'm not really interested in that bit since I carry my personal cart.
23:44:49 <elliott> Vorpal is a ruthless capitalist, caring not about anyone else but himself.
23:45:22 <Vorpal> elliott, if I was, why the heck did I build that all the way to mt. hoover?
23:45:27 <Vorpal> elliott, rather than just to my place
23:46:08 <elliott> Vorpal: To corner the market on long-distance transport, duh.
23:46:35 <Vorpal> elliott, anyway you have never built a booster before right?
23:46:40 <Vorpal> I would guess fizzie has
23:46:41 -!- Sgeo_ has joined.
23:47:05 <Vorpal> elliott, so there you have it
23:47:35 <Vorpal> elliott, I have the expertise :P
23:48:10 <Vorpal> anyway, this was a completely new booster design for me (though trivial, I assume someone already thought of it)
23:48:28 -!- Sgeo has quit (Ping timeout: 264 seconds).
23:48:30 <elliott> Vorpal: Patented, one presumes.
23:48:46 <Vorpal> elliott, it is a trivial academic result
23:48:52 <Vorpal> elliott, hardly worth even a lemma
23:51:03 * elliott removes reiserfs support from the default kernel
23:51:16 <elliott> I need to get more militant about stripping crap out of it; if anyone wants to add stuff they can always just use kerntool(1).
23:52:08 <elliott> Vorpal: yes, but nobody actually *uses* them nowadays in Linux.
23:52:09 <Vorpal> anyway that is your kernel
23:52:18 <Vorpal> elliott, not yet indeed
23:52:29 <elliott> Vorpal: I even like dynamic linking in the sense that it's late binding; it's just that the Unix implementations mess it up.
23:52:45 <Vorpal> elliott, why do you like late binding?
23:52:45 <elliott> i.e., I like the idea of ACLs in general (well, generalised as capabilities), but I don't think they're any use in Linux.
23:53:00 <elliott> Vorpal: I like really-late-binding (Smalltalk) and really-early-binding (Forth). :-)
23:53:13 * Vorpal replaces elliott's late binding with whatever way C++ implements it
23:53:15 <elliott> Vorpal: Late binding is really the heart of what people mean when they say "dynamic/flexible"; Lisp, Smalltalk, etc.
23:53:40 <Vorpal> presumably it involves vtables
23:53:45 <elliott> ...naw, I don't need quotas.
23:54:06 <elliott> (Adding any of this stuff consists of adding one line to ~/.kerntoolrc and running "kerntool install" as root, anyway.)
23:54:25 <elliott> (I'm making it so easy for all those fuckers with nvidia and ati cards. :p)
23:55:14 <elliott> Note to self: Look into integrating TuxOnIce at some point.
23:56:13 <elliott> I don't think I need in-kernel NTFS support, since anyone would use ntfs3g anyway.
23:56:25 <elliott> │ The only supported operation is overwriting existing files, without │
23:56:25 <elliott> │ changing the file length. No file or directory creation, deletion or │
23:56:25 <elliott> │ renaming is possible. Note only non-resident files can be written to │
23:56:25 <elliott> │ so you may find that some very small files (<500 bytes or so) cannot │
23:57:05 <elliott> │ │ [*] Tmpfs POSIX Access Control Lists │ │
23:58:13 <Vorpal> elliott, that has been there for ages
23:58:19 <elliott> Vorpal: I know, just, still. :)
23:58:29 <elliott> │ The Acorn Disc Filing System is the standard file system of the │
23:58:29 <elliott> │ RiscOS operating system which runs on Acorn's ARM-based Risc PC │
23:58:29 <elliott> │ systems and the Acorn Archimedes range of machines.
23:59:08 <Vorpal> elliott, I just go for freebsd slices and apple partition map when it comes to partitioning system support
23:59:17 <Vorpal> and more recently EFI GUID stuff too
23:59:24 <elliott> Vorpal: Go for? As in to remove?
23:59:36 <Vorpal> elliott, as in enable those apart from normal MBR
23:59:53 <Vorpal> elliott, I use things with freebsd slices sometimes