←2010-02-02 2010-02-03 2010-02-04→ ↑2010 ↑all
00:03:56 -!- SimonRC has quit (Ping timeout: 246 seconds).
00:05:04 -!- Slereah has joined.
00:06:25 -!- Slereah_ has quit (Ping timeout: 240 seconds).
00:09:56 -!- GreaseMonkey has quit (Read error: Connection reset by peer).
00:10:26 -!- GreaseMonkey has joined.
00:10:35 -!- cpressey has left (?).
00:11:45 -!- SimonRC has joined.
00:23:23 -!- Pthing has quit (Remote host closed the connection).
00:24:25 <olsner> d'oh, I'm missing pieces, this isn't even TC
00:56:28 -!- cheater4 has joined.
00:57:06 -!- cheater3 has quit (Read error: Connection reset by peer).
01:12:03 -!- Pthing has joined.
03:58:04 -!- olsner has quit (Ping timeout: 260 seconds).
04:00:37 -!- olsner has joined.
04:01:51 -!- jcp has joined.
04:09:41 -!- jcp has quit (Read error: Connection reset by peer).
04:11:45 -!- jcp has joined.
04:27:06 -!- MissPiggy has quit (Remote host closed the connection).
05:15:11 -!- MizardX has quit (Read error: Connection reset by peer).
05:43:12 -!- Pthing has quit (Read error: Connection reset by peer).
05:44:13 -!- Pthing has joined.
05:45:06 -!- Gracenotes has quit (Ping timeout: 240 seconds).
05:47:21 -!- Gracenotes has joined.
06:04:09 -!- Pthing has quit (Remote host closed the connection).
06:08:39 -!- jcp has changed nick to javawizard.
06:29:16 -!- tombom has joined.
06:42:50 -!- javawizard has quit (Quit: I will do anything (almost) for a new router.).
06:44:06 -!- jcp has joined.
06:58:36 -!- Gracenotes has quit (Ping timeout: 272 seconds).
07:01:20 -!- Gracenotes has joined.
07:33:49 -!- FireFly has joined.
07:49:40 -!- tombom has quit (Quit: Leaving).
07:49:41 <oklopol> hehe, they're doing something with the pipes and absolutely no water should go in the drain; i've already managed to use the sink twice and flush the toilet once
07:55:25 -!- lifthrasiir has quit (Ping timeout: 265 seconds).
07:55:57 -!- lifthrasiir has joined.
07:56:14 -!- FireFly has quit (Quit: Leaving).
07:59:06 -!- Gracenotes has quit (Ping timeout: 252 seconds).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:00:40 -!- lifthrasiir has quit (Ping timeout: 272 seconds).
08:02:47 -!- lifthrasiir has joined.
08:07:03 -!- lifthrasiir has quit (Ping timeout: 240 seconds).
08:09:38 -!- Pthing has joined.
08:11:10 <fizzie> oklopol: You literally give a crap about their instructions, then.
08:16:33 -!- Asztal has quit (Ping timeout: 276 seconds).
08:18:24 <oklopol> actually it was pee, if you must know
08:18:31 -!- lifthrasiir has joined.
08:18:40 <oklopol> i probably would've realized during a longer operation
08:21:51 <fizzie> In retrospect, I'm not sure I actually wanted to start a conversation about this.
08:24:26 <oklopol> in this context, i'll interpret "retrospect" as "looking at ass".
08:24:54 <oklopol> although i guess that makes no sense
08:27:08 -!- lereah_ has joined.
08:27:08 <fizzie> With all retrospect due, I really wasn't looking at your ass.
08:28:35 <oklopol> it's fine
08:29:34 -!- lifthrasiir has quit (Ping timeout: 258 seconds).
08:30:37 <fizzie> I'm going to have to take your word for that, though I'm sure it is.
08:30:42 <lereah_> Does linux have a standard way to read djvu files?
08:30:48 <lereah_> Or do I need to get one
08:31:25 <Deewiant> I don't know of any "standard" way but viewers do exist
08:31:41 <oklopol> you want to get a way from here?
08:32:31 <lereah_> Not sure I can get one though
08:32:34 <lereah_> I be at zerk
08:32:37 <lereah_> werk*
08:32:47 <fizzie> Evince, which is everywhere, has some support for it, I think.
08:32:50 <lereah_> Not sure you can install anything on that computer
08:34:14 <fizzie> At least djvu's mentioned on http://projects.gnome.org/evince/ and Evince's probably installed on anything Gnomish.
08:35:15 <lereah_> I'm on scientific linux, is it gnomish?
08:35:37 <lereah_> Yeah, I think it is
08:35:48 -!- lifthrasiir has joined.
08:37:09 <lereah_> Well, I tried typing evince on the console and it didn"t work, and that's about all my knowledge of linux
08:37:16 <lereah_> How do I open it?
08:38:54 <fizzie> If it's there, "evince" in a terminal should run it. You may be out of luck there.
08:39:05 <lereah_> Balls.
08:39:27 <lereah_> i tried finding an online converter to pdf but no luck
08:39:39 <lereah_> Can someone convert the file to pdf for me?
08:39:59 <lereah_> It is http://membres.multimania.fr/bewulf/Russell/Quantum%20Mechanics%20And%20Path%20Integrals%20-%20R.%20Feynman,%20A.%20Hibbs.djvu
08:40:17 <fizzie> KDE's "Okular" supposedly supports djvu too.
08:40:19 <oklopol> ooh math
08:41:11 <lereah_> No, SCIENCE!
08:41:20 <lereah_> (it are physics)\
08:41:53 <oklopol> hmm oh quantum mechanics
08:41:54 <fizzie> oklopol: I'm at the lectures of a course from the maths department at the moment; so technically this is math too. (But it's about numeric methods, so maybe it doesn't count.)
08:42:01 <oklopol> i just read blah blah path integrals
08:42:30 <oklopol> ofc quantum mechanics is certainly more interesting than path integrals
08:42:46 <lereah_> It is pretty neato torpedo
08:42:50 <lereah_> See, the concept is
08:42:59 <oklopol> fizzie: yell "woohoo math!"
08:43:06 <lereah_> A particle in this theory actually travels on EVERY POSSIBLE PATH
08:43:40 <oklopol> so how exactly does the path integral come in?
08:44:16 <oklopol> and does it travel in R^3?
08:44:18 <lereah_> Well, you have to do a sum over all path to find the probabilities of finding the particle here or there
08:45:22 <oklopol> i guess we could take the space of all continuous paths, and give it a natural measure obtained by some sort of integral average
08:45:34 <oklopol> "sum over all path" isn't specific enough
08:45:54 <lereah_> Yeah, it is rarely easy to find
08:46:07 <lereah_> I'm doing one right now, for instance
08:46:32 <oklopol> i mean you can't sum over all the paths, there's an uncountable number of them
08:46:35 <lereah_> You decompose time in little increments, and for the paths, you do all the possible turns you can do
08:46:42 <lereah_> Then you take the limit
08:47:17 -!- Gracenotes has joined.
08:49:00 <oklopol> all the possible turns is still an uncountable amount; and is this in R^3?
08:49:15 <lereah_> Iunno
08:49:21 <lereah_> Here's the example
08:49:27 <oklopol> you don't know or you don't know what R^3 is?
08:49:39 <lereah_> It is in R^3
08:49:50 <lereah_> A particle goes at the speed of light in 1D
08:50:10 <lereah_> At every time increment, it can do a turn
08:50:29 <oklopol> yeah in 1D you can do it combinatorially
08:50:42 <lereah_> So to go from point A to point B, you sum over all the possible turns at every increment of tine
08:50:57 <lereah_> In the end it's ~ an exponential
08:51:15 <lereah_> (by the way can someone convert it to pdf)
08:51:41 <oklopol> what i would do is google "dvu to pdf"
08:51:48 <lereah_> I did
08:51:56 <lereah_> But I can't install softwares here
08:52:12 <oklopol> well and "online converter" ofc, i never install anything
08:53:09 <oklopol> nothing on the first few pages
08:53:22 <oklopol> i guess no one's invented a converter
08:56:12 <lereah_> (that is why I ask)
08:57:43 -!- oerjan has joined.
08:59:45 <oerjan> <oklopol> hehe, they're doing something with the pipes and absolutely no water should go in the drain; i've already managed to use the sink twice and flush the toilet once
08:59:53 <oerjan> never trust a human to change his habits
09:00:06 <oklopol> or think about what he's actually doing
09:00:53 <oklopol> i was thinking about teaching math to innocent children
09:01:15 <oerjan> _evil_ math i assume
09:01:53 <oklopol> it's not the math that's evil, it just makes humans evil
09:02:04 <oklopol> MWAHAHAH
09:02:23 <oerjan> it turns them into rational, uncaring beasts
09:03:10 <lereah_> Even irrational numbers?
09:03:38 <oerjan> it's a bit complex
09:04:21 <oklopol> i think it's natural enough
09:04:27 <oklopol> we need more number fields
09:04:42 <oklopol> ...with funny names
09:04:44 <oerjan> oklopol: what a surreal idea
09:06:47 -!- augur has quit (Ping timeout: 246 seconds).
09:07:58 -!- augur has joined.
09:08:44 <lereah_> There's plenty of number sets
09:09:00 <lereah_> Constructible, computable, compressible, definable
09:09:22 <oklopol> and how are those names funny
09:09:41 <oklopol> that's the problem with contemporary mathematics, things aren't given funny names
09:09:59 <oklopol> and what's a contructible number
09:10:11 <oklopol> ohhh
09:10:22 <oklopol> right ofc i know those
09:11:13 <lereah_> Also algebraic and transcendental
09:11:16 <lereah_> Shit like dat
09:12:52 <oklopol> what's the definition of a compressible number?
09:13:13 <oklopol> i think you have to be careful or it might not be a field
09:13:32 <lereah_> A number you can write a finite sized program for
09:13:52 <oklopol> oh. then what's the definition of computable
09:13:55 -!- kar8nga has joined.
09:14:00 <oklopol> i thought you'd have to be able to compress it
09:14:16 <lereah_> Lemme check
09:14:50 <lereah_> Hm, I might remember the wrong name for it
09:15:46 <oerjan> no, finite sized program applies to compressible strings, not real numbers, i think
09:15:56 <oerjan> and the program must be shorter than the string
09:18:25 <oklopol> i wonder if there are puns involving Z[sqrt(2)]
09:18:37 <oerjan> compressible number doesn't seem to be a real term
09:18:46 <oklopol> or Z(sqrt(2)) if we want to emphasize fieldness i suppose
09:19:07 <lereah_> It might have been incompressible
09:20:12 <oerjan> no better
09:20:39 <lereah_> http://homepages.cwi.nl/~paulv/schedule06/intro.ps
09:20:57 <oerjan> if i don't put quotes around, google wants to give me (un)compressible flows instead
09:21:33 * oerjan doesn't do postscript
09:22:53 <oklopol> Unrecoverable error: configurationerror in setpagedevice
09:22:54 <oklopol> Operand stack:
09:22:54 <oklopol> false --nostringval--
09:22:54 <oklopol> Failed to open device or install ViewerPreProcess hook: returns -26
09:23:02 <oklopol> me neither, it seems
09:23:17 -!- GreaseMonkey has quit (Quit: HydraIRC -> http://www.hydrairc.org <- Nobody cares enough to cybersquat it).
09:28:19 -!- jcp has quit (Quit: I will do anything (almost) for a new router.).
09:42:40 <lereah_> oklopol : http://arxiv.org/pdf/hep-th/9302097v1
09:42:51 <lereah_> If you want the path integral mathematically
09:43:04 <lereah_> It describes it a bit
09:43:16 <lereah_> With the hilarious wiener measure
09:43:30 <oklopol> ah, now there's a funny name
09:45:04 <lereah_> It was waiting for you
09:45:35 <lereah_> ":This calculus in functional space (“Wiener measure”) attracted
09:45:36 <lereah_> several mathematicians"
09:45:42 <lereah_> They love wiener measures
09:47:41 <oklopol> "this file is damaged and could not be viewed"
09:47:43 <oklopol> :P
09:47:55 <lereah_> Nigga you got file proble;s
10:09:16 -!- Pthing has quit (Remote host closed the connection).
10:30:26 -!- oerjan has quit (Quit: leaving).
10:44:22 -!- iamcal has joined.
10:45:48 -!- cal153 has quit (Ping timeout: 256 seconds).
10:46:24 -!- comex has quit (Ping timeout: 260 seconds).
10:46:29 -!- comex has joined.
11:49:05 -!- MizardX has joined.
12:34:19 <lereah_> How can I check what Linux I have?
12:35:30 <fizzie> "lsb_release -a" if it happens to support that LSB thing, but not all do.
12:35:54 <lereah_> k
12:36:05 <fizzie> If it does, the "Description" field is most likely a human-readable sensible thing.
12:36:29 <lereah_> Scientific Linux
12:36:32 <lereah_> Damn it
12:36:43 <lereah_> I guess it's close to some other linux
12:36:46 <lereah_> Let's check
12:37:28 <fizzie> It seems to be at least slightly related to Red Hat Enterprise Linux.
12:38:15 <lereah_> It is a free and open source operating system based on Red Hat Enterprise Linux and aims to be 100% compatible with and based on Red Hat Enterprise Linux.
12:40:27 <fizzie> You talk like a free, web-based, collaborative, multilingual encyclopedia project.
12:40:51 <lereah_> I sure do
12:40:56 <lereah_> Forgot the quotes
12:41:09 <lereah_> I don't like to type too much here, the keyboard is qwerty
12:41:49 <oklopol> ugh, who still uses qwerty
12:42:09 <oklopol> well i do but you know that's not relevant.
12:42:17 <fizzie> ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvxyz ,.-?!";: -- there, you can copy-paste most of the things you'll ever need out of that; no need for typing.
12:42:24 <lereah_> France does not
12:42:58 <fizzie> Visited CERN once, the azerty keyboards in the terminals there were confusing.
12:43:29 <oklopol> isn't azerty even stupider
12:43:58 <lereah_> Well, confirmed : I can't install a software here without admin priviledges
12:48:24 <oklopol> they're drilling in harmonic major second outside, i love it
12:49:21 <lereah_> drill baby drill
12:49:24 <oklopol> haha, a tritone xD
12:49:57 <oklopol> i can't concentrate with all this beautiful music around me
12:50:26 <lereah_> Put on some shitty music then
12:51:15 <oklopol> or i could just not concentrate
12:56:58 <fizzie> Debian package single-line descriptions are sometimes less than informative; what'd you expect from "gaia - interface to the planet"?
12:57:17 <lereah_> Some Pandora shit
12:57:28 <lereah_> Just plug in your ponytail
12:57:50 <fizzie> It is, in fact, "an Earth viewer that lets you navigate around the globe and zoom into almost any place".
12:58:02 <fizzie> I'm not sure how that's interfacing with the planet, but.
12:59:19 <oklopol> okay what am i wearing
12:59:42 <oklopol> well okay that's not fair, i'm not outside
13:01:54 <fizzie> It would be much more interfacingy if I could also change what you're wearing.
13:02:06 <oklopol> hmm, true
13:02:28 <oklopol> that sounds really scary after your questions about my ass
13:02:46 <oklopol> or how did the conversation go again
13:22:23 -!- deschutron has joined.
13:22:48 <deschutron> RIP
13:24:45 <deschutron> java.sun.com is there, but they changed the colours, and it looks less welcoming.
14:08:51 -!- oklopol has quit (Ping timeout: 276 seconds).
14:49:33 -!- MissPiggy has joined.
15:07:17 -!- augur has quit (Ping timeout: 246 seconds).
15:23:31 -!- cpressey has joined.
15:33:22 -!- coppro has quit (Quit: I am leaving. You are about to explode.).
15:40:39 -!- augur has joined.
15:52:34 -!- FireFly has joined.
16:00:09 <Gregor> deschutron: Soon it'll forward to oracle.com/ogppl , and be renamed to the Oracle General Purpose Programming Language
16:00:45 <Gregor> Also, Solaris will be renamed to Oracle System Services
16:01:12 <Gregor> And they'll try to rename JavaScript to OracleScript
16:01:55 <Gregor> Then they'll rename VirtualBox to Oracle Virtual System and MySQL to the Oracle Database Sy--no wait.
16:02:55 <Gregor> MySQL will be renamed Oracle Database Demo :P
16:03:44 <Sgeo_> Can't MySQL just be forked?
16:03:53 <Gregor> Sure, but why? PostgreSQL is better.
16:04:15 <Sgeo_> Easier on systems already using MySQL?
16:04:34 <Gregor> I'm sure Oracle provides a very handy upgrade path :P
16:04:41 <deschutron> "Oracle General Purpose Programming Language" haha
16:05:49 <deschutron> yeah they could make a fork called OurSQL or something
16:05:53 <Sgeo_> Hey WOT plugin on Chrome: I love how long it takes you to warn me that I'm on a distrusted site
16:06:31 <deschutron> MySQL fans, that is
16:07:56 -!- puzzlet has quit (Ping timeout: 265 seconds).
16:10:25 -!- kar8nga has quit (Remote host closed the connection).
16:10:32 -!- kar8nga has joined.
16:11:00 -!- kar8nga has quit (Remote host closed the connection).
16:11:30 -!- kar8nga has joined.
16:20:53 -!- lereah_ has quit (Remote host closed the connection).
16:33:26 -!- kar8nga has quit (Remote host closed the connection).
16:33:47 -!- puzzlet has joined.
16:35:29 -!- BeholdMyGlory has joined.
16:36:24 -!- deschutron has left (?).
16:44:45 -!- scarf has joined.
16:53:45 <AnMaster> phew
16:54:09 <AnMaster> 1.5 dm snow during the night to today, then when I got home, another dm of snow
16:55:51 <AnMaster> also: open and mostly flat landscape around here and strong winds
16:56:04 <AnMaster> so really really bad roads today
16:58:02 <scarf> AnMaster: who measures in decimetres?
16:58:27 <scarf> IMO, 0.15m, 15cm, and 150mm would all be more plausible units there
17:02:04 <AnMaster> scarf, err, dm isn't all that uncommon in Sweden
17:02:33 <AnMaster> as in "a few decimetres" or such.
17:02:54 <AnMaster> scarf, oh and there is a class 1 warning for this part of the country
17:03:06 <AnMaster> (note: higher numbers are worse, it goes up to 3)
17:03:39 <scarf> hmm, 25cm of snow would be enough for a severe weather warning here in the UK
17:03:46 <scarf> but then, we're less used to large amounts of snow
17:04:20 <Gregor> Sweden.
17:04:23 <Gregor> Sweden is so weird.
17:04:33 <AnMaster> scarf, also, have you ever been traveling in an articulated bus that met another articulated bus on a already narrow street made yet more narrow due to the roughly 1.5 meter high heaps of snow on the sidewalks (also they extended a fair bit out in the street)
17:04:50 <AnMaster> scarf, really, this winter has been unusually bad
17:04:52 -!- tombom has joined.
17:04:54 <scarf> no, articulated busses are rare enough here as they are, and I don't think I've ever seen two at once
17:04:59 <scarf> *buses
17:05:15 <AnMaster> scarf, well same line, so two buses on the same line met
17:05:23 <AnMaster> really they had a hard time managing (I was in one)
17:05:55 <AnMaster> scarf, anyway what about non-articulated buses.
17:06:15 <scarf> AnMaster: oh, I see those all the time
17:06:23 <scarf> the record was something like 20 at once, that was a fun day
17:06:34 <AnMaster> scarf, I meant: the above scenario for other buses...
17:06:45 <AnMaster> or maybe you don't get enough snow to have a street made so narrow due to snow...
17:06:48 <scarf> basically, something had fallen off the roof of one of the big new skyscrapers in Birmingham and landed on one of the most major roundabouts there
17:06:55 <AnMaster> heh
17:07:02 <scarf> they had to close the roundabout as a result, while they did safety checks
17:07:19 <scarf> and pretty much every car driver avoided the city centre that day, but the busses didn't
17:07:26 <scarf> and they somehow managed to all be clumped in a really huge line
17:07:35 <scarf> nothing for half an hour, then 20 at once, on a huge range of different routes
17:08:33 <AnMaster> right, but that is still not quite like the scenario I described. Also it was snowing heavily at that point and the visibility was poor.
17:09:16 <AnMaster> scarf, oh and skyscraper, how many stories does it need to be that
17:09:22 <AnMaster> rather than just a tall house
17:09:37 <scarf> AnMaster: not sure, I think it's a relative term
17:09:51 <scarf> but it's one of the tallest buildings in Birmingham, many tens of storeys
17:09:53 <scarf> *stories
17:10:35 <AnMaster> could 15 stories or so be a sky-scraper anywhere?
17:10:42 <AnMaster> that is about the tallest around here I think
17:11:08 <AnMaster> and here = not just this town, but also the city a bit away
17:11:31 <AnMaster> in this town, I guess around 5 stories, maybe 7 in some of those newer houses...
17:12:52 <scarf> not sure
17:13:49 -!- oerjan has joined.
17:13:55 -!- cheater has quit (Ping timeout: 245 seconds).
17:14:44 <oerjan> dammit no oklopol
17:15:11 -!- oerjan has set topic: RIP sun.com | 16 days since last ehird sighting | http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
17:16:06 <cpressey> There's something I've wondered for a while but never got clear on. If I compile with "gcc -ansi", does that necessarily preclude using any and all POSIX calls?
17:16:21 <AnMaster> scarf, btw, the snow outside the university reached above my knees outside one of the doors I had to reach
17:16:30 <AnMaster> due to the wind blowing it up against the house
17:16:39 <AnMaster> not nice
17:16:40 <scarf> cpressey: the header files take clues from the compiler as to which POSIX calls they should include
17:16:52 <AnMaster> oerjan, bad weather in norway too?
17:16:58 <scarf> by default, I think gcc tells them to use a GNUish set; if you say -ansi, it tells them to use an ANSI set
17:17:04 <AnMaster> oerjan, some 2.5 dm or so since yesterday here
17:17:07 <AnMaster> and still snowing
17:17:08 <scarf> but you can write your own #defines in order to select one yourself
17:17:40 <oerjan> AnMaster: no new snow here, although below -10 temperatures, no wind to speak of. i don't really call that bad...
17:17:42 <cpressey> scarf: Thanks.
17:17:56 <AnMaster> oerjan, well strong wind here...
17:18:09 <scarf> cpressey: man 7 feature_test_macros describes the whole mess
17:18:26 <AnMaster> oerjan, official warning (lowest level) from SHMI and such
17:19:06 <oerjan> i saw some warnings on yr.no in weeks previous, although i think trondheim was mostly missed
17:19:26 <cpressey> scarf: I'd rather not get into that :) I guess my question becomes, is there any sanctioned way in -ansi (-std=c89) to sleep for a given number of milliseconds?
17:19:39 <AnMaster> oerjan, lucky you
17:19:43 <scarf> no, milliseconds are too fine-grained, and that's actually impossible in DOS
17:19:54 <scarf> (the minimum sleep length is 1/19.2 seconds in DOS, for some reason I don't fully understand)
17:20:08 <oerjan> we did have approx -20 temperatures for a short while, though
17:20:11 <AnMaster> oerjan, fun yr.no claims it isn't snowing around here any longer
17:20:18 <AnMaster> looking out I have to strongly disagree
17:20:32 <oerjan> true, it's not always accurate
17:20:50 <AnMaster> oerjan, well, they could update it with current data as well
17:20:56 -!- fungot has quit (*.net *.split).
17:20:57 -!- ineiros has quit (*.net *.split).
17:20:58 -!- pikhq has quit (*.net *.split).
17:20:58 <cpressey> scarf: I believe that's the frequency of one of the timer chips from the original PC architecture. But anyway -- OK, I might have to take a different approach.
17:21:11 <AnMaster> oerjan, also it is generally better than most other ones I know of
17:21:18 <AnMaster> thus when it isn't better it is fun to poke fun at it
17:21:21 <scarf> if you want a guaranteed millisecond sleep, you could require POSIX 2001 (that's #define _POSIX_C_SOURCE 200112L) and use select
17:21:46 <oerjan> oh do you look at yr.no often too? :D
17:21:50 <scarf> or you could require SUSv2 compatibility (#define _XOPEN_SOURCE 500) and use usleep
17:21:52 <AnMaster> <scarf> (the minimum sleep length is 1/19.2 seconds in DOS, for some reason I don't fully understand) <-- can't you change it iirc
17:21:54 <cpressey> scarf: I don't really need for it to be guaranteed. Actually, I don't really need for it to be compiled with -ansi, except for a perverse sense of pride.
17:21:57 <AnMaster> setting the PIT to a higher freq
17:22:00 <AnMaster> or something such
17:22:30 <oerjan> it's in cooperation with the norwegian meteorological institute, you'd imagine they had _some_ clue
17:22:35 <scarf> cpressey: if going for maximum portability, you could abstract out the sleep to a separate file and have it with a bunch of ifdefs for sleeps on differnt platforms
17:22:40 -!- fungot has joined.
17:22:41 -!- ineiros has joined.
17:22:41 -!- pikhq has joined.
17:23:11 <AnMaster> oerjan, nah, they just check the weather stone
17:23:20 <scarf> *different
17:23:28 <cpressey> scarf: Not maximum portability, but yes, I'm thinking now to just #ifdef on whether -ansi was specified. If so, approximate sleeps to the nearest second ('cos I think C89 has something for that)
17:23:41 <scarf> yep, time
17:24:12 <oerjan> :D
17:24:30 <cpressey> Then I can say "compiles with -ansi -pedantic" and satisfy my sick sense of pride in that. Just need to discover that that define is. Whee!
17:24:46 <scarf> cpressey: #ifdef __STRICT_ANSI__
17:24:52 <scarf> the feature test macros work both ways round!
17:25:11 -!- Asztal has joined.
17:25:12 <cpressey> And scarf will even volunteer that information. This must be my lucky day!
17:25:14 <scarf> although, I think that's a glibc definition
17:25:22 <scarf> which is a really perverse sense of portability
17:25:47 <scarf> (if you wanted to be less perverse, you'd instead check that, say, _POSIX_C_SOURCE /wasn't/ set)
17:27:20 <cpressey> Yes.
17:27:34 -!- cheater has joined.
17:32:01 <AnMaster> cpressey, -ansi is C89 iirc?
17:32:19 <cpressey> AnMaster: Yup.
17:32:28 <AnMaster> cpressey, so what is wrong with C99
17:32:37 <cpressey> The manpage even says it's equivalent to "-std=c89".
17:32:53 <AnMaster> don't claim it jumped the shark, it didn't. The difference to C89 aren't that large
17:32:54 <cpressey> AnMaster: Nothing's "wrong" with C99, I just don't want to require it.
17:33:01 <AnMaster> cpressey, meh
17:33:15 <cpressey> No, it didn't jump the shark :)
17:33:20 <AnMaster> everything supports the most useful subset of C99 these days. Windows doesn't count (and on windows icc does it)
17:33:24 <AnMaster> (and mingw)
17:33:27 <AnMaster> (and so on)
17:33:51 <cpressey> AnMaster: I guess you did not see me type "sick sense of pride" twice above?
17:33:56 <scarf> hardly anything supports all of C99, though
17:34:05 <AnMaster> scarf, sure, complex numbers thingy and such
17:34:17 <scarf> AnMaster: I know you dislike python 3, but at least it exists
17:34:19 <scarf> C99 doesn't, really
17:34:25 <scarf> it's just a standard that only half caught on
17:34:32 -!- rodgort has quit (Quit: Coyote finally caught me).
17:34:38 <AnMaster> "most useful subset" = stdint.h, restrict, stdbool.h, inline, // comments, for (int i ...) and a few such things
17:34:50 <AnMaster> oh and .fieldname = foo initialisers
17:34:54 -!- rodgort has joined.
17:35:23 <cpressey> Hey, at least there's only one C99. There's like three RSS 2.0's.
17:35:59 <scarf> there are two C89s, though
17:36:01 <scarf> well, C89 and C90
17:36:10 <scarf> hilariously, they're identical except for the section numbering
17:36:14 <scarf> which makes them rather hard to quote
17:36:32 <AnMaster> scarf, you just *love* to mention that right?
17:36:48 <scarf> AnMaster: not massively, but it's worth mentioning in this context
17:36:50 <AnMaster> I mean, I heard it so many times
17:36:53 <AnMaster> from you
17:37:33 <scarf> AnMaster: I don't think I've said it that many times (anyone care to grep the logs?), and it probably wasn't aimed at you each time
17:38:25 <AnMaster> not aimed at at me every time no
17:38:54 <AnMaster> scarf, also hard to grep for due to exact wording differing
17:41:15 <AnMaster> actually about that amount of snow, other member of household got home, apparently I missed out on two rounds of clearing away the snow during the day. So add about 20 cm to that above...
17:42:09 -!- MigoMipo has joined.
17:42:15 <AnMaster> scarf, so around 45 cm of snow then
17:42:24 <scarf> that's a lot for one day
17:43:11 <AnMaster> scarf, quite, 10-15 cm being more normal "lot of snow for one day" around here
17:44:45 <AnMaster> this has been both the coldest and the snowiest winter for a long time
17:44:56 <AnMaster> I think it even beat that amount some 5 years ago or so
17:51:47 -!- tombom has quit (Ping timeout: 246 seconds).
18:06:21 -!- tombom has joined.
18:11:49 <AnMaster> scarf, oh btw, about 40 cars has gone over the side of the roads in this <something that iirc is roughly eqiv in size to a county in UK>. Due to the extremely slippery roads, partly caused by snow blowing across them.
18:25:14 <cheater4> is there a funge like befunge-93 but with inserting and removing rows/columns?
18:25:49 <cheater4> and is it tc?
18:26:45 <scarf> cheater4: ooh, I wrote something a bit like that
18:26:52 <scarf> now I'm trying to remember the name
18:26:55 <scarf> http://esolangs.org/wiki/Shove
18:27:02 <scarf> no, not that
18:27:03 <cheater4> 'a bit'?
18:27:07 <cheater4> 'splain
18:27:21 <oerjan> sprain your brain
18:27:22 <cheater4> lol
18:27:46 <cheater4> i sprain my brain when i'm reading the mysql manual.
18:27:57 <AnMaster> cheater4, use postgresql
18:28:00 <scarf> doesn't seem to be online, how strange
18:28:49 <cheater4> fuck sql and their non-deterministic totalitarian regime
18:29:06 <oerjan> Shove is conspicuously absent
18:29:07 <cheater4> scarf: just write it again, what could be so hard. :p
18:29:42 <scarf> reference interp: http://pastebin.ca/1786097
18:29:47 <scarf> strangely, I seem to never have written a spec
18:29:49 <scarf> but I believe it's TC
18:30:45 <cheater4> aha
18:30:48 <cheater4> so it is shove after all?
18:30:53 <scarf> yes, just not online
18:31:11 -!- kar8nga has joined.
18:31:38 <cheater4> jesus christ perl
18:31:39 <cheater4> i hate perl
18:32:02 <scarf> why?
18:32:28 <cheater4> why not?
18:32:36 <cheater4> i hate it because i don't know it.
18:32:40 <MissPiggy> if your having perl problems I feel bad for you son, I got 99 programming problems but regex ain't one
18:32:46 <cheater4> isn't that good enough for you?
18:32:55 <cheater4> i don't have problems with regex
18:33:04 <cheater4> but RAP (regex ain't perl)
18:33:18 <oerjan> MissPiggy: that a quote?
18:33:22 <MissPiggy> yes
18:33:25 <MissPiggy> im rapping about perl
18:33:28 <cheater4> it's from an old blue song
18:33:31 <cheater4> blues
18:33:47 <oerjan> um i mean literally
18:33:51 <MissPiggy> no
18:33:53 <MissPiggy> it's modified
18:35:26 -!- Gracenotes has quit (Read error: Connection reset by peer).
18:36:01 -!- impomatic has joined.
18:36:05 <impomatic> Hi :-)
18:36:10 <oerjan> i note that googling "if you're having perl problems" gives several relevant hits, but not with "regex"
18:37:09 <AnMaster> cheater4, MissPiggy: to add some context to the above. in Swedish "rap" means "burp" (the verb form would be "att rapa" - "to burp")
18:37:11 <oerjan> i conclude that regexes _are_ a problem in perl ;D
18:37:28 <MissPiggy> There are rappers with guns after me
18:37:29 <MissPiggy> Enemies that want to make sure I'm dead
18:37:29 <MissPiggy> Rap critics that say, "he only cares about money and women"
18:38:24 <oerjan> we scandinavians frequently rape after dinner </old joke>
18:39:38 <oerjan> heck if it's old _enough_, it may not have been a joke when it started
18:42:23 <impomatic> Is anyone planning to enter the CROBOTS tournament? http://crobots.deepthought.it/home.php?link=91
18:45:39 <impomatic> Hmmm... not much BF_Joust action :-(
18:45:54 <scarf> unfortunately not
18:48:42 <pikhq> GCC hates tail-call optimisation when longjmp is in use.
18:57:37 <AnMaster> pikhq, tell me, what are you trying to do....
18:57:40 -!- cheater has quit (Ping timeout: 245 seconds).
18:57:49 -!- cheater99 has joined.
18:57:50 <AnMaster> pikhq, are you trying to implement call/cc?
18:58:00 <AnMaster> if so, arrgh, and that won't work after you returned
18:58:17 <pikhq> AnMaster: call/cc with continuation-passing style.
18:58:38 <pikhq> Meaning no returning.
18:58:51 <pikhq> It works, GCC just doesn't tail-call optimize it at all.
19:00:05 <MissPiggy> dude if you wantCWCC don't use longjmp
19:00:44 <pikhq> MissPiggy: But it works just fine if you write continuation-passing style.
19:00:56 <pikhq> This means that functions never, ever return.
19:01:05 <AnMaster> pikhq, so longjmp() isn't tail call optimised?
19:01:36 <pikhq> AnMaster: No, functions *with them* aren't tail call optimised.
19:02:12 <pikhq> A tail call to fact becomes "call fact;addq $248, %rsp;ret"
19:02:51 -!- Sgeo__ has joined.
19:02:58 <pikhq> And a tail call to longjmp becomes "call longjmp".
19:03:09 <pikhq> Not "longjmp and then ret", just
19:03:13 <pikhq> "call longjmp".
19:03:43 <pikhq> __attribute__((noreturn)) makes GCC fecking stupid.
19:03:56 <AnMaster> pikhq, well longjmp doesn't return
19:04:07 <pikhq> AnMaster: Yes. "jmp longjmp" would be valid.
19:04:24 <pikhq> That would be a tail-call optimisation.
19:04:35 <scarf> my guess is that longjmp is coded in such a way that it expects to be called by a call
19:04:36 <AnMaster> pikhq, well since it restores the stack, the stack growing is not an issue
19:04:50 <AnMaster> scarf, iirc longjmp is a builtin for gcc
19:04:59 <pikhq> AnMaster: It's not.
19:05:01 <AnMaster> that is just iirc though
19:05:05 <AnMaster> pikhq, hm? really?
19:05:05 <pikhq> I'm looking at the assembly.
19:05:12 <scarf> pikhq: it's a bit of both
19:05:18 <pikhq> Builtins get handled differently.
19:05:26 <scarf> it's special-cased, and so is a builtin in that sense (and you can call it __builtin_longjmp if you like)
19:05:32 <scarf> but it ends up being turned into a cal in the asm
19:05:34 <pikhq> It's almost certainly part of libgcc.a.
19:05:36 <AnMaster> ah it seems to be libc
19:05:37 -!- Sgeo_ has quit (Ping timeout: 264 seconds).
19:05:41 <scarf> *call
19:05:46 <AnMaster> extern void _longjmp (struct __jmp_buf_tag __env[1], int __val)
19:05:46 <AnMaster> __THROW __attribute__ ((__noreturn__));
19:05:46 <pikhq> But it's not a builtin.
19:05:54 <pikhq> The builtins are invariably inlined.
19:06:08 <AnMaster> pikhq, then what about libgcc
19:06:09 <scarf> pikhq: it has properties of a builtin, and properties of a nonbuiltin
19:06:30 <AnMaster> libgcc_s even
19:07:14 <pikhq> AnMaster: Mmm.
19:07:41 <AnMaster> or maybe that is just for unsupported instructions kind of stuff (like 64-bit division on 32-bit x86)
19:07:55 <pikhq> scarf: Anyways. It seems that longjmp is part of libc.
19:08:30 <scarf> I remember that with gcc-bf, it's in the libc, but has a crazy calling convention that's different from the normal one
19:08:31 -!- |MigoMipo| has joined.
19:08:33 <scarf> hmm, or is that setjmp?
19:08:43 <pikhq> setjmp is the crazy one.
19:08:48 <scarf> ah yes, it's setjmp that needs to be weird, longjmp can just be a normal function call
19:08:49 -!- cheater99 has quit (Read error: Connection reset by peer).
19:09:44 -!- cheater99 has joined.
19:09:45 <pikhq> Among other things, it's declared __attribute__((returns_twice)).
19:10:26 <scarf> that's a beautiful attribute
19:10:32 <scarf> and not one that's officially documented, I think
19:10:36 <scarf> besides, it can return more than twice
19:11:01 -!- MigoMipo has quit (Ping timeout: 264 seconds).
19:11:07 <pikhq> "Returns twice" just means "there is more than one return here. Make sure all registers are dead."
19:11:41 <AnMaster> scarf, err, setjmp has the "return twice attribute" yeah
19:11:55 <AnMaster> scarf, but what other strange calling convention would there be
19:12:16 <scarf> gah, now I'm going to have to look at libbf/setjmp.s
19:12:17 <AnMaster> <scarf> and not one that's officially documented, I think <-- yes it is
19:12:25 <AnMaster> I'm 100% sure I read the docs
19:12:50 <AnMaster> pikhq, isn't vfork() also returns_twice?
19:12:54 <AnMaster> not sure about fork()
19:12:57 <pikhq> AnMaster: Yeah.
19:13:03 <AnMaster> even though it returns twice it happens in different processes
19:13:05 <pikhq> fork() isn't.
19:13:15 <pikhq> It returns twice, sure, but once per stack.
19:13:20 <AnMaster> pikhq, but... on linux vfork() is just fork() iirc?
19:13:25 <AnMaster> as in, it is mapped like that
19:13:32 <scarf> AnMaster: not quite
19:13:33 <AnMaster> since linux does COW anyway
19:13:39 <scarf> vfork messes with the scheduling on linux
19:13:45 <AnMaster> scarf, in what way?
19:13:49 <scarf> such that the parent won't run until the child execs
19:13:54 <AnMaster> well yes
19:13:55 <pikhq> AnMaster: No.
19:13:58 <AnMaster> so you need a mutex or such
19:14:03 <AnMaster> to make it wait
19:14:07 <scarf> it's safe under the same circumstances as it's safe on non-cow systems
19:14:11 <pikhq> AnMaster: vfork() halts the parent until the child exits.
19:14:16 <scarf> pikhq: or execs
19:14:20 <scarf> which is the normal use-case
19:14:30 <pikhq> By any of: _exit, fatal signal, or exec.
19:14:38 <AnMaster> pikhq, or _Exit()
19:14:41 <pikhq> scarf: The *sane* use-case, you mean. :P
19:14:54 <scarf> pikhq: the *portable* use-case, I mean
19:15:03 <scarf> because vform means something rather different on, say, SunOS
19:15:05 <scarf> *vfork
19:15:05 <pikhq> AnMaster: _Exit is an alias for _exit.
19:15:15 <scarf> pikhq: different standards
19:15:39 <AnMaster> pikhq, well, _Exit is C99 too
19:15:43 <AnMaster> _exit() is posix
19:16:06 <pikhq> Ah.
19:16:17 <pikhq> Really should look at the feature test macros in the man page more.
19:16:28 <AnMaster> beats me why they didn't just use _exit() in C99
19:16:36 <AnMaster> after all, the behaviour is identical
19:16:38 <pikhq> POSIX requires that _Exit be an alias for _exit.
19:16:45 <AnMaster> well yes
19:17:05 <pikhq> But it makes no sense for the two of them to exist.
19:17:09 <AnMaster> huh
19:17:11 <AnMaster> wth
19:17:18 <AnMaster> okay this is strange
19:17:24 <AnMaster> 3p comes before 2 in my man
19:17:25 <AnMaster> on arch
19:17:33 <AnMaster> reverse (and saner) way on ubuntu
19:17:45 <AnMaster> now to find where to fix it on arch...
19:18:20 <AnMaster> ah /etc/man_db.conf
19:23:27 -!- MizardX has quit (Ping timeout: 276 seconds).
19:31:31 -!- tombom_ has joined.
19:33:19 <pikhq> I've got half a mind to do this with getcontext and setcontext.
19:33:23 <pikhq> (oh, the evil I can do!)
19:34:22 -!- oklopol has joined.
19:34:39 <AnMaster> pikhq, they are removed and/or deprecated in POSIX 2008 iirc
19:34:41 -!- tombom has quit (Ping timeout: 258 seconds).
19:34:51 <AnMaster> ah yes
19:34:53 <AnMaster> man page says:
19:34:56 <AnMaster> " SUSv2, POSIX.1-2001. POSIX.1-2008 removes the specification of getcontext(), citing portability issues, and recommending that applications be rewritten to use POSIX threads instead."
19:34:58 <scarf> pikhq: it's because _exit is in user namespace in C89
19:35:05 <scarf> rather than implementation namespace
19:35:16 <scarf> so adding it might have broken existing C89 programs that used it as a variable name (!)
19:35:20 <AnMaster> scarf, so what about printf()?
19:35:24 <AnMaster> isn't that user name space?
19:35:32 <pikhq> AnMaster: But POSIX threads aren't a replacement.
19:35:39 <AnMaster> pikhq, not my fault
19:35:49 <pikhq> I know, just debating.
19:35:58 <scarf> AnMaster: yes, unless you include stdio.h
19:36:04 <pikhq> Pthreads are... Threading. getcontext et al are continuations, with which coroutines can be done.
19:36:12 <pikhq> (and thereby userspace 'threads')
19:36:28 <AnMaster> pikhq, you could simulate getcontext/setcontext with posix threads and mutexes
19:36:34 <scarf> presumably, they had a choice of making you #include <c99stdio.h> to get _exit(), or just calling it _Exit() and putting it in a preëxisting header
19:36:37 <pikhq> o.O
19:36:38 <AnMaster> scarf, ....
19:36:43 <pikhq> Yes, you could.
19:36:52 <pikhq> Dear lord.
19:36:58 <AnMaster> pikhq, what?
19:37:04 <pikhq> This would be the most evil fibonacci sequence implementation ever.
19:37:10 <AnMaster> what?
19:37:13 <scarf> pikhq: I did something similar in continuation.i
19:37:20 <scarf> using INTERCAL threads to simulate continuations
19:37:24 <pikhq> Continuation-passing style using POSIX threads.
19:37:29 <AnMaster> pikhq, wonderful!
19:37:53 <AnMaster> pikhq, a bit irritating on implementations not using user space threads
19:38:02 <AnMaster> on m:n mappings it shouldn't be too bad
19:38:11 <AnMaster> on 1:1 mappings, well, it would be bad
19:38:24 <pikhq> AnMaster: Could just link against GNU userspace threads for those platforms.
19:38:32 <AnMaster> pikhq, such as linux?
19:38:38 <AnMaster> linux uses 1:1 threads
19:38:54 <AnMaster> freebsd used to use m:n but switched to 1:1 too
19:39:02 <AnMaster> most other are 1:1 I think
19:39:13 <pikhq> NPTL uses either 1:1 or m:n.
19:39:21 <AnMaster> ntpl can use m:n?
19:39:23 <AnMaster> link?
19:39:30 <AnMaster> I never heard about this before
19:39:34 <pikhq> Not sure how to switch it over to m:n, but Wikipedia says it can.
19:39:44 <AnMaster> nptl*
19:39:46 <pikhq> Argh.
19:39:52 <AnMaster> pikhq, what?
19:39:53 <pikhq> No it doesn't.
19:40:03 <pikhq> "An alternative to NPTL's 1×1 model is the m×n model where there are typically more userland threads than schedulable entities. In the m×n implementation, the threading library is responsible for scheduling user threads on the available schedulable entities; "
19:40:10 <pikhq> That's just saying "this is possible".
19:40:11 <pikhq> -_-'
19:40:13 <AnMaster> yeah
19:40:35 <AnMaster> but linux doesn't do it
19:40:53 <pikhq> GNU Portable Threads are M:1.
19:41:15 <pikhq> Slightly inefficient, but work just about everywhere.
19:44:16 <oerjan> (i think there was a big error in it when i found it, which i removed :D)
19:44:25 <oerjan> darn windows
19:44:30 <AnMaster> oerjan, big error in what?
19:44:54 <oerjan> well since it ended up here anyhow, in http://en.wikipedia.org/wiki/Formally_real_field
19:45:34 -!- BeholdMyGlory has quit (Remote host closed the connection).
19:45:37 -!- kar8nga has quit (Remote host closed the connection).
19:45:47 <pikhq> AnMaster: Well, doing that I would have *one* advantage.
19:45:57 <pikhq> Writing in a functional style makes everything automatically reentrant.
19:46:02 <pikhq> Mmm, immutable values.
19:49:21 <pikhq> Though, if I did everything with getcontext et al, I could do stuff like have garbage-collected stacks. Which amuses me.
19:49:46 <cpressey> Doing continuation-passing style with pthreads would be a bit like doing it with bash shell calling "exec $0"... no?
19:50:12 <pikhq> cpressey: Remarkably similar.
19:52:09 <cpressey> I should implement that idea I had for "continuators" a while ago, even though it was really vague and unimplementable. Something about passing continuations between processes/threads, and calling a service to provide new continuations...
19:52:34 <cpressey> I think I just liked the term "continuator".
19:52:55 <AnMaster> pikhq, how could you gc the stacks?
19:53:04 <AnMaster> pikhq, also I doubt boehm-gc would like it very much
19:53:12 <pikhq> AnMaster: Stacks are explicitly allocated.
19:53:34 <AnMaster> pikhq, boehm-gc would go mad
19:53:39 <pikhq> You literally have to allocate a stack and stick it into a context for creating a new context.
19:54:36 <AnMaster> pikhq, what if you realloc() it?
19:54:36 -!- GreaseMonkey has joined.
19:54:40 <AnMaster> or anyway
19:54:44 <AnMaster> how do you grow it
19:55:09 -!- BeholdMyGlory has joined.
19:55:13 <pikhq> Generally, you just allocate a sufficiently large stack.
19:55:32 <pikhq> But it's normally allocated memory.
19:55:39 <pikhq> Which happens to be where the stack pointer is.
19:55:51 <GreaseMonkey> 'lo.
19:56:27 <GreaseMonkey> hard disk's a little slow this morning
19:57:21 <AnMaster> mhm
19:57:39 <AnMaster> GreaseMonkey, why are you still using a legacy rotational storage device?
19:57:55 <AnMaster> also instead of flying cars: WHERE ARE MY DATA CUBES?
19:58:15 <GreaseMonkey> AnMaster: so how much are, um, 160DMGB+250DMGB SSDs?
19:58:26 <GreaseMonkey> DMGB = drivemakers' gigabytes = 1000^3B
19:58:27 <AnMaster> wth is DMGB
19:58:56 <AnMaster> GreaseMonkey, thanks, will use that against ehird (he uses GB/GiB)
19:58:56 <GreaseMonkey> i completely refuse to refer to them as just GB
19:59:12 <GreaseMonkey> GB was always 1024^3
19:59:27 <GreaseMonkey> those who insist we use GiB for that should learn to stop being so cheap
20:00:12 <GreaseMonkey> some country should bring in a legislation which enforces that a xB is 1024^y bytes
20:00:28 <AnMaster> GreaseMonkey, G = 1000 in the SI system though
20:00:31 <GreaseMonkey> because 1000^x is just false advertising
20:00:37 <AnMaster> or are you using kig for 1000 grams?
20:00:42 -!- augur has quit (Ping timeout: 252 seconds).
20:00:51 <AnMaster> GreaseMonkey, ^
20:00:58 <GreaseMonkey> well, note that there's kg, and KB
20:01:01 <AnMaster> also they should just use cyliders
20:01:04 <AnMaster> cylinders*
20:01:26 <GreaseMonkey> it was always 1024 for a KB, anyway
20:01:41 <cpressey> kig = 1024 grams?
20:01:52 <GreaseMonkey> thing is, with metric, you're working with real numbers
20:01:54 -!- augur has joined.
20:02:03 <GreaseMonkey> and a human is supposed to measure it
20:02:04 <scarf> wow: http://labs.ripe.net/content/pollution-18
20:02:08 <GreaseMonkey> (usually)
20:02:17 <GreaseMonkey> but with a computer, you're working with integers
20:02:17 <scarf> it's a graph of what happened when 1.1.1.0/24 was put live on the internet
20:02:23 <pikhq> GreaseMonkey: The SI prefixes are useless for computers, yes.
20:02:25 <AnMaster> <cpressey> kig = 1024 grams? <-- ah hm
20:02:29 <scarf> basically, the people who owned it got DOSed
20:02:37 <pikhq> This is why we shouldn't use SI prefixes.
20:02:40 <AnMaster> make it, dmkg for "drive makers kilogram"
20:02:45 <pikhq> Mmm, gibibytes.
20:02:51 <AnMaster> to keep with GreaseMonkey's logic
20:02:59 <scarf> should be Kig
20:03:07 <scarf> capital K on KiB
20:03:09 <GreaseMonkey> thing is, it was always 1024 bytes to a KB
20:03:19 <AnMaster> err
20:03:22 <GreaseMonkey> it was also always 1000 grams to a kg
20:03:39 <AnMaster> isn't 1.0.0.0/8 private?
20:03:41 <AnMaster> oh wait
20:03:43 <AnMaster> that is 10
20:03:45 <pikhq> AnMaster: No.
20:04:00 <pikhq> 1.0.0.0/8 was recently assigned.
20:04:08 <pikhq> However, it's been used as a private space.
20:04:25 <pikhq> (it was reserved by the IANA since '81)
20:04:28 <scarf> there's a huge amount of unofficial use of 1.0.0.0
20:04:35 <GreaseMonkey> at first i thought you said they allocated 10.0.0.0/8
20:04:50 <scarf> the authorities are holding on to 1.1.1.0/24 on the basis that it would be crazy to allocate it to anyone
20:05:08 <GreaseMonkey> good point
20:05:23 <scarf> and it looks like RIPE decided to put it online for a bit to see what happened
20:05:29 <scarf> (the result was chaos, as you can imagine)
20:06:20 <scarf> <RIPE> We can certainly conclude from this that specific blocks in 1/8 such as 1.1.1.0/24 and 1.2.3.0/24 are extremely polluted. Unless the traffic sent towards those blocks is significantly reduced they might be unusable in a production environment.
20:07:58 <GreaseMonkey> it's not so bad considering that they actually tested it before attempting to release it
20:08:54 <scarf> it's brilliant that they even thought of testing
20:09:29 <GreaseMonkey> that might be why AW's server was so damn slow
20:09:51 <AnMaster> scarf, I couldn't imagine before reading it that it was chaos
20:10:11 <AnMaster> mainly because I couldn't imagine people would be abusing reserved prefixes
20:10:20 <scarf> AnMaster: Wikipedia has edits from 1.2.3.4, you can check that yourself
20:10:29 <scarf> that's what cottoned me on to thinking that there might be a problem
20:10:35 <scarf> (it also has edits from 127.0.0.1, which is amazing)
20:10:38 <AnMaster> scarf, what does that signify btw
20:10:44 <AnMaster> the 1.2.3.4 thing I mean
20:10:54 <scarf> it means that someone, probably a dev, put a fake IP in for the edit
20:10:59 <scarf> although I'm not sure whyt
20:11:00 <scarf> *why
20:11:03 <AnMaster> hm
20:11:12 <AnMaster> scarf, as for those 127.0.0.1, what sort of edits
20:11:23 <AnMaster> I guess someone used an ssh tunnel...
20:11:23 <scarf> you can check for yourself, you know
20:11:29 <scarf> http://en.wikipedia.org/wiki/Special:Search?go=Go&search=Special:Contributions/127.0.0.1
20:11:33 <AnMaster> scarf, I don't know how to search on the ip
20:11:34 <AnMaster> ah
20:11:47 <scarf> looks like mostly interwiki links to the Russian Wikipedia
20:11:55 <AnMaster> yep
20:12:02 <AnMaster> so a bot that wasn't properly logged in?
20:12:31 <scarf> http://en.wikipedia.org/wiki/User_talk:127.0.0.1 is not that bad either
20:12:41 <scarf> "Note: In the event of vandalism from this address, half-baked abuse reports may be sent to Wikimedia's own network administrator for further investigation that will never be looked into."
20:12:57 <GreaseMonkey> hehe
20:13:14 <AnMaster> "Caution should never be used when blocking this IP or reverting its contributions without checking - if a block is unnecessary, administrators should consider using the banhammer without adding any templates as the block reason. Of course, most administrators are just testing the blocking and unblocking procedure."
20:13:18 <AnMaster> err
20:13:24 <AnMaster> is that backwards?
20:13:24 <GreaseMonkey> ARGH BALLS i can't get on wikipedia >_> <_<
20:13:40 <scarf> AnMaster: yes, it's backwards
20:18:00 <AnMaster> scarf, wth at http://en.wikipedia.org/wiki/User_talk:1.2.3.4
20:18:05 -!- |MigoMipo| has changed nick to MigoMipo.
20:18:46 <scarf> no idea there
20:19:20 <scarf> might be worth mentioning that it's changed hands since
20:19:43 <AnMaster> also why allocate 1/8
20:19:50 <AnMaster> just switch to ipv6 already dammit
20:20:01 <AnMaster> scarf, what changed hands?
20:20:26 <scarf> AnMaster: 1/8
20:21:21 <AnMaster> hm
20:30:21 <oerjan> um that's backwards? it seemed sort of logical to me to treat an address heavy-handed if it doesn't really have the right to exist as an editor...
20:37:15 <cpressey> AnMaster: I'm tempted to say that IPV6 is the "jumped the shark" of network addressing, but I won't :) It's not /quite/ the same effect...
20:37:45 <AnMaster> cpressey, it *is* about our only hope atm
20:37:58 <AnMaster> cpressey, any alternatives would take years to develop
20:38:00 <cpressey> But it's similar, I guess: "I'm only going to start doing it if all my friends are doing it, and my friends feel the same way"
20:38:09 <AnMaster> and we don't have that amount of time any more
20:38:16 <AnMaster> before ipv4 runs out completely
20:38:33 <AnMaster> cpressey, also I use a sixxs tunnel
20:40:34 <AnMaster> cpressey, my isp has an ipv6 block allocated, they don't use it though
20:40:39 <AnMaster> also I doubt my router could handle it
20:40:46 <AnMaster> could telnet to it and check it's menus
20:40:55 <AnMaster> (it's a speedtouch adsl modem/router)
20:41:48 <cpressey> I remember trying NetBSD once, and it insisting on using IPV6. That was not fun.
20:42:14 <pikhq> Comcast is beginning the switch to IPv6 *now*.
20:42:28 <cpressey> How very NetBSDish... "No! We will do it the RIGHT way, no matter what the cost!"
20:42:40 <GreaseMonkey> ...what the crap
20:42:49 <SimonRC> cpressey: heh
20:42:52 <GreaseMonkey> netbsd should still be using IPv2
20:43:01 <GreaseMonkey> THAT seems more netbsdish to me
20:43:08 <cpressey> GreaseMonkey: Heh
20:43:39 <GreaseMonkey> my netbsd experience is that it takes up sod-all RAM
20:44:15 <pikhq> GreaseMonkey: Garbage collection and manual memory management are highly overrated.
20:44:29 <pikhq> Just statically create a heap of a couple gigabytes and hope it works.
20:44:30 <pikhq> :P
20:45:06 <oerjan> just use your nanobots to create a few cubic meters of computronium
20:46:38 <scarf> pikhq: or the anagolf technique of just using whatever memory addresses happened to be in the variables at the time, and repeatedly running the program until they turn out to be writable
20:48:32 <AnMaster> cpressey, my isp hasn't started going ipv6 yet
20:48:33 <AnMaster> also
20:48:37 <AnMaster> things will be bad
20:48:45 <AnMaster> because almost no servers will have ipv6
20:48:52 <AnMaster> so you will have to go through 6-to-4
20:48:54 <AnMaster> and that
20:48:59 <AnMaster> will result in problems
20:49:06 <AnMaster> like connection limit counts for irc
20:49:07 <AnMaster> and what not
20:49:24 <AnMaster> also wikipedia banning
20:49:36 <scarf> I'd expect wikipedia to go to ipv6 some time
20:49:41 <AnMaster> well yes
20:49:48 <AnMaster> scarf, still, the issues are huge
20:50:00 <scarf> going home now, anyway
20:50:01 <scarf> bye everyone
20:50:04 <AnMaster> scarf, imagine rate limiting on lots of places
20:50:05 <AnMaster> scarf, cya
20:50:09 -!- scarf has quit (Remote host closed the connection).
20:50:09 <GreaseMonkey> cya scarf
20:56:44 <Ilari> Quite polluted /24 when announcing it causes tens of megabits of traffic (on block that should be quiet)... :-/
21:05:21 -!- augur has quit (Ping timeout: 259 seconds).
21:07:14 -!- augur has joined.
21:13:58 -!- MigoMipo has quit (Quit: co'o rodo).
21:14:07 -!- impomatic has left (?).
21:21:13 <Ilari> Often resource exhaustion problems don't feel that pressing until you really hit them, and when you hit them, you hit them _hard_.
21:30:44 -!- kar8nga has joined.
21:37:29 -!- adam_d has joined.
21:43:45 <pikhq> Well, I've got CPS with getcontext et al...
21:43:51 <pikhq> And it behaves... *Weirdly*.
21:44:38 <pikhq> It goes through 3 continuations just fine, before going up to a nulled out stack frame.
21:45:33 <pikhq> Yes, a nulled stack frame.
21:50:06 <cpressey> I'm sorry I ate your stack frame.
21:50:09 <cpressey> I was hungry.
22:01:29 <pikhq> And it is in fact a stack fram that was valid.
22:01:36 <pikhq> It just somehow gets nulled out.
22:01:46 <pikhq> I don't know how.
22:06:50 <pikhq> It would appear to be a result of a too-large new stack...
22:06:53 <pikhq> Somehow.
22:11:44 <pikhq> ... The program changes behavior with normal malloc.
22:12:10 <pikhq> With GC_MALLOC, it computes factorials validly for n in [0..11].
22:12:27 <pikhq> With normal malloc, it loops infinitely.
22:16:07 <pikhq> http://sprunge.us/RNie
22:18:06 <AnMaster> niht
22:18:08 <AnMaster> night*
22:20:59 <oerjan> ni!
22:22:40 -!- MizardX has joined.
22:23:28 <pikhq> I'm never using getcontext ever again.
22:23:50 -!- kar8nga has quit (Read error: Connection reset by peer).
22:27:02 -!- Pthing has joined.
22:28:46 -!- tombom_ has quit (Quit: Leaving).
23:06:50 -!- MissPiggy has quit (Ping timeout: 245 seconds).
23:07:24 -!- coppro has joined.
23:19:46 <cpressey> So I'm thinking an imperative function-level language would revolve around three concepts: storage locations (replace atoms), procedures (replace functions), and procedurals (replace functionals). Parameters would always be passed "by reference" -- so procedurals make new procedures by modifying (clobbering) existing procedures.
23:26:33 -!- FireFly has quit (Quit: Leaving).
23:34:29 -!- BeholdMyGlory has quit (Remote host closed the connection).
23:37:29 -!- adam_d has quit (Quit: Leaving).
23:45:21 -!- mquin has quit (Remote host closed the connection).
←2010-02-02 2010-02-03 2010-02-04→ ↑2010 ↑all