00:00:16 you need: memcpy, strlen, addition 00:00:19 shaddap :P 00:00:26 I decided that 00:00:29 char start[PATHLEN + LEN("run")] = SV "/"; 00:00:29 strcpy(start, s); 00:00:29 strcpy(start, "/run"); 00:00:34 is probably not going to be a performance problem. 00:00:53 olsner: WIMP = Weakly Interacting Massive Particle. The naming coincidence cannot be coincidental, as it is all too apropriate... 00:00:55 *appropriate 00:01:28 if appropriate means nonsensical, then yes 00:01:45 olsner: WIMP applications (i.e. "GUI applications") don't interact well with each other -- there's no pipes for GUIs, no easy moving of objects around -- and they tend to have many more interface aspects than they need or should have. 00:02:02 massive and weakly interacting 00:02:03 olsner: So, yes, WIMP programs are weakly-interacting massive particles. 00:02:18 massive and weakly interacting <-- ? 00:02:26 elliott is designing a computer that uses the weak nuclear force as a substitute for electricity (yes, I know that somehow in some way I don't understand they're the same thing) 00:02:59 Or um.. connected..ish? 00:03:33 olsner: still think it's nonsensical? :p 00:04:49 I'm actually not thinking at all at the moment, just writing code 00:07:29 How does one "shield" against the weak interaction? 00:08:09 Or maybe I'm reading too much into this 00:10:16 Sgeo: I have no idea what you're talking about. 00:10:34 " Gravitation:" 00:10:40 "cannot be absorbed, transformed, or shielded against" 00:10:55 I sort of read that as implying that the other interactions can 00:10:59 :/ 00:11:06 http://en.wikipedia.org/wiki/Fundamental_interaction 00:13:46 $ ./sv 00:13:46 usage: ./sv {u|d} service 00:13:48 now to make it do useful things 00:14:54 Why does the unlimited range of the graviton imply masslessness? 00:17:13 "For example, a detector with the mass of Jupiter and 100% efficiency, placed in close orbit around a neutron star, would only be expected to observe one graviton every 10 years, even under the most favorable conditions." 00:19:15 "A cheat: To prevent people from criticizing your magic numbers, make them deep magic numbers instead. Using the nearest power of 2 (1,2,4,8,16,32,64,128...) as your magical number makes people think there is some deeper magic going on, and they'll be less likely to criticize. *innocent look* 00:19:15 As an added bonus, they'll probably spend hours trying to figure out what that deeper magic is. :-) 00:19:15 That is the opposite of a trick that works. The most suspicious magic numbers of all are the round ones, in any base, but especially base 2 and base 10. PrimeNumbers, on the other hand, are essential for many algorithms to work correctly (e.g. many forms of hashing, with a prime table size). So use numbers like 131 and 6983 :-)" 00:20:37 Vorpal: Yellow: "Polynomial!" 00:20:44 base 10 is much more suspicious than base 2 00:21:32 Active Worlds has a ... really weird magic number 00:22:09 32764 is the highest coordinate 00:22:35 Which, at first glance, makes sense, because it's 3 coords away from 32767, which looks like it should make sense 00:22:45 But the units are dekameters, and AW does things in centimeters... 00:24:18 Hmm. Maybe internally, some int that has units of dekameters IS being stored in the client 00:24:20 As a signed int 00:24:36 Weird, but whatever 00:29:42 it could be using some bits for tags (e.g. unit = centimeters or dekameters) 00:30:09 hmm, that doesn't necessarily make sense 00:40:57 olsner: can i pay you to write a compiler to machine code in x86-64 assembly 00:41:24 I don't know how your finances are, so it's really impossible for me to tell 00:41:45 olsner: ok, what if i didn't pay you? 00:41:50 * elliott LOGIC KING 00:42:02 olsner: also the code is freestanding, so no libc 00:42:02 HAVE FUN 00:42:13 btw, my console driver can output single characters and newlines now 00:42:46 as in clear the current line and move to the next line *or the upper-left corner of the screen* 00:44:04 and I have programmed the APIC to give timer interrupts 00:47:46 olsner: scrolling is easy, even i did scrolling :P 00:47:49 admittedly in C, but c'mon! 00:48:06 I didn't bother implementing scrolling :) 00:48:08 just wrapping! 00:50:28 although haskell is my language of choice for writing compilers, it does seem interesting to write one in assembly 00:51:35 * olsner looks at topic: GREGOR HAS NOT WASHED HIS BEARD!? FILTHY GREGOR! 00:51:43 and good night :D 00:51:48 * Gregor strokes his bald chin. 00:53:26 olsner: well in this case 00:53:35 olsner: it'd be high-level-language to x86-64 machine code, written in x86-64 assembly 00:53:40 olsner: so lawl 00:53:48 olsner: (in OS-level code) 01:03:41 http://i.imgur.com/vIIc1.png Google: "Fuck you Facebook" 01:05:49 Someone's going to think that that means the contacts would be deleted from Gmail 01:08:49 No. 01:08:52 No they aren't. 01:09:21 elliott, you have an amazing amount of faith in computer users 01:09:37 Sgeo: No, I don't; I have a decent amount of faith in Google's writing team. 01:10:17 s/decent/amazing/ 01:10:18 I think the intersection of "people who are savvy enough to (1) use gmail and (2) put it into facebook to import contacts" and "people who manage to read that as 'we delete u conversations'" is very small. 01:10:30 Gmail isn't exactly ubiquitous. 01:10:57 Hmm, good point, I guess 01:11:02 Gregor: [[Q: GPL sucks! Now I can't compile my BSD programs with the diet libc! 01:11:02 A: Wrong. You can compile them, and you can use them. You just can't 01:11:02 redistribute the binaries. That said: I will not be sueing anybody 01:11:02 for distributing binaries of BSD programs linked against dietlibc, as 01:11:02 long as the source code is available somewhere publicly.]] 01:11:06 Gregor: That's new, I think X-P 01:11:19 Gregor: (Discuss now whether "I will not be sueing anybody [...]" counts as a license to do that.) 01:11:25 Argh. 01:11:32 It's irrelevant whether it is or not, he's still wrong. 01:12:49 lol dietlibc 01:13:43 coppro: why lol 01:13:53 The GPL is really really stupid for libraries. 01:13:59 *Especially* statically linked ones. 01:14:21 elliott: lolgpllol 01:14:23 Gregor: Yeah, he is. I'm torn between "hell, developers can go fuck themselves anyway" and "WOULDN'T IT BE SAFER TO SUFFER THROUGH THE QUAGMIRE OF CONFIGURATION AND COMPILATION THAT IS UCLIBC ON A REGULAR BASIS" 01:14:50 elliott: I should just write you a libv 01:14:52 *libc 01:14:57 coppro: yeah, yeah. at least he's wrong about that and it is legal to redistribute dietlibc-linked binaries, as long as you also offer the source to the program "under the GPL" (i.e. you can offer it under the BSD too, since that can be used as GPL) 01:15:01 coppro: sure, that'd be great 01:15:05 coppro: when will it be done 01:15:34 elliott: when you convince my university to let me take that instead of retarded 2nd-year courses 01:15:44 coppro: i'll pay you 3p a week 01:15:45 brb 01:16:06 elliott: not likely 01:20:21 coppro: That's a few thousand USD, though. 01:20:47 :P 01:21:02 pikhq: yeah, but that money is worthless here 01:22:49 coppro: Sadly. 01:31:51 pikhq: I'll pay you $7,000/week to do it. 01:31:53 (That's 6p!) 01:34:09 -!- augur has quit (Remote host closed the connection). 01:36:00 coppro: To be honest, coreutils would be more useful than a libc. 01:37:54 elliott: noted! 01:38:37 coppro: I'm gonna go crazy here. $8,000/week. That's 6.86p! 01:38:55 coppro: And I'll pay by the SECOND. 01:42:49 -!- jkenj999 has quit (Remote host closed the connection). 01:48:58 elliott: I told you 01:49:02 that money is worthless here 01:49:12 coppro: the CAD is next 01:49:20 coppro: besides, it has our Queen on it 01:49:25 elliott: 6.86p is only worth like $10 01:49:33 coppro: no, $4,000CAD 01:49:42 coppro: it's law; any piece of metal with the Queen's face on it is legal tender throughout the commonweahlth 01:49:58 coppro: heh i read that as $1 01:50:00 sneaky sneaky 01:53:06 elliott: I highly doubt that. 01:53:24 pikhq: Treasonerist. 01:53:32 elliott: rebel 01:53:39 coppro: Treasonerist. 01:56:01 comex: congratulations, Hillary Rodham Clinton! 01:56:24 elliott: Well, I am in a country formed by treason against the Crown. 01:57:07 pikhq: And look where that's got you! 01:57:34 It got us free of ridiculous libel laws? 01:57:50 elliott: Idea 01:58:00 I will make coreutils under the Forwardspace banner 01:58:06 (Just learned about the SPEECH Act of 2010) 01:58:09 Forwardspace is like the opposite of Backspace 01:58:13 coppro: Do I want to know what that is? 01:58:20 so you can also omit the name and also the character afterwards 01:58:30 elliott: BTW, your currency is freaking mad. 01:58:44 that way, Linux distributions that use Forwardspace coreutils should be referred to as "Forwardspace/Linux" or just "Linux" 01:58:47 There are 14 completely different entities that make it. 01:58:53 pikhq: "1 UK£ = 1.6121 U.S. dollars" --Google 01:59:05 coppro: Or how about just calling it "minimalist core utilities for Linux" :-P 01:59:09 Minutils! 01:59:12 pikhq: Decentralisation, bitch. 01:59:47 elliott, when did Hillary Rodham Clinton _call_ a CFJ? 01:59:54 elliott: Minutils/Linux sounds awful 01:59:56 She's been the SUBJECT of one, yes 02:00:11 Sgeo: In my imaginary version of the game, Hillary Rodham Clinton punched you to death. 02:00:15 elliott: With different currency designs from each entity. 02:00:17 Sgeo: Q.E.D. 02:00:24 pikhq: I think they're basically unified... 02:00:28 coppro: So don't call it that :P 02:00:36 coppro: The coreutils used is no more important than the libc, the httpd, the package manager. 02:00:46 coppro: Hell, the kernel -- it's pointless to call anything an "X distribution" really. 02:01:03 All "Linux distribution" means nowadays is "GNU, oh and some of that Torvalds stuff too". 02:01:23 elliott: No, they're radically different designs for the banknotes. 02:01:31 pikhq: oh the notes. kay 02:01:37 pikhq: at least our currency is strong :P 02:01:45 elliott: We'll fix that yet. 02:01:46 http://js.codu.org/wiki/ 02:01:51 * pikhq exports Bush 02:02:03 pikhq: Dude, we wouldn't let him in. 02:02:10 Gregor: Oh how Web 2.0. 02:02:24 Dude this shit's so Web 2.0 it's Web 2.2 02:02:40 * Sgeo is waiting for Web Gingerbread to come out 02:02:46 Gregor: When I look at that logo, I see someone trying to sell me something. 02:02:52 Gregor: Make it Comic Sans or something so I know you don't give a shit. 02:02:52 elliott: Oh, we'll export him alright. 02:03:02 elliott: And make him all the Lords. 02:03:11 elliott: I don't do aesthetics, I get other people to :P 02:03:44 * Sgeo WTFs at the OpenID he just put in 02:03:53 Gregor: hurf durf codu js 02:03:59 * Sgeo WTFs even more at it working 02:04:07 o.O' 02:04:17 Sgeo: ? 02:04:21 Scotland and Ireland have no legal tender. 02:04:26 Erm, Northern Ireland. 02:04:30 I tiredly put in sgeoster@myopenid.com 02:04:32 Ireland, of course, uses the Euro. 02:04:38 I got directed to myopenid.com 02:05:01 pikhq: "Repeal the 26th Amendment!" --Ann Coulter 02:05:07 Sgeo: foo@yahoo.com is a valid OpenID too. 02:05:11 pikhq: I DO BELIEVE that they use GBP in Scotland and Northern Ireland, they're not on the barter system :P 02:05:14 It's a Feature, it appears. 02:05:22 Gregor: Yes they are. 02:05:23 Gregor: It's not legal tender. 02:05:25 Gregor: They're also unwashed. 02:05:32 Gregor: It's the *de facto* currency. 02:05:34 pikhq: THE MARKET IS THE BLACK MARKET 02:05:43 elliott, that just translates to http://foo.yahoo.com ? That's what happened here. Or is that a MyOpenID thing? 02:05:51 pikhq: THE SCOTTISH BANK PRINTS SCOTTISH VERSIONS OF GBP 02:06:02 Sgeo: Both do that, clearly. 02:06:06 Gregor: BANKS 02:06:08 Gregor: I'll scot your version. 02:06:10 Gregor: NOT BANK, BANKS. 02:06:16 Gregor: ALSO, THOSE ARE PROMISSORY NOTES. 02:06:25 Banksy 02:06:45 pikhq: Last I checked, your mom is a promissory note. 02:08:01 I LOVE HOW SLOWLY UBUNTU 7.10 INSTALLS UNDER SOFTWARE EMULATION OF X86-64 02:10:51 "fgetty is actually a mingetty stripped of the printfs. Why would anyone do that? Because it can be linked against dietlibc then, yielding a 7k binary with a much smaller memory footprint." 02:10:53 I love crazy people. 02:11:22 Bah! It prints out /etc/issue. 02:11:24 This is unacceptable. 02:12:12 Gaaaaaaaawd this dietlibc guy's license incompetence irks me. 02:12:26 Gregor: It's one single mistake :P 02:12:35 UNACCEPTABLE. 02:12:44 * Sgeo angers at a rumor he was sure wasn't true not coming true 02:12:51 Gregor: I'm not entirely sure that FAQ section *is* new. Besides, he hasn't replied to the second email I sent him -- probably sick of my whining. 02:12:56 Although my reason for doubting the rumor turned out to be a bad reason 02:13:01 * Sgeo whargarbles 02:14:24 Gregor: It does irk me too a bit though but only because I'm planning to actually put it into practice. 02:14:50 Gregor: On one hand, of course in court both the truth and his statement that he won't sue would end it immediately. On the other, I don't want to be on the bad side of developers... 02:16:40 * Sgeo wonders how much memory Jolicloud requires 02:20:27 I'm going to try and redo my pcc toolchain with dietlibc. 02:24:16 Gregor: How come CVS isn't dead? 02:24:23 Seriously, I've used CVS SO MUCH in the past days. 02:24:28 Same 'ere, actually :P 02:24:29 *past few days. 02:24:43 Gregor: I thought svn had taken over for all the clods who won't use anything distributed. 02:25:05 endif 02:25:05 endif 02:25:05 endif 02:25:05 endif 02:25:05 endif 02:25:06 endif 02:25:08 ... 02:25:09 -- dietlibc/Makefile 02:25:17 (17 lines) 02:25:23 Apparently gmake can't do elseif :P 02:27:17 Tastes like lisp :P 02:27:33 Gregor: Good luck finding a Lisper who does 02:27:33 ) 02:27:33 ) 02:27:34 ) 02:27:36 Gregor: Rather than ))) :P 02:27:52 Trufe. 02:27:59 Gregor: Truse? 02:28:10 Haa, dietlibc almost compiles with pcc. 02:28:35 $ make CC="$K/stage1/bin/pcc -D__restrict__= -D__regparm__(x)=" 02:28:45 Oops, needs more escaping, of course. 02:29:02 Hmm, that doesn't apply inside __attribute__. 02:29:04 Oh well, /me patches 02:29:33 Gregor: Using a BSD compiler to compile something with GPL-related semi-controversy; do I get a medal? 02:29:35 Or a slap? 02:29:38 /home/elliott/kitten/stage1/bin/pcc -D__restrict__= -I. -isystem include -Os -fstrict-aliasing -momit-leaf-frame-pointer -mfancy-math-387 -W -Wall -Wextra -Wchar-subscripts -Wmissing-prototypes -Wmissing-declarations -Wno-switch -Wno-unused -Wredundant-decls -c lib/qsort.c -o bin-x86_64/qsort.o -D__dietlibc__ 02:29:38 major internal compiler error: lib/qsort.c, line 34 02:29:49 I think a medal. 02:30:01 The evil, evil line 34: 02:30:02 quicksort(base,size,l,j,compar); 02:31:32 I have no idea what it thinks is wrong with quicksort :P 02:31:43 elliott: stop being such a ricer 02:31:57 coppro: Uhh, because of the flags? Not my flags kthx. 02:32:03 coppro: Also, uhh, -Wanything != ricer 02:32:07 The only optimisation flags there are 02:32:14 -Os -fstrict-aliasing -momit-leaf-frame-pointer -mfancy-math-387 02:32:17 Which are hardly "ricer". 02:32:27 elliott: your pedantic obssession with tiny libraries and compilers is 02:32:34 In fact, is -fstrict-aliasing strictly an optimisation? (Pun, etc.) 02:32:51 but they're certainly nicer 02:32:56 coppro: I could respond to that seriously, or I could just roll my eyes and proclaim that, yes, you have successfully deduced my motive from IRC compile logs. 02:33:26 coppro: If you pointed me to the BEST LIBC EVER and it just happened to be gigantic, fine. I'd consider it. 02:33:50 But as smallness *is* a virtue -- a strong virtue, even -- of software, and those who excel in one software virtue tend to excel in the others (and contrariwise)... 02:33:51 -fstrict-aliasing is not directly an optimization 02:34:12 Besides, smaller code is prettier. And to hell with any system I can't look at and say is pretty in some way, even if it is Unix. 02:34:26 -fstrict-aliasing won't do squat at -O0 02:34:31 It's not like I'm wasting time doing this that I could be spending doing valuable things, this whole thing is an exercise in wheel-polishing^Wreinventing. 02:34:33 coppro: -Os != -O0 02:34:38 elliott: I know 02:34:43 but you asked if it was an optimization 02:34:45 and I answered 02:34:52 (also, I /do/ agree with -Os) 02:34:53 coppro: oh, okay 02:35:05 -Os is usually faster than -O3 these days, I'd wager. 02:35:13 (Also, y'know, small.) 02:35:19 yeah 02:35:24 also it generates correct code, etc. 02:35:44 (-ftree-vectorize produces some of the dumbest things ever) 02:35:57 a friend works a lot on ffmpeg 02:36:06 and they get to find all of GCC's wonderful compiler bugs 02:36:10 favorite: 02:36:12 coppro: Can you use your amazing powers of deduction to figure out what "major internal compiler error: lib/qsort.c, line 34" is telling me, other than "don't use pcc to compile dietlibc"? :P 02:36:15 int c = i; 02:36:15 Yeah, gcc is buggy software. 02:36:16 c--; 02:36:17 Possibly the buggiest. 02:36:18 int d = i; 02:36:19 wow 02:36:19 Ever. 02:36:22 assert(c != d); 02:36:24 ASSERTION FAILED 02:36:35 coppro: ...you're kidding me 02:36:38 BEST THING EVER 02:36:40 elliott: no 02:36:43 I wish I was 02:36:45 coppro: just with heavy optimisation, presumably? 02:36:49 yeah 02:37:04 coppro: i can *sort* of forgive that because heavy optimisation + assert = why, but on the other hand... gcc fails at basic comparisons WHOO 02:37:14 B's ruleset is surreal 02:37:16 at least when pcc fails, it just says "major internal compiler error" and gives up 02:37:20 * coppro <3 bugpoint 02:37:25 comex: linky 02:37:26 comex: B stopped being boring as soon as Era 5 ended! 02:37:27 also wrong channel 02:37:28 erm 02:37:31 comex: B stopped being interesting as soon as Era 5 ended! 02:37:41 That was the most fun bit of chaos EVER and I'm totally not biased. 02:37:56 it's like Japanese, where Chinese is the Agoran ruleset. 02:37:58 :< 02:38:52 comex: ALLOW ME LINK YOU TO THE SUPERIOR, ERA 5 RULESET (AS SOON AS THE WIKI LOADS) 02:39:08 elliott: which was Era 5? 02:39:14 also WHERE IS MY LINK 02:39:18 * comex searches for DimShips and is disappointed 02:39:21 coppro: The one me and teucer wrote that CHANGED EVERYTHING. 02:39:25 Era 5 came and went without my noticing? 02:39:34 Sgeo: ...era 5, the Agoran Era, and countless crises. 02:39:38 Era 5 was, like, 2008. 02:39:53 I think I gave up on B around the Agoran Era 02:40:00 Sgeo: Era 5 was BEFORE that >_ *>_< 02:40:04 Rule 50 would be fun to scam 02:40:13 Which was the Era with the Ring of 4E13? 02:40:15 coppro: Everyone then spent the whole time whining at me for making it SO BROKEN on a few weeks' notice (it was in an Emergency), spent the rest of the era fixing it up and generally having much fun in chaos without realising it, and then wiping it saying it was AWFUL. 02:40:17 I was around then 02:40:20 coppro: The game then promptly died. 02:40:22 ...Era 5 is the obvious answer 02:40:23 Sgeo: "4E13" 02:40:27 Huh. 02:40:34 elliott, _RING OF_ 02:40:35 Sgeo: 4E = Fourth Era, but yes, I think that was 5. 02:40:36 Sgeo: Right. 02:40:40 elliott: ah, right 02:40:43 In order to be an EU member state, the state must be European. 02:40:44 Bgoran era, you mean 02:40:50 5. MAY, PERMITTED: Performing the described action does not violate the rule in question. Where there is no more specific indication to the contrary, this also implies that the action is POSSIBLE. 02:40:53 *Bn 02:40:56 coppro: Err, no. 02:40:59 Where European means "European, as judged by the European Council". 02:41:00 elliott: LIES 02:41:04 coppro: Bgoran era was after that. 02:41:04 comex: WHERE IS MY LINK 02:41:17 I wanted to take people with me on a journey to retrieve the ring 02:41:21 Which allows for EU members to not exist geographically in Europe. 02:41:25 I had it all planned out 02:41:29 It would take real-world months 02:41:32 iirc 02:41:43 coppro: comex: Found it, second while Firefox unfreezes. 02:41:51 (and there is precedent on this: Cyprus, which joined the EU in 2008, is geographically in Asia) 02:41:52 coppro: comex: http://b.nomic.net/index.php?title=Rules&oldid=9553 02:41:56 coppro: comex: Behold the AWESOME. 02:42:13 I was around around then 02:42:14 elliott, you don't remember my plans? 02:42:15 vaguely 02:42:44 pikhq: that is far less dumb than it sounds, in my opinion 02:42:45 I think, sadly, Marr965 was involved 02:43:04 I wanted to negotiate with him for.. transportation to our departure point, I think 02:43:41 what is "Antient"? 02:43:57 coppro: Unbelievably, that Game Objects rule was MORE legalistic and programmerish than before E5. 02:44:00 why is 4E considered imaginary? 02:44:03 coppro: I actually TONED DOWN the crazy legalisticness. 02:44:08 comex: canon crisis 02:44:15 wasn't there some crisis about comments 02:44:19 comex: turns out $huge_swathes_of_history never happened for $trivial_ancient_reason 02:44:20 that applied way back to the original ruleset? 02:44:29 comex: yes, but i think it was after that 02:44:36 the comment crisis was what brought era 5 down I think 02:44:47 If you ask me, the Grid game was AWESOME. 02:45:13 * Sgeo decides to check his archives 02:45:16 Oh, I also refactored the proposal passing logic... BUT WHAT DO I GET FOR IT? 02:45:17 NOTHING! 02:45:32 # If it is Won, it Passes. 02:45:32 # Otherwise, it Fails. 02:46:09 [[Prize: A T-Shirt reading "I won B Nomic via at least three different Victory Conditions during an nweek, and had not won by the Win By Having Won condition during that nweek, and all I got was this stupid T-Shirt".]] 02:46:43 "Marr965, would you like to help Teucer, Warrigal, and myself shave 17 02:46:43 ndays off of what would be a 67 nday journey?" 02:46:59 Did the era even last 67, or even 50, ndays? 02:47:29 It lasted like 7 at the most :P 02:48:04 I would love to have a month-or-more-long era that lasts an nday or less. 02:48:19 "(Discussion 02:48:19 including whether a 50 nday journey is even worth it.)" 02:49:51 coppro: how to get past the major compiler error: compile just that file with gcc 02:50:50 http://freefall.purrsia.com/ 02:50:53 I AM OFFENDED 02:51:03 (Note: not really) 02:53:32 libpthread/pthread_internal.c:397: error: If-less else 02:53:33 SO NOT REASSURING 02:53:53 Notably, there's a mess of cpp here, so I don't know WHERE line 397 is :P 02:54:02 * comex registers 02:54:04 :q 02:54:06 whoops 02:54:12 comex: for C Nomic? 02:54:26 if (__likely(__modern_linux==1)) 02:54:31 * comex hates how :wq writes and quits, but :w!q writes to the file 'q' 02:54:44 comex: :wq! :p 02:54:56 elliott: you should help us destroy Agora 02:55:00 why does that code include a runtime optimization for something that should fold to a constant 02:55:06 coppro: that ... why 02:55:08 comex: i don't know 02:55:18 comex: it's pthreads, all bets are off 02:55:53 oh well, time-honoured trick; s/pcc/gcc/, carry on 02:56:05 :D that fails 02:56:57 * Sgeo will not stop attempts to destroy Agora unless they require the cooperation of more than 20 people 02:57:06 02:58:28 it succeeded 02:58:35 but it was just a sad truth 02:58:39 not actually funny 02:59:43 Maybe I could work on attempting to scam-proof the ruleset 03:00:05 Not as in, "impossible" 03:00:20 it's pretty scam-proof as is 03:00:27 you can escalate from 1->3 easily 03:00:30 but getting 1 is hard 03:00:49 Sgeo: have you considered doing /other/ things 03:00:55 yeah, except I think it's now scammable 03:01:01 maybe you could make activeworlds a partnership, and we could fine it a billion fucking dollars. 03:01:11 I thought partnerships were dead 03:01:31 * Sgeo still wants to revive Framework Nomic 03:01:33 Hmm 03:01:38 Is that like reviving a fetus? 03:03:35 * Sgeo has been fantasizing about MMONomic 03:04:14 Sgeo: You are banned from saying these words, indefinitely: 03:04:16 Sgeo: - nostalgia 03:04:18 Sgeo: - revive 03:04:25 Sgeo: - fantasise (and all derivative words). 03:04:29 Sgeo: Thank you and have a nice day. 03:04:47 - fantasy and derivatives too, to stop you loopholing that. 03:05:56 So, I'm not allowed to express my nostalgia for the days that I was more activie reviving a dead game that I've fantasized about playing again for years? 03:06:31 Sgeo: You do realise I'm an op? 03:07:40 No you're not. 03:07:53 Sgeo: You don't know that. 03:07:58 Yes I do 03:08:22 Sgeo: Wrong. 03:08:25 Unless you're secretly andreou or Aardappel 03:08:59 Or lament or oerjan or fizzie, for that matter 03:09:10 Sgeo: Or it's just hidden in the access list. 03:10:08 I don't see why I should believe that that's possible 03:10:30 Sgeo: Are you *sure* you want a demonstration...? 03:10:38 Yes 03:11:26 Sgeo: Are you sure you know how unreliable my memory is at making me remember to reverse things? 03:11:35 Not doing it. :P 03:11:50 Bannation isn't the only way to demonstrate oppness 03:12:11 But let's pretend it is. 03:12:45 Sgeo: Well, considering I don't have make-op privileges, only ChanServ privileges... 03:12:58 -!- RandSgeo has joined. 03:13:09 Hello 03:13:23 Surely you can ban just my nick+web ness 03:13:48 RandSgeo: Well, yes, but I don't particularly care to. 03:14:08 If only I could offer something of value :/ 03:14:23 I don't exactly have a million dollars on me 03:14:41 Sgeo: You could never say the words I listed ever again >__> 03:15:24 How about for the period of a month 03:15:41 Sgeo: Or forever. 03:16:19 * Sgeo 's confidence should not be wavering 03:16:30 Confidence in what? 03:16:41 You're the one offering me things so you can *get banned*. 03:17:02 Confidence in the idea that you have no capacity to ban me 03:17:28 I have lost track of your layers of thought. Now why is the Ubuntu installation only at 31%? Could it go slower, theoretically? (No.) 03:18:14 Fine. I'll keep on assuming that you have no bannination powers 03:18:45 -!- RandSgeo has quit (Quit: Page closed). 03:18:54 *fake gasp* 03:27:08 $ CC="$K/stage1/bin/pcc -nostdlib -nostdinc -nostartfiles -static -isystem $K/stage1/include -D__dietlibc__ -Os -L$K/stage1/lib-x86_64 $K/stage1/lib-x86_64/start.o $K/stage1/lib-x86_64/libc.a" ./configure --prefix=$K/stage2 03:27:10 Dear god. 03:47:28 elliott@dinky:~/kitten$ du -sh --exclude=include --exclude=lib-x86_64 stage1 03:47:29 572Kstage1 03:47:29 elliott@dinky:~/kitten$ du -sh --exclude=include --exclude=lib-x86_64 stage2 03:47:29 556Kstage2 03:47:35 Static linking, it's what plants crave. 03:47:45 Now let's see if I can compile dietlibc with pcc. :p 03:52:08 My guess: No. 03:52:15 Gregor: Actually... pretty damn close. 03:52:27 Gregor: I've even managed to fix the "major internal compiler error" that made me have to compile a certain file with gcc before. 03:53:07 Also, damn, pcc is fast. 03:53:18 Gregor: Mostly what I have to fix is removing unsupported __attribute__s. 03:53:25 Such as the format(printf,x,y) stuff. 03:54:56 Gregor: Oh, there is this utterly queer issue though: 03:54:59 /home/elliott/kitten/stage2/bin/pcc -D__restrict__= -I. -isystem include -Os -fstrict-aliasing -momit-leaf-frame-pointer -mfancy-math-387 -W -Wall -Wextra -Wchar-subscripts -Wmissing-prototypes -Wmissing-declarations -Wno-switch -Wno-unused -Wredundant-decls -c libpthread/pthread_internal.c -o bin-x86_64/pthread_internal.o 03:54:59 libpthread/pthread_internal.c:397: error: If-less else 03:55:02 Note: No if-less elses in the file. 03:55:26 #define INTR_RETRY(e) ({ long ret; do ret=(long)(e); while ((ret==-1)&&(_errno_==EINTR)); ret; }) 03:55:26 #define __NO_ASYNC_CANCEL_STOP } 03:55:30 I am already inspired with confidence at this file. 03:55:53 if (__likely(__modern_linux==1)) 03:55:55 Still love this. 03:56:03 Just the way it reads. 03:59:30 Oh. 03:59:34 That's the C preprocessor complaining. 03:59:37 REAL UNAMBIGUOUS THERE 04:00:11 The way it reads isn't what it means 04:00:13 Haa, pcc can't count. 04:00:14 Can't say I love that 04:00:26 Sgeo: It's called amusement. 04:00:59 * Sgeo should learn how microprocessors work 04:03:14 /home/elliott/kitten/stage2/bin/pcc -D__restrict__= -I. -isystem include -Os -fstrict-aliasing -momit-leaf-frame-pointer -mfancy-math-387 -W -Wall -Wextra -Wchar-subscripts -Wmissing-prototypes -Wmissing-declarations -Wno-switch -Wno-unused -Wredundant-decls -c libpthread/pthread_sys_close.c -o bin-x86_64/pthread_sys_close.o 04:03:14 libpthread/pthread_sys_close.c:7: error: expdef: out of mem 04:03:29 The file is 9 lines long. 04:09:11 Gregor: Ha ha, it worked. Pay up. Whoops, look at that error. 04:10:48 Now it really did work. 04:11:55 configure:2642: /home/elliott/kitten/stage3/bin/diet -V >&5 04:11:55 execvp() failed! 04:12:01 Oh, whoops. 04:12:19 /home/elliott/kitten/stage3/include/dietref.h:-21: error: __PCC__ redefined 04:13:51 #if (__WORDSIZE == 64) 04:13:51 .quad __you_tried_to_link_a_dietlibc_object_against_glibc 04:13:52 #else 04:13:52 .long __you_tried_to_link_a_dietlibc_object_against_glibc 04:13:52 #endif 04:14:46 I wonder where exactly line -21 of that file is. 04:21:42 /home/elliott/kitten/stage3/include/dietref.h:-21: error: __PCC__ redefined 04:21:43 previous define: x.c:-21 04:21:44 i... 04:25:52 yay, fixed it 04:26:32 $ CC="$K/stage3/bin/diet -Os $K/stage2/bin/pcc" CFLAGS="-Os" ./configure --prefix=$K/stage3 04:26:51 actually CFLAGS="", diet adds -Os itself 04:35:05 WAIT WHAT 04:35:22 Or not 04:36:13 Sgeo: ? 04:36:45 Tried some Facebook thing that attempted to automatically import Google Contacts 04:36:47 No fuss 04:37:00 But it's clear that no contacts got exported 04:37:10 It claimed that everyone's already on FB or got an invite 04:37:21 And somehow, I doubt that Agora got a Facebook invite 04:39:31 Oh: "Today's IANA depletion date estimate:2011-02-21". That's months earlier than anything recent I have seen. 04:40:04 pikhq: I have a meta-compiled-to-the-N dietlibc/pcc toolchain working except for one... minor... detail. 04:40:06 Ilari: Nice. 04:40:25 Ilari: We should hold a party on IPv4 Doomsday (defined as: the day when the last allocation *really* gets made, ever, no delegation). 04:41:10 Based on this model: 2012-08-29 IANA ANNOUNCE ALL DEPLETED 04:42:14 pikhq: Specifically: Calling a macro like MACRO() doesn't work. Seriously: zero-argument macros don't work. I looked at the pcc code. malloc is failing. ?! 04:42:34 Ilari: Got any links to that model? Also, that's amusingly close to December 21, 2012. 04:43:29 elliott, Kitten is obviously having difficulty with more than a few bytes of memory 04:43:32 http://www.ipv4depletion.com/?page_id=77 04:44:04 Ilari: Thank you. 04:44:21 Ilari: I really need to switch to an IPv6 provider. 04:48:04 pikhq: gcc - 5.0K, pcc - 4.6K 04:48:06 0.3K BETTER 04:48:09 (pcc that is) 04:48:55 elliott: So, pcc produces smaller code. 04:49:00 Awesome. 04:49:09 pikhq: Well, with diet's -Os, yes. 04:49:22 pikhq: Now help me DEBUG: 04:49:24 if ((args = malloc(sizeof(usch *) * (narg+ellips))) == NULL) 04:49:24 error("expdef: out of mem"); 04:49:34 pikhq: WHY WOULD THIS (1) EVER HAPPEN (2) ONLY HAPPEN WHEN YOU CALL A ZERO-ARGUMENT MACRO AND NO OTHER TIME 04:49:37 WHYYYYYYYYYY 04:49:47 pikhq: A-HA. 04:49:54 pikhq: Perhaps that number ends up ZERO! 04:49:59 That would do it. 04:50:08 malloc(0) should return NULL... 04:50:10 if ((args = malloc(sizeof(usch *) * ((narg+ellips) || 1))) == NULL) 04:50:14 pikhq: Right, but glibc's doesn't. 04:50:16 And perhaps BSD libc too. 04:50:20 Now let's try recompiling. 04:50:28 elliott: Hmm. Actually, that may be undefined behavior. 04:50:37 Regardless. That's plausible. 04:50:43 pikhq: IIRC POSIX says that malloc(0) = NULL. 04:51:01 pikhq: Hmm, got a nicer way to write my ||1 there? 04:51:09 I could just do +1, but that's wasteful! :P 04:51:19 Oh, (narg+ellips || 1) should work. 04:52:23 $ CC="$K/stage3/bin/diet -Os $K/stage2/bin/pcc" CFLAGS="" ./configure --prefix=$K/stage3 04:52:42 pikhq: Ha, wait, I really need to start from scratch, as my *current* compiler can't handle it. 04:52:47 pikhq: Still -- I have the build process down, yo. 04:53:43 Ilari: Let's try and get the earliest date! 04:53:47 2011-03-22IANAANNOUNCEDEPLETED 04:53:52 2012-08-13IANAANNOUNCEALL DEPLETED 04:55:08 2013-04-24IANAANNOUNCEDEPLETED 04:55:12 2014-09-22IANAANNOUNCEALL DEPLETED 04:55:31 2015-06-15LACNICANNOUNCEDEPLETED 04:55:32 2016-01-24IANAERROR NO MORE DATA 04:56:33 -!- zzo38 has joined. 04:57:47 pikhq: Say, do you know what diff format a CVS-using project will like best? "cvs diff" spews an awful lot. 04:58:08 (Moreover it is ugly. :P) 05:01:43 Factor has <<<<< and >>>>> words etc. to raise an error if someone tries using a conflicted Factor file 05:05:56 Keywords that just raise errors and do nothing else? 05:06:28 Ilari: probably 05:06:41 Ilari: would be amusing if they started up an interactive merge, and then continued loading the file after it was resolved 05:08:36 http://docs.factorcode.org/content/word-__lt____lt____lt____lt____lt____lt__%2Csyntax.html 05:12:10 -!- augur has joined. 05:15:59 $ patch -p0 < ../dietlibc-for-pcc.patch 05:16:01 Yay, it works. 05:16:26 -if ((args = malloc(sizeof(usch *) * (narg+ellips))) == NULL) 05:16:26 +if ((args = malloc(sizeof(usch *) * (narg+ellips || 1))) == NULL) 05:16:33 Seriously though, this is ugly. Can anyone think of a better way to write that? 05:17:21 How that does even work? 05:17:52 Ilari: Sometimes narg+ellips is 0. pcc, being a naughty boy and defiant of standards, assumes that malloc(0) is a proper pointer, and malloc(N) being NULL for any N just means that we're out of memory. 05:17:59 Ilari: glibc has malloc(0) point to memory (!). 05:18:10 Ilari: But dietlibc doesn't. So, when narg+ellips = 0, the code is 0 || 1, which is 1. 05:18:17 That 'narg+ellips || 1' thing 05:18:26 Ilari: (narg+ellips || 1) = max(narg+ellips, 1) 05:18:33 Ilari: (narg+ellips || 1) --> (42 || 1) --> 42 05:18:39 Ilari: (narg+ellips || 1) --> (0 || 1) --> 1 05:18:43 Taking advantage of the fact that false=0, true=1 in C. 05:19:23 Eh... Where is that || -> max guaranteed? 05:19:35 Ilari: Erm, it's not; it only works in this case. 05:19:53 Ilari: It's just boolean logic. 0 || X = X. Y || X = Y iff Y>0. 05:20:07 Ilari: In this case, X = 1. So if Y=0, Y||X = 1. 05:20:41 Ilari: I can try and explain more :) 05:22:50 Lua 'or' operator is whacky: If first argument is "true", return that, otherwise return the second argument. ('and' returns second argument if first is "true", otherwise the first). 05:23:05 Ilari: That's actually ubiquitous in dynamic languages. 05:23:12 Python and Ruby do that too. I think Perl might. 05:23:26 Ilari: Also, I imagine that in Lua, (3 or 5) is 3. 05:23:34 Ilari: Since all non-"false" values will be considered true. 05:23:39 Rather than it checking for true specifically. 05:23:44 (Unless Lua is super-freaky.) 05:24:49 3 or 5 is indeed 3. 05:25:25 Bit whackier: 0 or 1 is 0. :-) 05:25:43 Yeah, most dynamic languages don't treat ints as bools either. 05:25:58 Python is an exception; it didn't use to have a bool type, and things like (0 or x) == x and also ([] or x) == x. 05:26:01 Lua has exactly two false values: nil and false. 05:26:07 *and, not and things like 05:26:10 Ilari: Same in Ruby. 05:26:31 Jesus! I just rm'd my patches by mistake. But I saved the big one. 05:31:10 1. Build pcc (stage1). 05:31:10 2. Build pcc-libs (stage1). 05:31:10 3. Build dietlibc with pcc (stage2). 05:31:10 4. Build pcc with dietlibc and stage1 pcc (stage2). 05:31:15 pikhq: See any holes in my bootstrapping there? 05:32:11 Not currently. 05:32:31 pikhq: Yay. Previously, I had three stages, but then I was playing it by ear. 05:33:05 That appears to be the typical bootstrapping procedure, actually. 05:33:57 I know that in Javascript, using || returns the left side if the left side is a value considered true, while && is the other way around. 05:34:07 pikhq: Here's my HOLY FUCKING SHIT CRAZY patch against dietlibc to make it build with pcc: http://sprunge.us/LJaX 05:34:22 - tcbhead_t* me=alloca(sizeof(tcbhead_t) 05:34:22 -#ifdef WANT_TLS 05:34:22 - +__tmemsize); 05:34:22 +#ifndef WANT_TLS 05:34:22 + tcbhead_t* me=alloca(sizeof(tcbhead_t)); 05:34:23 +#else 05:34:25 + tcbhead_t* me=alloca(sizeof(tcbhead_t)+__tmemsize); 05:34:27 pikhq: I cannot believe that is required. 05:34:34 pikhq: It makes me very suspicious of pcc's cpp. 05:35:08 Hey, those __TEST_CANCEL() changes are totally unneeded, now that I have the fix in pcc. 05:35:09 Sweet. 05:36:22 pikhq: Let me just say now that building things with dietlibc, even when it's difficult, is so insanely less painful than building them with gcc. 05:36:25 erm. 05:36:27 pikhq: Let me just say now that building things with dietlibc, even when it's difficult, is so insanely less painful than building them with uClibc. 05:36:34 uClibc does some kind of black magic and fuck that. 05:36:46 pikhq: Not that dietlibc is perfect... it warns you every single time you use stdio. 05:36:50 Well, printf. 05:36:50 When I asked which way you prefer to make GF-Magick work with GraphicsMagick, it is not a good answer to just say you like any ways. A better answer is the advance of one or more ways, and which way you would probably use if you were using this program. 05:37:04 zzo38: I am not sure. I would have to use it to see. 05:37:14 elliott: It warns about stdio? 05:37:35 Yes; stdio functions are rather big so it warns that they inflate the binary by 7k or so. 05:38:08 ./libugly/system.c:link_warning("system","warning: system() is a security risk. Use fork and execvp instead!") 05:38:21 pikhq: Gotta love a library that doesn't have system() by default... I am not sure dietlibc will work for the whole system. 05:38:51 ./libpthread/pthread_errno.c:link_warning("errno","\e[1;33;41m>>> your multithreaded code uses errno! <<<\e[0m"); 05:39:03 pikhq: I swear to god it highlights the warning with colour codes. 05:39:22 ./libcruft/tempnam.c:link_warning("tempnam","\e[1;33;41m>>> tempnam stinks! NEVER ! NEVER USE IT ! <<<\e[0m"); 05:39:31 elliott: Ah many functions such as printf would be large, so, perhaps put printf and stuff in a separate file so that you can link separately, and also have perhaps printf optimizing at compile time?? 05:39:44 zzo38: It does not include printf unless you use it. 05:40:19 pikhq: Aha, I can just disable WANT_LINKER_WARNINGS. 05:40:53 * that is realloc-able; means realloc(..,size) gives a NEW object (like a 05:40:53 * call to malloc(size)). 05:40:53 * WARNING: this violates C99 */ 05:40:57 erm 05:41:00 /* do you want that malloc(0) return a pointer to a "zero-length" object 05:41:01 * that is realloc-able; means realloc(..,size) gives a NEW object (like a 05:41:01 * call to malloc(size)). 05:41:01 * WARNING: this violates C99 */ 05:41:04 Even if you do use printf (and sprintf, and fprintf, and so on), that there could be a way of optimizing some uses of printf style functions at compile time 05:41:09 pikhq: So yeah, malloc(0) != NULL is invalid C99. 05:42:34 I think that whether malloc(0) == NULL should be up to the implementation, but that if malloc(0) == NULL then realloc(NULL,x) should do like malloc(x). At least this is my opinion of what it should do. 05:44:39 What is your opinion of this? 05:48:52 One night without a proper dinner shouldn't kill me 05:48:53 Right? 05:48:59 * Sgeo is too tired to eat 05:49:12 $ make CC="$K/stage1/bin/pcc -D__restrict__=" prefix=$K/stage2 05:49:13 Right. 05:49:16 The moment of truth. 05:49:34 *** glibc detected *** cpp: free(): invalid next size (fast): 0x0000000000de4d30 *** 05:49:47 pikhq: I... ??????????????????????? 05:50:15 If I'm dead in the morning I'm blaming Ilari 05:50:37 Sgeo: Have you *never* gone a night without dinner before or something? 05:51:05 Not that I can recall 05:51:09 o_O 05:51:25 Maybe.. once. Or I think I forced myself to eat a sandwich that night 05:51:35 Of course, when I was a little kid, all bets are off 05:52:08 Anyways, night all 05:52:50 -!- Sgeo has quit (Read error: Connection reset by peer). 05:54:01 Can you have a program that unpacks only what is used, to have both speed and small? 05:57:02 zzo38: Isn't that already how malloc(0) is specced to work? (Except that realloc(NULL, x) == malloc(x) even if malloc(0) does return an actual, free()able pointer.) 05:57:02 zzo38: Only unpacks what? 05:57:13 fizzie: malloc(0) is specified to return NULL in C99, I think. 05:57:23 fizzie: Or at least, "you're not allowed to use the return value of malloc(0), so STFU". 05:58:05 Description 05:58:05 2 The malloc function allocates space for an object whose size is specified by size and 05:58:05 whose value is indeterminate. 05:58:05 Returns 05:58:05 3 The malloc function returns either a null pointer or a pointer to the allocated space. 05:58:13 fizzie: Or let's just go with "the C99 draft is hopelessly vague on this point". 05:58:27 fizzie: Aha: " If the size of the space requested is zero, the behavior is implementation- 05:58:28 defined: either a null pointer is returned, or the behavior is as if the size were some 05:58:28 nonzero value, except that the returned pointer shall not be used to access an object." 05:58:33 elliott: "If size is 0, either a null pointer or a unique pointer that can be successfully passed to free() shall be returned." (Posix 2008, which is supposed to be aligned with C99) 05:58:37 Right. 05:58:40 fizzie: tl;dr malloc(0) can be anything, but you can't look at it. 05:59:07 fizzie: Anyway, assuming that malloc(0) == NULL means that you're out of memory is definitely wrong, so there. Now rewrite (foo+bar || 1). 05:59:44 Huh? It can't be "anything": it's implementation-definedly (means it has to be documented) either NULL or a real pointer you can free, just not dereference. 06:00:42 fizzie: Okay, true. 06:00:45 Missed that detail. 06:00:51 Apparently ARIN will grab its two blocks first... 06:00:59 fizzie: Now you get to tell me why free(args) would fail after doing: 06:01:01 if ((args = malloc(sizeof(usch *) * (narg+ellips || 1))) == NULL) 06:01:05 fizzie: where narg+ellips = 0. 06:01:07 fizzie: Gogogogogogo. 06:01:16 (And no, there are no previous frees.) 06:01:20 *** glibc detected *** cpp: free(): invalid next size (fast): 0x0000000000de4d30 *** 06:01:23 is the inexplicable error. 06:01:26 elliott: O, that is how the C99 draft works? In my opinion how it should work is that if malloc(0) returns NULL then you can free(NULL) safely and that if malloc(0) returns NULL then realloc(NULL,x) acts like malloc(x), but that if malloc(0) does not return NULL, then its return value must be safely pass to free() but 06:01:53 that in case of malloc(0) != NULL then there is no guarantee that free(NULL) or realloc(NULL,x) has to be meaningful. 06:02:06 zzo38: You can free(NULL) and realloc(NULL,x) safely in all cases, is how it works. 06:02:10 At least this is what makes sense to me. 06:02:17 fizzie: SRSLY WHUT ;__; 06:02:55 elliott: get some sleep 06:03:28 coppro: fuck you fix my bug 06:03:37 fizzie: If you can safely free(NULL) and realloc(NULL,x) safely in all cases, then it should make sense that malloc(0) should return NULL (although the Posix 2008 note above that says "either a null pointer or a unique pointer that can be successfully passed to free() shall be returned" is very sensible) 06:03:42 elliott: Tried valgrinding it yet?-) (Can' see any obvious problems there, except that (nargs+ellips || 1) is always 1.) 06:03:52 fizzie: "(Can' see any obvious problems there, except that (nargs+ellips || 1) is always 1.)" 06:03:54 fizzie: Gaah, of course. 06:04:06 fizzie: Remember when I said "rewrite that"? Rewrite it so that it means max(nargs+ellips, 1) plz :P 06:04:14 It's not a Perl ||. :p 06:04:30 fizzie: I suppose I could do nargs+ellips ? nargs+ellips : 1, but, er, ew. 06:05:11 You can do ((nargs+ellips) | 1) if you don't mind that it's sometimes one larger than necessary. :p 06:05:23 fizzie: If I didn't mind that, I'd do something called +1. 06:05:51 But then it's one larger always; at least my thing is that only half of the time. 06:05:58 fizzie: I want this to be hypothetically accepted into pcc, and I imagine the BSD guys might be a bit CAGEY about wasting sizeof(usch *) bytes like that. 06:07:53 fizzie: Factored it out into a conditional like the cool dudes. 06:10:09 Is it possible to make Frama-C to work without the special ACSL but instead by commands that can be inserted anywhere in the C code where a statement is expected: assume(), assert(), assume_portable(), assert_portable(), etc. 06:10:39 ((nargs+ellips)|1)^(((nargs+ellips)&~1) && !((nargs+ellips)&1)) for maximum confusion. 06:10:54 fizzie: I am so glad the if statement was invented. 06:10:57 SO GLAD 06:11:45 fizzie: However, for your ASSISTANCE, you are now an official Kitten developer. Expect drudge work. 06:12:20 elliott: Okay, it's really odd that you would actually *want* an allocation for a malloc(0)... 06:12:33 pikhq: Indeed. 06:12:36 elliott: Especially with the reasoning that you might want to pass the result to realloc... 06:12:42 realloc of NULL works just fine. 06:12:47 pikhq: I am so tempted to build dietlibc with malloc(0) == pointer turned on, though. 06:12:52 pikhq: Just to give me less shit. 06:13:18 Well, it's *permitted* behavior, and apparently some crazy junk relies on it, so I say go for it. 06:13:18 pikhq: (Similarly, although I've disabled the GNU program invocation name crap that only util-linux and GNU software uses, I'm tempted not to...) 06:13:29 pikhq: But then why use dietlibc :) 06:13:45 Okay, go for it if it ends up actually being problematic. 06:13:59 Say, "HOLY FUCK I NEED TO PATCH EVERY SINGLE PROGRAM". 06:14:19 If realloc of NULL works fine, then the Posix specification that malloc(0) should either return NULL or a pointer passable to free or realloc, works fine. (It says "a unique pointer that can be successfully passed to free()", but it ought to include realloc() as well, or else it isn't very good) 06:14:53 zzo38: That's pretty bizarre specing there. 06:14:57 * file support on kernel 2.2 or 2.0 */ 06:15:00 er 06:15:03 /* you need to define this if you want to run your programs with large 06:15:03 * file support on kernel 2.2 or 2.0 */ 06:15:07 dddisable 06:15:25 Disable-tastic. 06:15:39 Compile! 06:16:03 pikhq: I am shocked to the core that this shit actually works, and in fact all the patches I've made could be made into sane ones that condition on __PCC__ or something rather than just chainsawing up dietlibc. 06:16:15 Still, the dietlibc patch is only 200 or so lines; the pcc patches are both tiny. 06:17:37 pikhq: What is bizarre? 06:17:42 Vorpal: night, GENERA LOSER <-- ? 06:17:51 elliott: That's pretty nice. 06:17:54 Vorpal: Yellow: "Polynomial!" <-- heh 06:18:23 elliott: You might want to clean up the patches a bit and send them upstream. 06:18:53 Vorpal: night, GENERA LOSER <-- ? 06:18:55 I've forgotten :P 06:19:09 pikhq: I'm going to send them as-is; they're not *too* messy and I don't know how they'd want to do this stuff, so, leave it to the maintainer! 06:19:21 -!- TLUL has joined. 06:19:21 pikhq: The pcc ones need no clean-up and are totally unobjectionable. 06:19:44 pikhq: Well, except that the pcc-libs one mentions the name "dietlibc"; sure, it also mentions SunOS in that file, but perhaps mentioning GPL'd software is just TOO FAR for the BSD guys. 06:20:24 elliott, what are you doing? 06:20:43 elliott@dinky:~/kitten$ $K/stage2/bin/diet -Os $K/stage2/bin/pcc hello.c -o hello 06:20:43 execvp() failed! 06:20:48 Whoopsy. 06:20:51 Oh, I neglected to make install. 06:21:14 Vorpal: Building the Kitten toolchain. 06:21:19 elliott, dietlibc? kitten will use dietlibc? 06:21:25 Vorpal: I have dietlibc built by pcc, and pcc built by pcc linked with dietlibc. 06:21:29 Vorpal: Yes, probably. 06:21:43 elliott, huh, what is wrong with, say, newlib or uclibc? 06:21:50 Vorpal: (It *is* legal to distribute the binaries, as long as I offer source; felix is mistaken about this.) 06:22:01 elliott, felix? 06:22:08 Vorpal: the dietlibc developer 06:22:09 Vorpal: newlib's Linux code is just... glibc's. Also it's primarily for standalone environments. 06:22:23 elliott, and uclibc? 06:22:35 Vorpal: uClibc is... well, let's put it this way: They use goddamn menuconfig. And if you don't compile your own gcc, your resulting programs segfault. 06:22:52 Vorpal: Also, uClibc doesn't build with pcc. dietlibc does, with some patches that I have written. 06:22:57 elliott, uh? do they? I used it with my normal system compiler 06:23:00 I would not want to patch uClibc. I hear it is very gcc-specific. 06:23:02 (uclibc that is) 06:23:09 Vorpal: Well, it was the issue I had. But whatever. 06:23:13 right 06:23:27 Vorpal: Also, dietlibc tends to produce leaner executables, I think. Admittedly it is less strictly compatible with stuff. 06:23:39 I will have to see how much stuff breaks with it. I doubt X.Org will build with anything but gcc. 06:23:47 elliott, how feature complete is dietlibc wrt posix? 06:24:20 Vorpal: Quite; by default, it tells you off for using stdio and other functions (they bloat the binary), but I've disabled that. Some functions like system() are in a separate library because they suck. 06:24:26 Vorpal: (But that's just a link to fix.) 06:25:03 pikhq: Vorpal: It works! http://sprunge.us/WYFe 06:25:17 What might be good is having a Frama-C that will create reports that can be included into a Enhanced CWEB program. (You first run tangle, and then analysis, and then weave, and you will get the reports included in a new chapter of the program.) 06:25:22 mhm 06:25:24 bbiab 06:25:31 pikhq: Now I just need a cc(1) wrapper script that works like diet(1), except that it doesn't behave differently depending on where -Os is. 06:25:42 pikhq: (diet's -Os switch appends a few extra arguments depending on your arch.0 06:25:44 *.) 06:25:58 actually, make that "bbl, several hours" 06:26:53 -rwxr-xr-x 1 elliott elliott 4.4K Nov 12 06:24 hello 06:26:53 -rwxr-xr-x 1 elliott elliott 6.5K Nov 12 06:26 hello2 06:26:59 pikhq: Second is gcc -Os, dynamically linked to glibc. 06:29:29 pikhq: I just had the most perverse idea. Compile plan9port with this. 06:29:43 ((nargs+ellips)+!(nargs+ellips)) also occurs to me, but I think they're all messier than the conditional. 06:30:26 elliott: Please do. 06:31:03 pikhq: How depressing: I just realised I'm spending all of this time and effort on Unix. Not even that, Linux. 06:31:09 Maybe I'll shoot myself when all's said and done. 06:31:42 elliott: It's at least non-sucky Unix. 06:31:49 It's Linux. 06:31:57 ... Okay, except for the kernel. 06:32:06 dietlibc is Linux-only :P 06:32:13 Baaah. 06:38:10 pikhq: What does irritate me is that I'm going to have to wrap pcc. 06:38:25 Why don't compilers support an easy way to compile them to automatically link to a static libc? *sniff* 06:42:16 pikhq: LETS FORK PCC 06:59:16 pikhq: Haw, plan9port does not like diet pcc :P 07:01:04 pikhq: http://www.fefe.de/embutils/ almost coreutils 07:02:31 pikhq: Can I poll you about my service manager's interface? Or should I shut up and code :p 07:05:29 -!- TLUL_ has joined. 07:06:50 -!- Zuu has quit (Ping timeout: 260 seconds). 07:07:23 -!- TLUL_ has left (?). 07:16:05 -!- TLUL has changed nick to Teh_Lawl. 07:16:11 -!- Teh_Lawl has changed nick to TLUL. 07:16:31 pikhq: BAH 07:33:11 -!- wareya_ has joined. 07:33:12 -!- wareya has quit (Ping timeout: 255 seconds). 07:35:24 -!- TLUL has quit (Quit: *disappears in a puff of purple smoke*). 07:37:35 Is && a sequence point? 07:37:38 i.e. is 07:37:41 *q++ && *q == foo 07:37:41 ok? 07:37:43 *!= 07:39:14 -!- ttm_ has joined. 07:40:30 It is. 07:41:43 "Unlike the bitwise binary & operator, the && operator guarantees left-to-right evaluation; there is a sequence point after the evaluation of the first operand. If the first operand compares equal to 0, the second operand is not evaluated." 07:42:30 Yay. 07:53:59 Can poker be combined with Magic: the Gathering in the following way: That whatever cards you have available to make up a poker hand are the cards you play with in your hand in Magic: the Gathering cards. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:03:11 fizzie: I have used things like (x+!x) in C codes, before (occasionally; not all the time). 08:04:49 Writing (x|!x) should also produce the same answer (I think), also (x^!x), and if you are using GNU C, also (x?:1) 08:05:36 x>>=!y; 08:09:14 -!- zzo38 has quit (Quit: *reappears in a ball of invisible ink*). 08:13:18 The slowness of qemu-system-x86_64 is profound. 08:15:15 I am staring at one gnome panel while waiting for the other to start. 08:15:15 Seriously. 08:16:15 WOO THE BACKGROUND IS UP 08:54:22 Gregor: http://beesbuzz.biz/art/web/statistically_significant_2.php What an average Dinosaur Comic. 09:57:13 i should just start using ie6 for all my linux web browsing 09:57:18 just to be contrarian 10:10:58 Okay, seriously, does anyone know of a faster x86-64 emulator than qemu? 10:20:33 -!- augur has quit (Ping timeout: 276 seconds). 10:21:00 pikhq: BE MORE AWAKE 10:25:24 -!- augur has joined. 10:30:10 augur: ed is the best editor 10:30:37 elliott: what 10:30:42 it is 10:44:44 -!- ineiros has quit (Ping timeout: 240 seconds). 10:46:34 -!- ineiros has joined. 11:40:01 Vorpal: ping 12:03:50 -!- ais523 has joined. 12:04:57 hi ais523 12:05:04 hi elliott 12:05:16 I managed to do a rm * in my home directory last night (thought I was in /tmp...) 12:05:22 luckily, I lost hardly anything, maybe about 15 minutes of work 12:05:36 (an rm -r * would have been rather more destructive...) 12:06:57 (and the reason I was doing an rm * of /tmp is that I'd just tarbombed it) 12:07:11 (good thing I forgot the -rf, which would have been needed...) 12:07:53 ais523: oh dear... 12:08:18 it's at times like this that I'm glad I'm using my crazy Emacs backup system, that's designed specifically to protect against accidental rm *s 12:08:26 (it backs up all files I edit in a /different/ directory from the file itself) 12:08:52 ais523: err, I do that too 12:08:58 ais523: and many, many others 12:09:03 ais523: but -- mostly to stop cluttering up directories. 12:09:14 indeed, clutter in directories is what causes the rm * ~ typo in the first place 12:09:14 ais523: (in fact, i'd say that in the "emacs community", >50% do that) 12:09:25 really, it should be the default, backing up to a dotfile somewhere 12:09:34 ais523: emacs isn't exactly one for sane defaults 12:09:43 indeed (GNU-style indentation?) 12:09:46 I've almost got a basic .emacs in muscle memory from reinstalling so many times 12:09:57 heh 12:10:09 (tool-bar-mode -1) (menu-bar-mode -1) (setq inhibit-splash-screen t) (setq-default indent-tabs-mode nil) 12:10:21 anyway, it turns out that the only actually important thing I lost in the rm * of my homedir, as far as I can tell, was the public key for accessing the wireless here 12:10:24 *search "amit patel backups emacs", copy snippet in* 12:10:26 tada 12:10:36 ais523: probably retrievable/regeneratable? 12:10:36 and I managed to download that over JANET roaming 12:10:39 right 12:10:55 indeed, the issue is, I thought I couldn't connect to the Internet to download it without it 12:11:03 ais523: uh, care to take bets on whether ubuntu 7.10 will recognise this fancy new ethernet card that supposedly older ubuntus didn't? 12:11:11 In fact, it might handle my wifi but not ethernet. Wouldn't that be a thing. 12:11:18 (And yes... I need 7.10. For Genera!) 12:11:21 7.10? that's pretty old 12:11:32 ais523: every newer version has a too recent X.Org 12:11:38 and as a result, I have no idea 12:11:47 ais523: which breaks the hideous monstrosity of a hack^W^W^W^W^WOpenGenera port to Linux 12:11:57 elliott: use an old X.Org as an application within Wayland! 12:12:00 ais523: (it actually replaces the Alpha assembly-generating routines with routines that generate *C*...) 12:12:11 ais523: hey now, i support the move to wayland :P 12:12:17 why? IT'S NOT X 12:12:30 I don't yet oppose it, but I don't have enough data to make up my mind 12:12:34 I'd probably prefer Windows Display Server, the Linux port than X. 12:12:45 (Also known as "1/3 of the Windows kernel, the Linux port".) 12:13:14 Windows keep replacing their display code, though 12:13:19 I still remember GDI32.DLL 12:13:25 in fact, I still remember the 16-bit version 12:13:32 but Microsoft have deprecated not only it, but most of its successors 12:13:40 which is a pity, as I actually liked GDI 12:13:43 ais523: afraid i have to reboot to get debian to realise i have a usb stick now... 12:13:51 ugh 12:13:56 ais523: (not sure why it didn't just work, but) 12:14:05 The Global Defense Initiative. 12:14:22 ais523: I'm even using my Kitten work partition to install 7.10 on! What blasphemy. (In other Kitten news you don't care about, I got a bootstrapped pcc/dietlibc -- pcc built with pcc, linked with dietlibc -- working today.) 12:14:55 for my benefit: 12:14:55 http://sprunge.us/ZAWh 12:14:56 http://sprunge.us/ZAWh 12:14:56 http://sprunge.us/ZAWh 12:14:56 http://sprunge.us/ZAWh 12:15:05 -!- elliott has quit (Remote host closed the connection). 12:15:16 hmm, why would you paste that four times? to make it easier to see? 12:15:21 it's a mystery! 12:17:12 Maybe one of us should paste the output of "banner 'hey elliott here is your paste'" on-channel to make it even more visible. 12:18:11 I think elliott uses a proportional-width font for IRC (although I'm not sure), so it'd probably get munged 12:18:13 With -w 30 it's only 207 lines. 12:18:32 arguably, though, he or she uses a fixed-width font for the logs 12:18:40 I somehow assumed he'd be looking for it in the logs, which, ... right. 12:19:29 # # 12:19:29 # # 12:19:29 ### #### 12:19:29 ## ####### 12:19:29 ## ###### # 12:19:30 ###### 12:19:32 ######### 12:19:34 # ####### ## 12:19:36 ###### ## # 12:19:38 ### # 12:19:40 # # 12:19:42 The X marks the spot, you see. 12:19:47 that's an X? 12:19:58 Banner prints things sideways. 12:20:10 it's not an X any direction I look at it 12:20:20 hi 12:20:20 It has serifs and all. 12:20:28 Vorpal: ping <-- pong 12:20:32 ah, serifs + proportional-width font is what's confusing me 12:21:09 it's at times like this that I'm glad I'm using my crazy Emacs backup system, that's designed specifically to protect against accidental rm *s <-- I use rm -I 12:21:20 also shell function that wraps rm 12:21:25 Vorpal: that likely doesn't help, I did rm * deliberately but in the wrong directory 12:21:30 and examines the arguments for sanity 12:22:16 With -I you'd see the file names, though. Maybe seeing rm: remove regular file `important_stuff'? would help you notice sooner. 12:22:17 for example it will do nothing if one of the files is a guard file in ~ 12:22:26 which is just to avoid rm * in the wrong dir 12:22:41 also if an argument is ~ or $HOME 12:22:58 -I doesn't show the filenames 12:22:59 to avoid rm * ~ 12:23:00 or such 12:23:00 -i does, but not -I 12:23:16 ais523, indeed, but -i is annoying when you have multiple files 12:23:34 I do like the idea of a guard file which stops rm working; you could design it so you needed to use unlink instead, say 12:23:34 anyway, the shell wrapper protects against the most common stupid mistakes 12:24:02 ais523, well, it just examines the arguments. I could bypass it by calling /bin/rm rather than just rm for example 12:24:02 Something like http://www.nilfs.org/en/ would make your file system also mistake-proof. 12:24:11 I still haven't had the occasion to try that out. 12:24:17 fizzie, I wanted to try it too 12:24:28 and never got around to it 12:24:35 fizzie, isn't it in the kernel nowdays? 12:24:40 as in 12:24:43 vanilla 12:25:03 Yes, I think it is. 12:25:20 I wonder how good the performance is 12:25:23 -!- sebbu2 has joined. 12:25:33 -!- sebbu has quit (Ping timeout: 272 seconds). 12:25:35 "NILFS was merged into the Linux kernel 2.6.30. -- NILFS2 is available in Ubuntu 9.10 or later (karmic, lucid)." 12:25:56 -!- sebbu2 has changed nick to sebbu. 12:26:33 There's that one year-old benchmark article. 12:26:54 fizzie, hm is nilfs2 not in vanilla? 12:27:00 assuming they are different 12:27:12 They're the same thing in that context. 12:27:15 ah 12:27:36 Just two different snippets of the page, one talking about where the tools and such are packaged, the other about the file system. 12:28:03 http://www.nilfs.org/en/current_status.html <-- fsck on todo list 12:28:04 heh 12:28:49 I think I'll stay with my mix of ext4 and jfs partitions for now. 12:32:06 Can't seem to find any recent benchmarks. I just like the idea of continuous-snapshotting. 12:34:31 Our work /home share has /home/.snapshot/hourly.0 .. hourly.3, nightly.0 .. nightly.3 and weekly.0, weekly.1 which provide complete snapshots of everyone's ~ at the times you'd expect from the names. 12:35:37 It's not exactly continuous snapshotting, but probably good enough to catch many mistaken deletions, unless you're unlucky enough to have both made and deleted the stuff within the same one-hour period. 12:42:19 -!- FireFly has joined. 12:43:39 -!- Leonidas has quit (Ping timeout: 264 seconds). 12:43:43 -!- Leonidas has joined. 13:08:53 a hidden file in /home is an interesting way to violate the FHS 13:08:59 shouldn't it technically be in /var? 13:10:48 Possibly, but I think due to technomagical reasons it needs to be in the same NFS share, which is mounted at /home. 13:11:11 And, well; our /home is a symlink to /m/fs/home anyway. 13:11:43 What is even stranger is that /home/.directory is a symlink to /etc/kubuntu-default-settings/directory-home. 13:11:49 I have no clue what that's about. 13:12:31 It contains: 13:12:32 [Desktop Entry] 13:12:32 Icon=user-home 13:12:39 So I guess it's some sort of desktop environment magic. 13:21:42 -!- ais523 has quit (Quit: Page closed). 13:22:55 -!- elliott has joined. 13:22:57 04:28:49 I think I'll stay with my mix of ext4 and jfs partitions for now. 13:23:00 weren't you an xfs user? 13:23:07 joined the light side? :) 13:24:20 Vorpal: Genera update: qemu-system-x86_64 is slower than you can possibly imagine -- and yes, slower than that. Utterly unworkable; even booting was a five-minute operation and even starting an xterm in a Xephyr on the root machine was way too slow. 13:24:37 Vorpal: Then tried doing it as a chroot, ended up changing my system hostname and starting nfs, and on top of that swap4 didn't work so fuck that. 13:24:47 Vorpal: Now I'm going to install Ubuntu 7.10 literally onto another partition. 13:25:03 brb 13:27:06 -!- elliott has quit (Read error: Connection reset by peer). 13:33:59 -!- sftp has joined. 13:37:02 weren't you an xfs user? <-- I use xfs on some partitions. Those with few, but huge, files. So for the partition with disk images basically 14:14:25 -!- digimunk has joined. 14:24:55 -!- Phantom_Hoover has joined. 14:35:29 -!- elliott has joined. 14:35:35 Indeed, the ethernet driver was added in karmic. *sigh* 14:35:57 Vorpal: Are you *sure* Genera doesn't work with a more recent X? Anyway, wouldn't it be easier just to install Gutsy's X package onto Maverick? 14:38:43 __████████___ ██ 14:38:44 _█▓▓▓▓█░▓▓███▓█ 14:38:44 __█▓▓▓█░░▓▓▓█▓█ 14:38:44 ___█▓▓█░▓▓▓▓▓▓█ 14:38:44 ____█▓▓▓▓▓▓▓▓▓█ 14:38:44 ___█▓▓▓▓▓__▓▓▓▓█ 14:38:46 __█▓▓▓▓▓___▓▓▓▓█ 14:38:50 _█▓▓▓▓▓_░__▓░__█ 14:38:52 █▓▓▓▓▓▓_█__▓█__█ 14:38:54 █████▓▓_░__▓░__█ 14:38:56 ___█▓▓▓░░░░░░███ 14:38:58 __█▓▓▓▓░░░░░░░█ 14:39:00 _███████░░░░░█ 14:39:02 ___██░░░█████ 14:39:04 __█░░███░░░█ 14:39:05 fizzie 14:39:06 fizzie 14:39:06 _█░█__█▓░░░██ 14:39:06 fizzie 14:39:06 fizzie 14:39:08 _██____▓░░█__█ 14:39:10 __█____▓▓██__█ 14:39:11 fizzie 14:39:11 fizzie 14:39:12 fizzie 14:39:12 fizzie 14:39:12 fizzie 14:39:12 ___█__█▓██__█ 14:39:14 ____███▓█▓██ 14:39:15 fizzie 14:39:16 _______█▓█▓█ 14:39:20 _______█▓█▓██ 14:39:22 ______█___█__█ 14:39:24 ______█▓▓▓█▓▓███ 14:39:27 _____█____▓█▓▓_▓▓█ 14:39:28 ____█▓▓▓▓_█▓▓_▓▓▓█ 14:39:30 ____██████████████ 14:39:32 fizzie fizzie fizzie fizzie fizzie 14:39:32 Is that... Sonic? 14:39:37 yes 14:39:44 the coolest hedgehog! 14:40:06 The illest hog. 14:44:01 Uh, oh: "I think I’m going to write a post about the fact that Sprint allocated around 4 million v4 addresses yesterday." --ipv4depletion 14:44:16 Ilari: Looks like the bank run has begun! 14:44:36 If it begins in APNIC region... 14:45:41 Vorpal: I'm going to try this whole wazoo with 10.10 and an old X package. 14:46:10 You could get double the resolution with U+2596..259F, but that'd lose the shading. 14:46:17 ▟▄▟▖ 14:46:17 ▟▙█▙█▖ 14:46:17 ▌▛▀▀▌▌ 14:46:17 ▝ ▝ 14:47:25 ooh they're playing space invaders 14:47:49 It's a bit elongated with this font: it's not exactly a rectangular block. 14:48:02 Er, square, I mean. 14:48:26 is it ment to be? 14:48:54 The pixels should be square-ish, yes, I think. I just copied them directly to the quadrant-blocks, which aren't. 14:49:12 * yorick is implementing printf in c 14:49:46 What a strange language to implement printf in. 14:49:59 elliott: true, but I'm wondering what else 14:50:11 yorick: "what else"? 14:50:15 To change the "endgame" with IANA allocations, ARIN (which will probably request&get 2 blocks soon), APNIC or RIPE would have to allocate so fast that they can get 2x2 blocks before IANA runs out... 14:50:20 elliott: yes 14:50:28 yorick: no, as in, please restate so i can understand that 14:50:41 elliott: I have myself a set of "asm, C, C++" 14:51:05 why would you ever touch c++ 14:51:06 I can pick any (combination) of those to write a printf function 14:51:11 yorick: do printf in haskell! 14:51:13 and export it to C 14:51:16 elliott: templates. 14:51:24 yorick: yes, that's a reason not to use C++ ever, correct 14:51:38 I rather like templates 14:51:53 (especially when they can do the world some good compile-time instead of runtime) 14:52:16 Actually, severe run on the bank scenario on ARIN could force second set of allocations before IANA pool runs out, changing the endgame. 14:52:54 yorick: c++ makes the amount of patience and energy you have left to do anything after compiling shoot itself 14:53:56 elliott: most of the time, the only thing I actually have to do after compiling is see if it still boots and else fix it 14:54:05 What do templates do again? 14:54:27 they're like the C preprocessor 14:54:35 Phantom_Hoover: You know what a generic is? 14:54:46 elliott, in the CLOS? 14:54:46 but much more C++-y (and they have recursion :)) 14:54:50 Phantom_Hoover: No, in e.g. Java. 14:54:54 Phantom_Hoover: List is a list of Ts. 14:55:10 yorick, ah, so they're trying to solve a problem Lisp solved 50 years ago. 14:55:24 Phantom_Hoover: You know what a typeclass is? Imagine a feature that makes doing what generics do (simple parameterised typed) a pain in the ass, and not letting you do any useful typeclass stuff, but *yes* letting you abuse it as the most tarpitty thing ever. 14:55:25 http://pastebin.com/UF0Ci5F6 14:55:47 Phantom_Hoover: Also it's turing-complete and makes compile times go up faster than ... uh ... a really quickly-erected structure. 14:56:13 Than a self-assembling skyscraper? 14:57:03 I suppose some people might call it that. 14:57:05 http://pastebin.com/UF0Ci5F6 14:57:28 as we can see by this, templates are a verbose, strangely-arbitrarily-limited functional language, masquerading as a simple generics system. 14:57:41 not laughing at this too much produces Boost 14:57:43 *enough 14:58:53 So it is basically just a crippled, awful and inelegant version of Lisp's macros? 14:59:15 crippled...meh 14:59:17 Phantom_Hoover: It can't transform code. So no. 14:59:22 yorick: Yes crippled (you can't do SKI calculus). 15:00:05 elliott, its functionality is entirely a subset of macros, though? 15:00:12 umm. 15:00:14 that question makes no sense 15:00:57 Well, does it do anything in C++ which Lisp can't do with macros? 15:01:12 Vorpal: Are you *sure* Genera doesn't work with a more recent X? Anyway, wouldn't it be easier just to install Gutsy's X package onto Maverick? <-- pretty sure. Could work with some in between the two I tried (that old ubuntu version I mentioned in the "guide" and the X version in the next ubuntu release after that) 15:01:19 -!- augur has quit (Remote host closed the connection). 15:01:25 The answer is obviously going to be no, since you can do _everything_ with macros, but anyway... 15:01:26 Vorpal: is it the server or the libs that is the issue? 15:02:23 elliott, unknown. I didn't investigate it further than noticing that googling indicated it was X related, and that fitted with the error message, which was some X message (forgot which) 15:07:05 -!- MigoMipo has joined. 15:09:59 elliott, xkcd is hiatusing. I'm not sure whether I'm allowed to be happy. 15:10:28 Considering there's a family illness and also nobody's forcing you to read xkcd, I think not. 15:11:10 Yeah, I was thinking that. 15:15:23 Phantom_Hoover: The xkcdsucks people are all going "THANK GOD NO XKCD", but as is well-established they are all (a) douchebags and (b) morons. 15:18:10 Indeed. 15:28:05 Also, (d) fluffy puppies. 15:32:24 (We don't know what (c) is yet) 15:32:46 -!- augur has joined. 15:35:55 -!- Sgeo has joined. 15:37:18 I was about to confusedly ask something along the lines of "Is the amount of infinite sets of rationals uncountably infinite somehow? Because if not, I don't see how reals defined in terms of Dedekind cuts are uncountably infinite" 15:37:46 But then, I _think_ the diagonalization thingy says that there is an uncountably infinite amount of those sets 15:37:50 Which is head-breaking 15:38:53 Infinite sets of rationals? 15:39:30 Sgeo: the amount of infinite subsets of rational numbers is uncountable 15:40:03 That seems ... unintuitive, I guess 15:40:12 But intuition counts for nothing, really 15:40:19 well there's an uncountable number of subsets of integers 15:41:01 look at all numbers of the form 0\.[0-1]+ 15:41:13 suppose they're countable, so line them up 15:41:15 or just do diagonalization right away, it's easier than for reals 15:41:47 in fact in general, it's easy to show that for any set S, there is no surjection S -> 2^S 15:41:52 now, you can define a new number, which flips 0 to 1 and 1 to 0 on the nth digit of the nth number. 15:42:13 in fact in general, it's easy to show that for any set S, there is no surjection S -> 2^S 15:42:16 either i'm tired or wrong way around 15:42:38 you're tired 15:42:41 hard to say which one 15:42:52 wait 15:42:59 i misunderstood you 15:43:01 you're tired 15:43:07 surjection means the function values are the whole of the codomain 15:43:20 oklopol: oh right my brain ignored "surjection" 15:43:25 kinda important part of the sentence 15:43:28 :D 15:43:38 well there's a function the other way around too 15:43:42 According to my flawed understanding of Wikipedia, the left can be larger than the right in a surjection 15:43:45 oklopol: what i saw: "set S is no S -> 2^S" 15:43:53 oklopol: now you're fucking with me 15:43:55 or i just don't understand you 15:43:58 or i'm *really* tired 15:44:15 Sgeo: yes. but can't be of smaller cardinality. 15:44:28 elliott: you can take a function that takes everything to s \in S, for some element s in S 15:44:39 maybe you don't know what a function is? 15:44:51 Sgeo, a function f : A → B is surjective iff forall x:B, exists y:A, f y = x. 15:45:04 So yes, the left can be larger. 15:45:10 a function from X to Y is a subset Z of XxY such that for each x \in X there is exactly one y \in Y such that (x, y) \in Z 15:45:18 oklopol: i know what a function is :p 15:45:25 elliott: okay 15:45:29 oklopol: but yes, yes 15:45:31 i misread you 15:45:32 because tired 15:45:36 surjection is >=, injection is =<. 15:45:45 How does anyone remember all the jections? 15:45:53 study maths 15:45:59 Sgeo, surjection is what I just said. 15:46:00 "sur" is french for "on" 15:46:03 there's only surjection, injection, bijection. 15:46:04 that's all. 15:46:08 Oh 15:46:11 and in english, onto is a synonym for surjection 15:46:15 you call surjections "functions onto" 15:46:20 yes 15:46:33 Injection is basically the same as surjection except the other way around. 15:46:43 injections are also called one-to-one 15:46:46 no, it's not. 15:46:49 Phantom_Hoover: no 15:46:56 you're confused. 15:46:58 injection = at most one preimage 15:46:59 oklopol, yes, I just realised that. 15:47:03 surjection = at least one preimage 15:47:05 No, I know the actual definition. 15:47:15 (per point) 15:47:16 I was oversimplifying to the point of incorrectness. 15:47:21 yeah 15:47:22 Isn't it bijection that's 1 to 1? 15:47:30 Sgeo: it's a confusing name, yes 15:47:43 bijections are often called "one-to-one correspondence" 15:47:48 :D 15:47:57 So what was " injections are also called one-to-one" about? 15:47:57 but for some reason one-to-one just means injection 15:48:04 injections are 1-1 15:48:04 Injection is forall x y:A, x!=y->f x!=f y, yes? 15:48:11 HackEgo: yes 15:48:12 Phantom_Hoover: yes 15:48:15 *phanqpqokiepok 15:48:39 If something is both a surjection and an injection, is it a bijection? 15:48:44 that's the def 15:48:57 sur = at least one preimage, inj = at most one, bij = one 15:48:58 Sgeo, yep. 15:49:50 now once you get to the definition of the continuous function 15:49:53 that gets confusing 15:49:56 they should be called heterojections, homojections and bijections really 15:49:58 that'd be easier to remember 15:50:11 basically during undergrad you learn 2 new definitions of continuous functions per year 15:50:15 or one per term or so 15:50:21 continuous functions have 1 definition 15:50:49 The topological one? 15:50:51 tell that to kuratowski 15:51:03 what are there besides the topological one? 15:51:12 um helloo 15:51:15 maybe i'm just not aware of others 15:51:17 cauchy? 15:51:28 oh there are multiple characterizations, yes 15:51:31 I don't know; I know the one for functions and derivatives and stuff is the same. 15:52:04 i don't know what cauchy is tho, do you mean limit's image is images' limit 15:52:08 epsilon-delta, sequence-based, topology, algebraic (element of ...), category theory 15:52:23 cauchy is epsilon delta 15:52:39 yeah i don't know what the category theory definition is 15:52:48 epsilon delta is topological continuity 15:52:50 me either 15:52:52 there are more too 15:52:54 well 15:53:16 it's just the first few i came up with 15:53:28 okay, it's the characterization that uses the base of the topology instead of arbitrary open sets 15:53:40 Does OpenCourseWare have good math stuff? 15:53:49 what is the algebraic definition? what's the category theoretical one? 15:54:21 the algebraic definition says that a function is continuous if it's an element of ContinuousFunctions 15:54:30 the CT one, i don't care. 15:55:02 haha 15:55:11 well right, that's a different definition :D 15:55:29 "a function is continuous if we call it continuous" 15:55:39 It occurs to me that math _wouldn't_ be a good universal language for talking to aliens 15:55:49 They'd almost certainly use different axioms than we do 15:55:53 http://pl.wikipedia.org/wiki/Funkcja_ci%C4%85g%C5%82a#Przestrze.C5.84_funkcji_ci.C4.85g.C5.82ych 15:55:55 lawl, Firefox mobile beta requires 32MB internal storage :P 15:56:19 you can for example define L_n(X, Y) and just go from there to define C(X, Y) 15:56:25 wait, actually the sequence definition is only equivalent in metric spaces 15:56:30 -!- zzo38 has joined. 15:56:38 also there's another definition which talks about that sup f(x) = f(sup x) or something like that 15:56:50 sup f(X) = f(sup X) 15:57:03 those are characterizations, they aren't "different definitions" 15:57:06 for this you just need an order and that's it 15:57:18 they are different definitions which can be shown to be equivalent 15:57:21 yeah f(closure of G) \subset closure of f(G) 15:57:41 historically they were not known to be equivalent 16:00:42 perhaps 16:02:24 so, let S be a set, and consider a surjection f: S -> 2^S, take the set T = {s \in S | s \not\in f(s) }, and let f(t) = T for some t, then t \in T => t \not\in f(t) = T; and t \not\in T => t \in T, so in fact there can't be such an f 16:02:36 a bit confusing linearized like that 16:04:22 cheater99: it's just i prefer to say "different characterizations" once we know they're the same thing, because when new mathematics happens, there will be multiple actually different definitions for the same intuitive concept, until a consistent naming scheme appears (if it ever does) 16:04:50 but, maybe it should've been clear what you meant 16:05:19 it's just it could be that that topological definition was different from the one for metric spaces, or the epsilon delta one, but because it's not, you really don't have to remember but one definition. 16:06:04 i mean, in theory :P 16:06:55 elliott, toroidal lightspeed bubbles. 16:06:59 ADMIT THE AWESOME 16:07:14 so what about f : 2^S -> S be an injection 16:07:43 i can't say i see it. 16:07:56 Phantom_Hoover: k 16:08:58 -!- Zuu has joined. 16:09:52 erm 16:09:57 i'm fucking retarded xD 16:09:59 i'm fucking retarded xD 16:09:59 i'm fucking retarded xD 16:09:59 i'm fucking retarded xD 16:10:00 i'm fucking retarded xD 16:10:00 i'm fucking retarded xD 16:10:17 kebab time -> 16:10:47 actually Phantom_Hoover's "injection is a reversed surjection" is the answer 16:11:41 elliott, FWIW, the spherical geometry did exactly what oerjan predicted. 16:12:07 Phantom_Hoover: did they listen? 16:12:15 elliott, no, they did not. 16:12:16 What did oerjan predict? And is this spherical geometry of a CA? 16:12:21 And who is "they"? 16:12:22 anyway how can anyone think continuity is confusing, when there's the roughly as important concept of compactness lying around 16:12:25 Phantom_Hoover: what did they say 16:12:31 Sgeo, that you can't make a planar CA spherical. 16:12:36 elliott, they didn't respond. 16:12:46 DKS REPLIED! 16:12:53 "A complete 3620 system weighs about 200 lbs when packed for shipping." 16:12:54 ha ha ha 16:12:56 can't make a planar CA spherical? 16:13:09 in what sense? 16:13:25 Psht, he's trying to sell me the more expensive and Symbolics-keyboard-lacking MacIvory. 16:14:25 well, i don't really care a shit anyway 16:14:31 oklopol, in the sense that you need to have a point at which the symmetry breaks down. 16:14:31 -> 16:14:37 :( 16:14:40 erm 16:15:00 i'm not smart enough for an explanation that vague 16:15:12 when are two CA's equivalent? 16:15:16 conjucagy? 16:17:17 conjugacy is when you take a homeomorphism between the spaces that commutes with the CA functions on both sides 16:17:29 -!- elliott_ has joined. 16:17:33 highly intuitive definition of sameness don't you think 16:17:36 -!- elliott_ has quit (Client Quit). 16:17:43 Now why am I so terrible at talking to salesmen? 16:17:58 do you always end up buying them 16:18:15 erm right the leaving -> 16:21:50 -!- elliott has quit (Ping timeout: 240 seconds). 16:21:52 oklopol, when you get back: if you have Life on a sphere, gliders crash at some points; the whole space isn't isotropic. 16:22:37 I do not know what homeomorphism is, though 16:23:58 It's a morphism that likes other morphisms. 16:24:24 oklopol: no, they're different definitions, because their application doesn't overlap 16:24:34 the topological definition doesn't work on ALL ordered sets 16:24:45 and the ordered sets definition doesn't work in ALL topological spaces 16:24:55 there are sets on which those definitions are equivalent 16:25:12 but that intersection is of measure zero 16:37:14 cheater99: oh cool, what's the ordered set definition 16:37:32 i told you already 16:37:33 sup 16:37:35 of course they are different definitions by any definition of definition if they're not equivalent 16:37:52 erm that always works with the order topology, doesn't it? 16:37:59 if it doesn't, okay, you win 16:38:03 100-0 16:38:32 erm 16:38:54 just sup, not inf? then i'd guess it's continuity w.r.t. half interval topology 16:39:00 * cheater99 looks away from oklopol and toots 16:39:07 :\ 16:39:46 look oklopol 16:40:00 everyone's got their talents 16:40:02 if there seriously is a concept of continuity that isn't subsumed by the topological one (and isn't a ridiculous definition like your algebraic one), then my life will change completely. 16:40:03 >_> 16:40:20 for the worse 16:40:36 you just have to find something you're good at, y'know? 16:40:37 >_> 16:41:56 so what's the sup f(X) = f(sup X) def exactly, do we assume sups always exist, do we assume a chain? 16:42:26 you know what i did 16:42:28 i went to wikipedia 16:42:31 it had all the answers 16:42:37 it's full of stars. 16:42:55 good to know, then maybe you'll be able to answer 16:43:13 we assume sup's always exist. 16:43:28 by saying that we have a total order. 16:43:39 that doesn't guarantee it 16:43:42 lol 16:45:36 sez who 16:45:55 are you serious 16:46:03 have you heard of Q 16:47:44 um 16:47:49 X has to be a directed set 16:47:56 Is there a style of static analysis for C programs which is suitable for literate programming? 16:48:02 obviously 16:48:07 doesn't directed just mean the sup of two elements exists? 16:48:43 Q has that property, 16:48:58 *-, 16:49:50 well seems it just requires *an* upper bound, in any case this is trivially true if you have total order 16:52:28 o rite 16:52:35 i mistransratet 16:52:43 we want a complete partial order 16:52:48 there u go 16:52:58 yeah that works 16:53:07 oh you had the wrong definition for total order 16:53:09 a supremum is not "an" upper bound 16:53:14 it's the minimal upper bound 16:53:22 no, i just used the wrong words 16:53:23 yeah but apparently directed sets don't need there to be a minimal one 16:53:29 yeah, that's what i meant 16:53:40 it's not like i care about this conversation >_> 16:53:45 you had the wrong definition = you associated the word total order to something other than what it usually means 16:53:53 j/k 16:53:55 yeah fuck you too 16:53:55 i care. 16:53:59 OMG 16:54:06 I HATE U!! 16:54:21 * cheater99 runs out of the room crying and slamming the door! 16:54:55 i was j/k but u were like, totally unfriendly, man. 16:55:46 * Sgeo sneezes on chat 16:57:17 cheater99: You misinterpreted him. 16:57:22 cheater99: He was propositioning you. 16:57:25 cheater99: Not trying to insult you. 16:57:44 hmmm... that puts a different twist on the whole conversation. 16:58:45 i was insulting cheater99? 16:58:53 sorry, that was not the intention :D 16:59:08 what did he do to get me to insult him? 16:59:36 hmm maybe the wrong definition thing 17:04:28 what I mean is if it can create the reports in a Enhanced CWEB include file (written with TeX codes) and then can be included into your program as an additional chapter (using the @i command). 17:09:58 " i was j/k but u were like, totally unfriendly, man." i don't think i was 17:11:12 u wuz. 17:11:49 well, sorry again then; does it say on wp that the order thing is a different definition? 17:11:58 that sounds like a very hard thing to prove 17:12:15 that there is no topology in which the continuous functions are the order-continuous functions 17:12:48 i couldn't find it 17:13:49 (not that i looked very hard) 17:17:08 http://www.newscientist.com/blogs/culturelab/2010/11/its-life---only-much-much-bigger.html 17:17:09 WTF 17:17:13 Watch it 17:17:37 From 0:57-1:10 17:17:50 Am I the only one who sees something blatently impossible happening? 17:18:52 Why? 17:19:07 A block disintegrates for no reason 17:19:57 Hmm, yeah. 17:20:02 Glitch? 17:21:15 they even say on the vid that occasionally the sensor reads a ghost's presence 17:22:06 Hmm, good point 17:22:09 Nevermind then 17:22:21 well, they said "random mutation" 17:22:31 DIE SGEO_ DIE 17:22:33 -!- Sgeo_ has joined. 17:23:08 -!- Sgeo_ has quit (Client Quit). 17:23:47 Oh come on, Freenode doesn't show quit messages? 17:24:22 maybe you need to be identified 17:24:37 -NickServ- Sgeo_!~Sgeo@ool-18bf618a.dyn.optonline.net has just authenticated as you (SgeoBot) 17:24:42 hmm. 17:28:00 -!- Wamanuz2 has quit (Ping timeout: 245 seconds). 17:30:06 -!- Wamanuz2 has joined. 17:36:12 -!- Wamanuz2 has quit (Ping timeout: 276 seconds). 17:40:15 -!- elliott has joined. 17:40:28 Vorpal: now watch as I perform the impossible! 17:40:33 X versions be damned. 17:40:56 -!- elliott has left (?). 17:41:11 -!- elliott has joined. 17:41:25 also, I am doing it from Ubuntu Netbook Edition (installed with alternative CD), because I defy logic 17:43:16 -!- Wamanuz has joined. 17:49:51 elliott, when does Fine Structure get a coherent interchapter narrative? 17:50:00 Phantom_Hoover: Soon enough. Bear with it; it's well worth it. 17:50:02 I'm at "The Astronomer's Loss". 17:50:24 Phantom_Hoover: Believe me that the reason it is not connected at first is not due to his incompetence. 17:50:44 I wasn't saying it was; just that it's a bit hard on the brain. 17:50:59 Phantom_Hoover: Yeah, it muddled me a bit too. But then it all starts making sense. :P 17:51:36 It's like the opening of Banks' novels times 10. 17:51:58 -!- Naamah1 has joined. 17:53:52 Naamah1, welcome to the legion on a dam! 17:54:11 Hey, thanks for the welcome! 17:54:58 It could be a silly question, but how can I change my username? : ) 17:55:07 Vorpal: Pingpingping, what is an acceptable date and time format? 17:55:11 Naamah1: /nick foo 17:55:29 elliott, for what? 17:55:35 -!- Naamah1 has changed nick to Adramelech. 17:55:37 Phantom_Hoover: Genera. 17:55:51 Phantom_Hoover: That is, the Symbolics Lisp Operating System. 17:56:02 Phantom_Hoover: It is currently asking me for a date and time. 17:56:17 -!- Adramelech has changed nick to Baal_Zebel. 17:56:24 The best date and time format is objectively YYYY:MM:DD HH:MM:SS 17:56:34 Wait! 17:56:44 YY:YY:MM:DD:HH:MM:SS! 17:56:58 Going to reboot into non-netbook, this is too irritating. 17:57:05 -!- elliott has quit (Remote host closed the connection). 17:57:44 -!- elliott has joined. 17:58:00 * Phantom_Hoover wants to write something TeXy called LoX. 17:58:41 elliott, what's irritating about it? 17:58:46 I was considering installing it today 17:59:19 Sgeo: The dock at the side likes to move the icons around when you click an icon in any way that's even slightly like a drag. For some reason the icons tilt when the dock fills up in a way that I cannot figure out what it is. 17:59:34 Sgeo: Additionally, the bar on the left is so large that screen real estate feels *significantly* reduced. 17:59:47 Sgeo: Also, it is difficult to, say, bring a file system browser on /. There is no facility to do this. 17:59:53 You have to open a USB stick, say, and go from there. 18:00:01 WAIT WHAT? 18:00:02 elliott, alt-f2? 18:00:05 That makes... no sense 18:00:07 Phantom_Hoover: Well, maybe. 18:00:11 Sgeo: The built-in browser looks at ~ only. 18:00:14 (~ = home) 18:00:30 C-l? 18:00:31 Ctrl-L doesn't bring up anything useful in the browser? 18:00:39 Sgeo: I doubt it. I didn't try. 18:00:44 It's very minimal. 18:01:45 "If you find you need more direct access to your file system, click the folder icon to open that folder in Nautilus." 18:01:47 * Sgeo ponders 18:01:53 * Sgeo still wants to try it 18:01:58 Sgeo: That worked in *old* Netbook Edition. 18:02:01 Which is fine. 18:02:05 It manifestly doesn't in 10.10. 18:02:38 Unetbootin should be fine, right? 18:02:44 * Sgeo doesn't exactly have a CD burner 18:04:04 Okay, seems 18:04:08 *that 18:04:21 18:03 Nov 12, 2010 18:04:25 Should do the job. 18:04:27 cheater99: i couldn't prove it so checked on wp, the sup definition is subsumed by the topological one according to the article 18:05:00 so guess i can continue living my life for now 18:07:57 Vorpal: Eurgh. 18:08:00 Vorpal: Have you ever had this time issue> 18:08:04 *issue? 18:08:08 OH WAIT I need to enable it in inetd don't I 18:09:31 also maybe the reason i couldn't prove it was it's not the [a, b) topology :D 18:09:33 21min to download the .iso? 18:09:38 -!- Gregor has set topic: Topic revoked by the Spirit of OER-JAN. | http://tunes.org/~nef/logs/esoteric/?C=M;O=D. 18:09:55 That's incredibly fast 18:10:05 Sgeo: you have a strange definition of fast ... 18:10:11 * Sgeo has a sad Internet connection 18:10:38 * Sgeo vaguely wonders if it's the ISP or the router or the fact that I'm using the wifi 18:10:45 Instead of wired 18:13:17 * Sgeo o.Os at xPUD 18:13:45 * Sgeo is tempted to try it 18:14:20 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds). 18:14:47 It occurs to me that I, for one, don't just restrict myself to web browsing 18:15:48 Dear Chrome: I DID NOT MEAN TO CLICK CANCEL 18:16:08 -!- Sgeo has quit (Read error: Connection reset by peer). 18:16:57 Vorpal: Have you ever had this time issue> <--- ? 18:17:02 -!- Phantom_Hoover has joined. 18:17:26 OH WAIT I need to enable it in inetd don't I <-- well that is why it was mentioned in the guide iirc? 18:18:21 Vorpal: actually no 18:18:25 Vorpal: you install xinetd and never mention it again 18:18:37 elliott, hum, You need to enable daytime and such 18:18:40 i did 18:18:44 daytime, time, echo (why not) 18:19:05 Vorpal: ok, I don't get an XIO error; now I get the big ol' Genera window coming up, but solid white; it POSTs ok, says the log. It traps my keyboard and mouse and all I can do is kill -9 it (kill doesn't work) from a console. 18:19:26 -!- SgeoN1 has joined. 18:19:45 Unebootin bootloader is being very broken 18:19:57 elliott, huh 18:20:02 never ran into that 18:20:05 It counts down the automatic boot... then goes back up to 10 18:20:10 Vorpal: still, no trace of XIO or anything 18:20:26 elliott, well, XIO was in the terminal you ran it from iirc 18:20:30 Vorpal: yep 18:20:33 Vorpal: I checked 18:20:33 elliott, wait, are you using new X? 18:20:41 Vorpal: yes, that's sort of the point 18:20:52 then: no clue, you are on your own. Tell me if you manage though 18:20:54 Vorpal: The XIO error *definitely* no longer happens. Instead, we get an exciting new problem. 18:21:09 Vorpal: HOW-EVER. It *did* manage to bug me for a date and time when I didn't have xinetd. 18:21:17 This then dropped me into a debugger and I could evaluate Lisp expressions. 18:21:21 This was with the fonts and everything. 18:21:24 So it *is* basically working. 18:21:26 heh 18:21:40 elliott, well tell me when you get it actually working 18:21:50 Vorpal: i expect pay-quality support 18:22:07 elliott, XD 18:22:21 Vorpal: (in return you can have remote access to my 3620 when i get it) 18:22:30 Vorpal: (even if DKS is trying to make me buy a macivory instead) 18:22:47 Vorpal: fucking thing doesn't even come with a symbolics keyboard 18:22:47 in other news I implemented a way to prevent my mine carts from running away when I'm exiting them at the stations. 18:23:16 elliott, well, I have no idea what the issue is... 18:25:17 Dear computer: Acknowledge my USB Drive! 18:25:47 Vorpal: Well then. Time to transplant Gutsy packages onto Maverick! 18:26:14 I think this USB drive is broken. Drive feels like the wrong word 18:27:41 elliott, what's so great about the Symbolics keyboards? 18:28:08 Phantom_Hoover: http://www.asl.dsl.pipex.com/symbolics/photos/IO/index.html 18:28:17 Phantom_Hoover: Hyper Super Meta Shift Control. 18:28:26 It had loads of modifiers? 18:28:39 Phantom_Hoover: It was and is THE keyboard. 18:29:20 elliott, OK, more specifically, what about it is superior to the common or garden European keyboard? 18:29:38 Phantom_Hoover: Would you plug a USB keyboard into a Commodore 64? A ZX Spectrum? 18:29:59 Phantom_Hoover: Would you, say, plug, into an Amiga from the glory days, into a 30" LCD? 18:29:59 So it's nostalgia value or what? 18:30:07 Phantom_Hoover: It's *awesome*. 18:30:29 elliott, you're being like Sgeo, but with keyboards rather than virtual worlds. 18:30:51 Phantom_Hoover: No, you just don't realise how awesome Lisp Machines were. 18:31:12 Phantom_Hoover: Even Stanislav, who was trying to get the console hooked up to an LCD and USB mouse (non-trivial), was adamant about keeping the keyboard. 18:31:21 The thing was a legend. 18:31:55 that's a stupid keyboard 18:31:55 elliott, I have actually heard *reasons* for the Lisp machines being awesome. 18:32:13 yeah unlike you elliott, you stupid noob 18:32:14 For instance, they taste like butterflies. 18:32:37 Gregor, well, it's nice, but it's hardly a vital feature. 18:33:28 Unetbootin is simply not installing xPUD for some reason 18:36:05 It hates you. 18:36:34 I think it loves me. It's trying to protect me from xPUD 18:37:40 -!- zzo38 has quit (Remote host closed the connection). 18:37:57 Vorpal: [[Your password file needs to use "unix crypt" style passwords instead of the now-common md5 passwords. On Ubuntu with the default installation I use, this is configured in the file /etc/pam.d/common-password by commenting out the string "md5": 18:38:01 $ grep md5 /etc/pam.d/common-password password required pam_unix.so nullok obscure min=4 max=8 # md5]] 18:38:03 Vorpal: That wasn't in your guide :p 18:38:37 1. The world images that are included in the snap3 and 4 releases won't boot if the system date is after ~2000! y2k thing?? 18:39:14 # the default is Unix crypt. Prior releases used the option "md5". 18:39:16 ah 18:39:23 There is something utterly demented about its defaults for the trackpad 18:39:43 -!- poiuy_qwert has quit (Ping timeout: 245 seconds). 18:40:13 Why is Facebook app fullscreen on me? How do I get back to the menu? 18:40:33 #xpud 18:40:36 I am trapped in Facebook 18:41:31 Vorpal: http://www.cliki.net/Linux%20VLM%20workarounds genera without root 18:42:05 I officially loathe xPUD 18:42:48 Now that we know TUN is avaible, grant full local access to the TUN device: 18:42:52 $ sudo iptables -A INPUT -i tun+ -j ACCEPT 18:43:32 -!- poiuy_qwert has joined. 18:46:14 tx ip: 10.0.0.2 10.0.0.1 18:46:16 rx ip: 10.0.0.1 10.0.0.2 18:46:18 over and over 18:46:20 hmm 18:47:51 SgeoN1 is stuck in Facebook. 18:48:00 What. 18:48:44 xPUD had a list of web apps. I clicked Facebook. It went fullscreen, with no way to get out. I bothered to log in, and it didn't work 18:49:13 * SgeoN1 suddenly wonders if xPUD is an elaborate phishing scheme 18:50:12 I wouldn't be surprised. 18:50:15 What the hell is it? 18:50:39 Phantom_Hoover: a stupid linux-based os built around firefox. 18:50:49 Why... why would anyone do that? 18:50:54 And why would SgeoN1 install it? 18:50:55 bcuz cloud 2.0 web 18:50:58 bcuz sgeo 18:51:35 Well, I'd guessed that was why Sgeo did it. 18:51:53 "why sgeo x?" "bcuz sgeo" 18:52:56 Phantom_Hoover: I am so close to this: http://collison.ie/img/genera.png 18:53:01 Phantom_Hoover: well, the genera bit. 18:54:02 It still boggles me that after well over a decade of effort, noöne has resurrected the Lisp Machines. 18:54:33 Phantom_Hoover: because nobody wants lisp machines apart from the people who had lisp machines, and Symbolics, in their arrogance, destroyed the market and then, well, what can you do? 18:54:37 and then everyone who used them moved on. 18:54:52 I want Lisp machines! 18:54:58 Phantom_Hoover: to Unix. And hated it. In fact the UNIX-HATERS list was founded a few days after a Lisp Machine user was forced to switch to a Sun box. 18:55:14 Phantom_Hoover: You know what Unixes like SunOS were like in the 80s? 18:55:18 Shit, that's what. 18:55:21 Genera to that. AWESOME 18:55:28 I can imagine. 18:56:32 http://conservapedia.com/ConservaMath_Medal 18:56:41 Anything like switching from a MOO to a MUSH? 18:56:53 Phantom_Hoover: Just seeing that title: read more Hughes to cleanse your brain. Ommmm. 18:57:08 'cos the Fields Medal is Liberal lies! 18:57:21 Ben Green apparently earned this award four years ago when it was given for an achievement he helped obtain, but giving the Fields Medal to him now might dim the star of Obama-supporter Terence Tao, making Tao less effective politically. 18:57:23 :D 18:57:25 elliott, I'm up to the Man Who Can Pass Through Things. 18:57:45 Phantom_Hoover: Wow, what a chapter naming opportunity Sam missed there. 18:57:50 THE MAN WHO CAN PASS THROUGH THINGS 18:58:04 Phantom_Hoover: "Indistinguishable from magic", right? 18:58:28 Yes 18:59:33 Seth Gold was an ordinary basement dweller, until he was torn about by ActiveWorlds' destruction, when he became.... THE MAN WHO CAN PASS THROUGH THINGS 18:59:44 HE CAN PASS THROUGH 18:59:47 PURE 18:59:49 NOSTALGIA 18:59:55 THRILL as he falls through his chair! 19:00:05 CRY as he is unable to type! 19:00:29 CHEER as he passes through the basement ceiling! 19:02:01 I need to ditch this laptop for a real computer. 19:04:33 Phantom_Hoover: Tell me. Why are AMD processors terrible? 19:07:10 My dad was uneasy about buying this laptop because of AMD 19:07:45 SgeoN1: I will never say a bad word about you ever again if you stop listening to every damn thing your dad says even when he clearly has absolutely no idea what he's talking about. Or hell, I'll compromise, just don't mention it. 19:09:44 Phantom_Hoover: NOW WATCH ME FLAIL AS I TRY AND FIND AN AMD PROCESSOR THAT (1) HAS DECENT PERFORMANCE AND (2) HAS LOW ENOUGH THERMAL SPECS THAT I CAN GET IT IN A FANLESS SYSTEM 19:10:36 my dad says you shouldn't move the mouse when the computer is starting because... well i don't really know why but you shouldn't. 19:10:44 he was a programmer for like 15 years 19:10:47 oklopol: wise words, wise words 19:10:57 oklopol: it could disturb the magnetics 19:11:00 :D 19:11:22 or when pretty much anything is happening, like you're installing something, you can't touch the computer 19:11:27 the installation might fuck up 19:11:47 oklopol: well if it's processing your mouse movements and rendering to the screen it can't write to the disk at the same time 19:11:56 so it might miss out bits, with obvious consequences 19:12:01 :D 19:12:04 yeah 19:12:33 elliott, I don't know why AMD processors suck. 19:12:58 Perhaps it's because they didn't sell their souls to the devil so he would suck the heat out of the processors. 19:13:17 Phantom_Hoover: They're a lot hotter and a lot more power consuming than *faster* Intel processors. 19:14:29 Yes, that does suggest that Satan is being used as Intel's cooling system. 19:15:34 Indeed. 19:16:31 or when pretty much anything is happening, like you're installing something, you can't touch the computer <-- that makes no sense 19:16:53 why? 19:17:11 oklopol, moving the mouse around should be harmless 19:17:24 really?? i'll tell him as soon as i can 19:17:33 at least on sane systems 19:17:51 of course, on classic mac os, if you downloaded something then moving your mouse would slow down the download 19:18:04 that was on a performa 19:18:08 with 28 kbit modem 19:18:40 but that hardly applies to a modern computer 19:18:49 elliott, Fine Structure blind guess: are Anne Poole and the Man Who Can Pass Through Things both due to superposition of stuff? 19:20:28 -!- kar8nga has joined. 19:21:15 it surprises me greatly that conservapedia has an article on infinity, i was pretty sure they were all finitists 19:21:20 The fact that the Man Who Can Pass Through Things can See Through Things as well. 19:21:33 *strongly suggests hyperspacy stuff. 19:21:44 1/0 NO! This isn't allowed! Infinity is not a number, and division by zero is illegal! 19:22:06 Well, it is in non-standard analysis. 19:22:35 it's a number if you call it a number, by the mathemtical definition of being something 19:22:56 Phantom_Hoover: elliott, Fine Structure blind guess: are Anne Poole and the Man Who Can Pass Through Things both due to superposition of stuff? 19:23:00 Not that I know of. 19:23:04 Phantom_Hoover: But you should pay more attention to names. 19:23:33 Phantom_Hoover: Hint: Crushed Underground. 19:23:33 infinity is a number on the extended real line, i'm not sure nonstandard analysis really has a number called infinity, although the hyperreals have many numbers that are bigger than any real 19:23:43 Phantom_Hoover: Just click it, and look for a name. 19:23:44 "The Man Who Can Pass Through Things" is far too cool a name to bother correcting. 19:24:06 The Man Who Can Pass Gas Through Things 19:24:15 oklopol: it surprises me greatly that conservapedia has an article on infinity, i was pretty sure they were all finitists 19:24:19 god is infinity duh 19:24:26 Ah. 19:24:51 "In measure theory, one sometimes defines the "extended reals", allowing plus and minus infinity to be considered to be numbers, but this is a special construction, which makes certain arithmetical operations impossible. It can't be done in general" 19:24:55 "it can't be done in general" 19:24:57 :D 19:25:20 Phantom_Hoover: elliott, Fine Structure blind guess: are Anne Poole and the Man Who Can Pass Through Things both due to superposition of stuff? <-- the latter seems a rather awkward name if you have to use it more than once or twice on a single page :P 19:25:24 this may be funnier than finitism 19:25:33 (book page size block if not in such a format) 19:25:34 http://www.conservapedia.com/Special:Random <-- how to kill yourself 19:25:55 SgeoN1 being the Man Who Can See Through Things as well would be too icky to consider. 19:25:57 oklopol: finitism is boring cuz they usually accept the naturals imo 19:26:13 oklopol: ultrafinitism is great because you can basically deduce that there's only one of anything, if you apply ultrafinitism to ultrafinitism 19:26:37 "The Conservative Bible Project is a project utilizing the "best of the public" to render God's word into modern English without liberal translation distortions." HOW WAS I UNAWARE OF THIS X-D 19:26:48 Gregor: Old :P 19:26:54 Gregor, very old 19:26:56 Gregor: They get rid of comrade because it's communist 19:26:57 I didn't claim it wasn't. 19:26:57 It was on Colbert, too. 19:27:04 erm wtf 19:27:10 they believe in diagonalization too 19:27:20 oklopol: but not relativity 19:27:24 really? :D 19:27:28 oklopol: yes, they link it to moral relativity 19:27:37 oklopol: seriously 19:27:40 oklopol: http://www.conservapedia.com/Counterexamples_to_Relativity 19:27:43 elliott, they said that Jesus telling his disciples to cast their nets to the right of the boat is an endorsement of conservatism. 19:27:56 oklopol: [[# In Genesis 1:6-8, we are told that one of God's first creations was a firmament in the heavens. This likely refers to the creation of the luminiferous aether.]] 19:28:02 oklopol, those counterexamples include Jesus doing action at a distance. 19:28:05 oklopol: that's gotta be a troll :) 19:28:10 :D 19:28:11 xD 19:28:17 oklopol: "Liberal pseudoscience: Black holes • Dark matter • Moral relativism • Wormholes" 19:28:44 Since the writers of the New Testament were clearly carefully checking to see if Jesus' miracles obeyed causality. 19:29:01 Phantom_Hoover: of course, god is very careful 19:29:03 Luminiferous æther. Seriously. That's astoundingly retarded. 19:29:12 pikhq: It's 101% certain a troll, that one. 19:29:20 * Phantom_Hoover checks. 19:29:45 pikhq: Why isn't there a six-core processor with a TDP of around 100W, I ask you. 19:29:46 Dear CMOS: Please try to retain settings that I set. 19:29:46 oklopol: ultrafinitism is great because you can basically deduce that there's only one of anything, if you apply ultrafinitism to ultrafinitism <-- hm.... so lets see... that means there are only one "thing". That must be the universe. Which must be an atomic entity (in the sense "can't be divided"). Since if we could divide it, there would be multiple things. 19:29:49 Whyyy 19:29:51 (x86-64 :P) 19:29:55 is only* 19:30:04 Vorpal: You must be SO HIGH right now. 19:30:14 "Duude... what if the universe is only one thing and it can't be divided..." 19:30:15 "7.The universe shortly after its creation, when quantum effects dominated and contradicted Relativity. " 19:30:15 elliott, ;P 19:30:16 :D 19:30:34 [10] 19:30:35 elliott, well, I didn't sleep well last night 19:30:36 oklopol: yeah the quantums became conscious and talked back whenever relativity told them to do shit 19:30:39 that probably explains it 19:30:45 Vorpal: I *didn't* sleep last night. 19:30:52 Vorpal: 24.5 hours I've been awake now. 19:30:56 Perhaps we should forbid the mentioning of quantum and relativistic mechanics to people too stupid to comprehend them. 19:31:08 Vorpal: I've gone past the "falling asleep in my chair" stage to the "lucid once more" stage. 19:31:11 (I think that's how high my CMOS is, which just decided that my USB drive is an fdd 19:31:11 we should try 19:31:17 saying something retarded on there 19:31:22 and adding "because god" 19:31:30 oklopol: like "god made the universe in 7 actual days" 19:31:32 lol 19:31:34 nobody could believe that 19:31:40 elliott, anyway it can obviously only have one property, which would be it's existence. Which means it that is all we can describe it by. 19:31:54 Vorpal: No, man, the property would be another thing. 19:31:57 Vorpal: Its only property is ITSELF. 19:31:58 elliott, oh right 19:31:59 Vorpal: WHOA. 19:32:04 elliott, :D 19:32:14 elliott: Some people actually believe that God made the universe in 7 actual days. 19:32:22 pikhq: That was sort of entirely what my joke was based on. 19:32:35 elliott: The same morons can often be found believing that the King James Bible is the only valid Bible. 19:32:53 elliott, you haven't seen Schlafly trying to justify the fudge to Newtonian gravity. 19:33:02 Phantom_Hoover: JUSTIFY FUDGE TO ME 19:33:13 pikhq: The Christian Bible needs a fucking name, I'm sick of it being called "The Bible". 19:33:21 pikhq: Torah. Qur'an. ??? 19:33:32 Erm. 19:33:36 Phantom_Hoover, what fudge? 19:33:36 *Tanakh 19:33:39 if there isn't an article about perfect spaces 19:33:47 Vorpal: THE fudge. The single object that exists. 19:33:50 Vorpal: The fudge's property is the fudge. 19:33:54 elliott, you can't name it 19:33:55 maybe we could add one saying only god can create perfect spaces 19:33:56 The fudge is made out of a single the fudge. 19:33:59 because the name is another thing 19:33:59 Vorpal: It IS its name. 19:34:02 elliott, ah 19:34:08 Vorpal: The whole universe, when taken together, is the name "the fudge". 19:34:11 Ok. Unetbootin isn't working like it should 19:34:24 elliott, but that consists of separate letters 19:34:24 because man is not perfect 19:34:27 elliott, brb phone 19:34:40 man will always leave a couple points in every neighborhood 19:35:00 Vorpal: no it's all joined together 19:35:03 Vorpal: and fractal 19:35:11 oklopol: :D 19:35:25 elliott: Biblia sacra. 19:35:31 elliott, the fudge is replacing r^2 with r^2.anindecentnumberof0s1 in the gravity formula. 19:35:51 Phantom_Hoover: lawl 19:35:59 pikhq: Hey, I might have found an acceptable AMD processor. 19:36:04 Oh? 19:36:18 pikhq: Phenom II X4 925. 2.8 GHz quad core, 4x512KB L2, 6MB L3, and 95W TDP. 19:36:30 Sure, that's still too high... but it's certainly decent. 19:36:36 elliott, Schlafly doesn't seem to have realised that that value would imply that space was curved *anyway*. 19:37:02 pikhq: Disadvantage: It's $129.99, which is not far away from the $179.99 prices of X6 monstrosities with the newer core architectures. 19:37:52 elliott, what do you actually need your megaprocessor for? 19:38:03 Phantom_Hoover: Umm, I actually just want... a processor. 19:38:07 Phantom_Hoover: Four cores isn't much to ask for these days. 19:38:17 Phantom_Hoover: See, I want a decent computer. So I want to assemble one. 19:38:20 So I want components. 19:38:23 elliott, well, get two. 19:38:27 FUCK YEAH "Christians qualify this statement further by stating that God is infinite, as are all things associated with Him (His Might, His Forgiveness, etc), but our finite minds cannot comprehend them. For this further reason we should not include infinite sets into our mathematical universe. " 19:38:30 FUCK YEAH :D 19:38:33 You'll have EIGHT cores! 19:38:37 Phantom_Hoover: Which requires a dual-socket motherboard. And also uses more heat. 19:38:49 Phantom_Hoover: I'm a freak, so I don't want a single moving part in my computer other than the bulk storage hard disk. 19:38:51 elliott, DAMN YOU THERMODYNAMICS 19:39:03 Phantom_Hoover: Meaning: I have thermal limits I need to adhere to. 19:39:09 http://conservapedia.com/Axiom_of_Infinity 19:39:10 And Intel is evil because? 19:40:12 -!- jcp has quit (Ping timeout: 276 seconds). 19:40:18 Phantom_Hoover: Not offering virtualisation in lower-end processors just to make people buy more, illegal demand to a laptop (IIRC) manufacturer that they cannot use AMD processors in more than N models otherwise Intel would break their deal (they got fined SHITLOADS for this), Trusted Computing bullshit, and most recently the "You buy a processor, but we've locked up some of the features. You can pay us online, and we'll software-unlock them. Fuck 19:40:20 nd have a nice day." 19:40:48 "Not offering virtualisation in lower-end processors just to make people buy more" 19:40:51 Right. 19:40:51 that is 19:40:55 there is no technical reason or cost to disable VT-x in the processors 19:40:58 and AMD offers it on all models 19:41:05 so it's a scummy marketing technique 19:42:08 Phantom_Hoover: So the only problem is that AMD processors suck. :) 19:42:13 I know! I'll buy a TRANSMETA PROCESSOR. 19:42:26 So AMD sucks and Intel is evil? 19:42:57 Phantom_Hoover: Pretty much. Well, AMD don't suck too much. 19:43:21 Phantom_Hoover: If you don't mind a dozen or so extra watts on your power bill, and you don't care how loud your computer is, you can run AMD just fine. 19:43:28 It's freaks like me that have issues. 19:43:50 elliott, move to Siberia? 19:44:13 Phantom_Hoover: IIRC someone in Finland did use an external radiator... 19:44:22 Phantom_Hoover: But hooking a heatpipe system up to a radiator is not fun. :) 19:44:29 There's encryption in the processor for those features? 19:44:45 Phantom_Hoover: Much easier if you can just poke holes in your case, and put a GIGANTIC FUCKING HEATSINK on your CPU. 19:44:49 To gain access to those features, I mean? 19:44:52 Phantom_Hoover: The other option is this: http://www.endpcnoise.com/cgi-bin/e/std/sku=zalman_fanless.html 19:44:55 -!- kar8nga has quit (Remote host closed the connection). 19:44:57 Phantom_Hoover: Which I'd just buy, except it's Intel-only. 19:45:16 SgeoN1: I think it's just a bit-flip, but I'm not sure. 19:45:17 [htkallas@triton ~]$ salloc -p short grep -c ^processor /proc/cpuinfo 2>/dev/null 19:45:17 12 19:45:21 You need at least that many cores. 19:45:28 Preferrably a strictly greater number. 19:45:36 fizzie: Donate all your computers and I TOTALLY WILL 19:45:42 So then why doesn't someone distribute gpld software to do t? 19:45:49 fizzie: Hmm, I could just use a qemu emulating a 13-core machine as my main computer. 19:46:04 SgeoN1: Obnoxious lawsuit. 19:46:04 SgeoN1: Because you'd have to reverse-engineer it? Also, haha, GPL, more likely some Windows shitsoftware written by a script kiddie. 19:46:10 And what pikhq said. 19:46:15 Incidentally, have I ever expressed my disdain for Linux's support for my graphics hardware? 19:46:26 Phantom_Hoover: It's sort of not Linux's fault, but yes, everyone has that problem. 19:46:30 Except for Intel card users. 19:46:36 And we have the problem of underpowered GPUs. 19:46:42 I think you said you use Intel. Which just means that you're on crack. 19:47:02 I have some Intel onboard laptop rubbish. 19:47:08 Well, that's unfair. 19:47:26 It's what I have and WFM. 19:47:43 It actually ran *perfectly* until Ubuntu 10.04 IIRC. 19:47:43 Phantom_Hoover: Intel should just work. 19:48:03 As the official driver for it is part of X... 19:48:07 *Now*, the GPU actually hangs upon a number of things. 19:48:16 Requiring a complete reboot. 19:48:38 So either some solder melted or something is badly wrong. 19:49:15 Phantom_Hoover: http://www.silentpcreview.com/forums/viewtopic.php?f=14&t=58457 19:49:32 Phantom_Hoover: Gets awesomecrazy a few posts in. 19:50:36 Hah. Universal USB Installer uses a temporary copy of 7zip 19:50:52 pikhq: Have you ever noticed how awesome mice are? 19:51:07 -!- Baal_Zebel has left (?). 19:51:38 [[00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)]] — lspci 19:51:51 [[00:02.1 Display controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)]] 19:52:07 So it's evil to its core, but it's certainly Intel. 19:52:14 Phantom_Hoover: Oh, and here's one of my favourite cases: http://www.silentpcreview.com/files/images/epcn-tnn500/epcn1.jpg 19:52:34 Large surface area? 19:52:37 Phantom_Hoover: That thing -- case and cooling system -- costs about $1,000. 19:52:44 Phantom_Hoover: It is UNBELIEVABLY heavy. 19:52:56 It weighs more than the sun. 19:53:03 Phantom_Hoover: It is a case. Inside, you have small heatsinks and the whole thing is wired up with copper pipe like a watercooled system except... solid copper. 19:53:09 And then, it's attached to the outer shell of the case. 19:53:15 Which is just REALLY THICK, REALLY HEAVY METAL. 19:53:17 elliott, stick a Stirling engine somewhere. 19:53:19 Oh look. Now that I do what Canonical recommended, and told Unetbootin to fuck off, things work 19:53:20 Or a thermocouple. 19:53:26 Not only is it an iron maiden, it just dissipates the heat into HELL. 19:53:38 (see wut i did thar heavy meatl iron ma) 19:53:44 Phantom_Hoover: "# Weighs 26Kg." 19:53:50 Madness. 19:53:53 I weigh less than double that! 19:53:55 Although right now it looks like it's stuck 19:54:00 It better not be stuck 19:54:12 Phantom_Hoover: http://www.silentpcreview.com/files/images/epcn-tnn500/epcn9.jpg TECHNOLOGY 19:54:16 Phantom_Hoover: it was discontinued in 2004 :( 19:55:24 I wonder if bsmntbombdood still uses that insane rig I specced out for him. 19:55:31 Still can't believe it cost less than $1,600. 19:55:36 elliott, how about: you have a radiator array somewhere out-of-the way; you run the pipes to that, and use a Stirling engine to pump it. 19:55:40 It looks like it's stuck on NET: Registered protocol family 1 19:56:00 Phantom_Hoover: Pumping of any sort is out. Movement of any sort = sound. 19:56:08 Why is nothing happening grrrrb 19:56:11 you don't need a cooler if you use a GoL processor 19:56:14 elliott, hence "inconspicuous". 19:56:15 Phantom_Hoover: The only sound I'll accept is a one terabyte bulk storage drive, and that grudgingly :P 19:56:19 oklopol: tru dat 19:56:40 Phantom_Hoover: Hell, you can get fans that are literally inaudible in any sane (non-anechoic chamber) environment. 19:56:44 The Stirling engine is at the radiator. 19:56:44 well, you do but you can just use gliders to send heat away from the action 19:56:46 Phantom_Hoover: But I'm a crazy bastard, you see. 19:56:46 maybe 19:56:49 Phantom_Hoover: hmm 19:56:53 Phantom_Hoover: well it could work 19:57:00 Phantom_Hoover: find me a big ass-radiator that'd work without too much modding :P 19:57:00 information = heat right 19:57:06 And it runs from the heat gradient across the radiator. 19:57:12 Phantom_Hoover: also surely a regular pump would work 19:57:16 oklopol: totally. 19:57:24 oklopol: that's why the hotter you get, the smarter you are 19:57:34 hohooo 19:57:35 elliott, sure, but it's less cool. 19:57:37 http://www.newegg.com/Product/Product.aspx?Item=N82E16820231392 anyone want 48 gigs of ram? 19:57:40 includes four ram fans 19:57:44 Also inelegant. 19:58:34 http://www.newegg.com/Product/Product.aspx?Item=N82E16820226096 Or, much more reasonably, 12 gigs of ram :P 19:58:35 Indeed, it would be *best* to have a huge amount of heat with the Stirling system. 19:58:43 This isn't working. Why isn't it working? 19:58:58 SgeoN1, what are you doing this time? 19:59:22 Attempting to get Ubuntu 10.10 Netbook Edition to start off USB 19:59:24 Moving the water takes constant power, which comes entirely from the heat. 20:00:00 * Phantom_Hoover ponders. 20:00:29 How much force would you expect to need to run a radiator in such a system? 20:02:50 Why wont the bloody thing work? 20:02:57 I don't know. 20:03:05 Maybe there's some CMOS setting that I need? 20:03:11 Perhaps if you actually described the problem that would help. 20:03:20 Have you got the boot order right? 20:04:38 The boot order wont stick, but Ive been using the boot menu 20:04:49 It keeps getting stuck at that NET thing 20:04:59 That... NET... thing. 20:05:08 Which I quoted above 20:07:25 Oh fun 20:07:57 The CMOS deletes the USB entry from the boot order when there's a boot up without a USB stico 20:09:53 http://i.imgur.com/TEVSt.jpg this is where it's stuck 20:11:02 SgeoN1: Toshiba T150? 20:11:07 Looks very much like my machine. 20:11:19 T215 20:11:29 *T130 20:11:43 SgeoN1: Cool, the retarded brother of my laptop. 20:13:25 For the record, protocol family 1 is Unix domain sockets, and necessarily doesn't have all that much to do with whatever it gets stuck with. 20:14:08 So then, why is it getting stuck? 20:14:29 SgeoN1: I don't know I'll use my psychic powers 20:14:33 OMMMMMMMMMMMMM 20:14:37 It occurs to me that I may have downloaded the 32 bit version 20:14:54 Irrelevant. 20:14:59 It would still boot just fine. 20:15:38 It probably won't help, but my Ubuntu boot-time dmesg looks quite a lot like that, and the immediately following lines are: 20:15:40 [ 0.404185] NET: Registered protocol family 1 20:15:41 [ 0.482223] Freeing initrd memory: 9188k freed 20:15:41 [ 1.430015] pci 0000:00:12.1: OHCI: BIOS handoff failed (BIOS bug?) 00000184 20:15:41 [ 1.521275] pci 0000:01:05.0: Boot video device 20:15:46 SgeoN1, what do you have installed right now. 20:16:00 Phantom_Hoover: WINDOZ probably 20:16:18 Windows 7 20:16:21 SgeoN1, have you actually tried AW on Wine? 20:16:34 Or better yet NOT tried it? 20:16:34 Tried to use the noapic/nolapic/acpi=off boot parameters with that Ubuntu? 20:16:50 elliott, incidentally, WP had, in its comparison of TeX editors, "Compatibility with Windows 7 math panel". 20:17:00 Phantom_Hoover: heh 20:17:00 (Those should be there in the boot-time menu.) 20:17:02 I was planning on first trying this, and if I liked it, dual-boot 20:17:02 I've tried it in the past 20:17:04 Exactly one of the editors had anything other than "No" in that column. 20:17:23 I say "had" in the past tense because I deleted it. 20:17:39 Phantom_Hoover: Information elitist. 20:18:29 Gyaah, wait. 20:18:35 Fuzzier, hadn't tried it yet 20:18:44 WP's table format doesn't allow you to get rid of columns easily. 20:19:49 elliott, opinion on paternoster lifts? 20:20:14 What's the difference between them? 20:20:17 elliott: Opinion on Valery Nikolayevsky? 20:20:27 fizzie, I have a reason for this seemingly random question! 20:20:31 fizzie: Opinion on global warming? 20:20:37 Vorpal: You did it with Minecarts. 20:20:47 elliott, no. with boats. Or I'm going to do so 20:20:49 elliott, http://www.minecraftforum.net/viewtopic.php?f=35&t=22428 20:20:50 see that 20:20:53 very cool 20:20:58 Fizzie? 20:20:59 Vorpal: Well, mine was a Wikipedia "Random article", unsurprisingly. 20:21:14 fizzie, hah 20:22:21 fizzie: If you don't spoon-feed SgeoN1 some answers now, he might have to think and where will we be then? 20:22:22 SgeoN1: Uh, they do different things? noapic/nolapic disable some newer sort of programmable interrupt controllers, and acpi=off may help on systems where the ACPI bios stuff is somehow strangey-weirdo. 20:22:35 I just tried all of them 20:22:46 Sure, that's reasonable. 20:23:02 SgeoN1, apic != acpi 20:23:08 same letters yes 20:23:28 ...believe it or not, I realized that 20:23:39 Just had no clue what any of them meant 20:23:47 hm I guess another way to phrase that would be that words are _ordered_ sets of letters. 20:23:47 Vorpal: If only they would've called it Advanced Power and Configuration Interface, instead of Advanced Configuration and Power Interface; then we'd have identical acronyms there. 20:23:58 fizzie, :D 20:24:30 Erm, or actually APCI is still not APIC. Oh well. 20:24:39 hah 20:24:55 Maybe call the latter Advanced Programmable Controller of Interrupts instead of ... Interrupt Controller. 20:24:56 oklopol: very good 20:25:34 Vorpal: aka lists 20:25:41 words can have duplicates 20:25:43 fizzie, they could have called PCIe APCI too. For Advanced PCI 20:25:47 ordered multiset i.e. list :P 20:25:53 elliott, ah indeed 20:25:58 I thought 10.10 Netbook Edition was supposed to use Chromium by default 20:26:30 elliott, I think "ordered multisets" sounds better than "lists" though. Mainly because it takes slightly longer to parse. 20:26:35 SgeoN1: You thought wrong. 20:27:03 SgeoN1: Did you mention what actually happened with those options? 20:27:35 It worked 20:27:45 Oh, how strange. 20:27:48 I don't know which did it though 20:28:03 My guess would be acpi=off, but that's just from personal "thing that breaks often" experiences. 20:28:04 cheater99: what is, the thing i told you last or the thing i said last? 20:28:13 actually i guess i haven't said anything for a while so 20:28:13 so does anyone here know cobol? 20:28:18 SgeoN1 20:28:31 oklopol: -- You are sitting in a chair. -- 20:28:31 knows pretty much everything about it 20:28:50 My guess would be acpi=off, but that's just from personal "thing that breaks often" experiences. <-- I just hope that is a desktop and not a laptop 20:28:53 Well, #intel-gfx is apparently too dead to help me. 20:28:54 -!- SgeoN1 has quit (Remote host closed the connection). 20:29:07 oklopol: i was replying to what you said to me last 20:29:21 cheater99: yeah realized 20:29:29 omg 20:29:33 we can almost like 20:29:36 ;) 20:29:38 read each other's thoughts. 20:29:41 finish each other's... 20:29:43 darn 20:29:51 you ruined it :D 20:30:05 -!- jcp has joined. 20:30:51 Vorpal: It's a laptop, which might make power-management and/or suspendy stuff quite a tricky thing. 20:31:35 i was busy surfing the net 20:31:47 i'm sorry sweetheart :( 20:31:58 fizzie, indeed it would 20:32:00 oklopol: so does sgeo really know that much about cobol? 20:32:03 no prob 20:32:06 cheater99: well umm 20:32:15 here's the thing 20:32:17 i didn't really read the conversation from which i deduced this 20:32:20 i have a huge codebase 20:32:24 in COBOL 20:32:32 i need to have it translated to php 20:32:40 and i think it was more like sgeo said about cobol that wasn't completely retarded and everyone hated on him 20:32:55 there's already one team working on it, but their approach is crap 20:33:53 so i'm sort of looking for ideas 20:34:02 if you have some sort of real problem, i'm not the guy to talk to 20:34:10 well 20:34:26 if you have an imaginary problem then maybe 20:34:32 the problem is unusual enough that it's highly unlikely someone will actually have experience doing that exactly 20:34:36 ok 20:34:40 imagine i have a cobol codebase 20:34:46 and i need to translate it to php 20:34:48 hmm umm well i can give you all the insight i have: 20:34:50 WHAT DO 20:34:50 your problem 20:34:54 sounds really trivial 20:34:58 get some monkey to do it 20:35:06 cheater99, cobol... php. Both are very wtf. 20:35:11 i didn't think cobol-> could ever be a bad thing 20:35:13 Vorpal: don't ask why 20:35:15 but putting php on the end of it 20:35:17 yup 20:35:19 you succeeded 20:35:20 :D 20:35:27 cheater99: #unesoteric 20:35:31 elliott, :D 20:35:40 they can help with all practical questions. especially ones about COBOL and PHP. 20:35:40 elliott: #angst 20:35:54 Yes... those messages were so angsty... 20:36:09 I'm cutting myself just to let the ANGST out of my body. 20:36:21 It can't just be that you never talk about anything interesting or even vaguely on the esoteric side of things, nope. 20:36:24 + Now talking on #unesoteric 20:36:25 + Phantom_Hoover (~phantomho@cpc3-sgyl21-0-0-cust116.sgyl.cable.virginmedia.com) has joined #unesoteric 20:36:25 + oklopol (~oklopol@xdsl-83-150-123-242.nebulazone.fi) has joined #unesoteric 20:36:28 elliott, ^ 20:36:35 anyways.. 20:36:35 Vorpal: see, a thriving community alraedy 20:36:37 there's some dime involved 20:36:39 elliott, :D 20:36:50 so if anyone's got good ideas, i'm all ears. 20:37:15 i don't need money i have a joba 20:37:17 *job 20:37:30 nobody here does COBOL. nobody here does PHP. nobody wants to. nobody wants to help whatever shitty conversion that is. especially if you won't even justify it. 20:37:42 i think i read a book about cobol once? 20:37:44 is *job just like a virtual idea of where your actual job would be? 20:37:55 no i actually do have a job 20:38:05 i thought it was a job pointer 20:38:06 the joke is i'm not actually getting payed that much 20:38:24 elliott: #AANNGGSSTT 20:38:42 oklopol: ya :-\ 20:38:43 Vorpal: can I borrow a dictionary? I think cheater99 needs some definitions clarified for him. e.g. "angst" 20:38:59 angst (uncountable) 20:39:03 1. A feeling of acute but vague anxiety or apprehension often accompanied by depression, especially philosophical anxiety. 20:39:05 2. More commonly, painful sadness or emotional turmoil, as teen angst. 20:39:06 -!- SgeoN1 has joined. 20:39:22 elliott, sure you can. Cambridge unabridged 20:39:23 elliott: how is what you're doing not apprehension? 20:39:23 i guess he could theoretically think that he manages to induce emotional turmoil in me rather than chronic irritatedness. 20:39:29 haha elliott has an uncountable amount of angst 20:39:44 oklopol: he has 2^angst 20:39:52 cheater99: "homophobia (uncountable) 20:39:56 1. (obsolete, individual occurrences) A pathological fear of mankind." 20:39:56 get it, his angst has a subset that has strictly less cardinality than his angst that can still be put in bijection with its proper subset 20:39:58 cheater99: ergo homophobia means "fear" 20:40:04 i know it's true, because one of the words was in there 20:40:08 oklopol: SO TRUE 20:40:35 COBOL should be easy to mechanically translate from 20:40:37 cheater99, #unesoteric has competent COBOL and PHP experts, FWIW. 20:40:42 But it wont produce readable code 20:40:44 SgeoN1, don't get involved. 20:40:53 elliott: are you done having your attack? 20:40:55 elliott and I will never forget. 20:41:07 SgeoN1: oh? i would've thought it's especially hard to parse 20:41:09 cheater99: sorry i will stop this libel and slander 20:41:17 elliott: thx 20:41:23 cheater99: i apologise for any losses you may have incurred as a result of my hideous character deformation 20:41:29 elliott, don't feed the troll 20:41:37 I can't even begin to imagine under what circumstances cobol->PHP is even meaningful though 20:41:42 cheater99: my only condition is that you administrate #unesoteric for the good of COBOLkind. 20:41:50 Vorpal: everyone else does, so obviously it must be fun 20:41:51 SgeoN1: someone's paying, that's meaningful enough :P 20:41:57 elliott, what about PHP->COBOL? 20:42:06 * elliott considers paying cheater99 to rape babies all day long 20:42:20 elliott: i thought you were done now? 20:42:27 cheater99: you didn't agree to my condition 20:42:31 elliott: what are you, 14? 20:42:35 12 20:42:39 and a half 20:42:43 and three quarters of a half 20:42:49 Seriously, do COBOL -> Python or Java or Ruby or somethinf 20:42:57 sorry *8 20:43:01 8 and a half and three quarters of a half 20:43:09 elliott, exactly? 20:43:22 Vorpal: yes. i have a time machine set up to ensure this age stays constant 20:43:27 like peter pan 20:43:29 elliott, you need to divide it down to plank time :P 20:43:34 COBOL should not be involved with web sites. PHP should not be involved with... well, anything really, but especially not nonwensites 20:43:39 Vorpal: is plank time the time it takes for a plank? 20:43:40 SgeoN1: i would love to be doing python, but unfortunately the project's been doing php for years now :p 20:43:50 elliott: you go forward in time all the time you mean? 20:43:59 oklopol: yes 20:44:00 elliott, err did I typo? 20:44:01 SgeoN1: in this case, cobol was involved in websites, and now php is involved in doing websites, and non-websites 20:44:03 oklopol: at a rate of 1 second per second! 20:44:07 Vorpal: planck 20:44:08 SgeoN1: which is fairly crazy 20:44:10 elliott, oh right 20:44:27 SgeoN1: how would you do this mechanical translation, anyways? 20:44:46 cheater99: kittens and rainbows 20:44:48 cobol doesn't have much in the way of an AST that can be meaningfully used, since it has lots of goto's 20:44:54 you put them on a treadmill and put a furnace behind the treadmill 20:44:56 elliott, actually plank time is the same. But you can check how many rings they have to find out... uh... how many rings they have! 20:45:00 and lots of weird jumps and entry points 20:45:01 and make the treadmill go super-fast if they don't type out PHP code 20:45:05 problem solved 20:45:06 Well, if you want readable PHP, do translation by hand, then spend a few months getting rid of the globalism and Vito's and alters and everything else that makes COBOL so bad 20:45:09 that doesn't really translate to any sane language being used nowadays 20:45:11 the rainbows are just there for decoration 20:45:20 SgeoN1: it doesn't have to be readable.. it just has to do the same thing, really 20:45:35 elliott, Note: do not confuse plank time and tree time. Tree time is when you drink sap. 20:45:46 at tree o'clock 20:45:57 Globalism? Vito's? Fuck you autocorrect 20:46:06 * Gregor stabs Vorpal with a ... wrench. 20:46:15 Vorpal: Plank time is when your plank gets hard and then sap comes out of it. 20:46:22 elliott, XD 20:46:24 yeah, what's the "vitos" supposed to be? :o 20:46:26 Gregor: Worst stabbing instrument ever :P 20:46:39 elliott: Yeah, you say that until there's a wrench stickin' out yer kidney. 20:46:46 SgeoN1: ^ 20:46:49 How difficult can it be to mechanically translate gotos to something... that exists in PHP? 20:46:54 * elliott stabs Vorpal with a cardboard box 20:46:58 STAB STAB STAB 20:46:59 Gregor, why did I just imagine a google results page saying "Did you mean: stab with a wench" 20:47:00 STABBY STAB STAB 20:47:02 Gotos, I think 20:47:03 SgeoN1: php added goto recently 20:47:11 SgeoN1: php has goto's, but it's more about the zillions of entry points, and global references 20:47:28 Is there a way to mechanically get rid of alter? I remember seeing it once 20:47:33 SgeoN1: also, php using goto's isn't really recommended 20:47:42 php isn't really recommended. 20:47:48 I thought you said it has to work, not be readable 20:47:51 elliott: agreed 20:47:52 A nice (read: terrible) way to emulate goto which I think PHP has the equivalent of is with a labeled switch with fallthrough. 20:47:58 * Vorpal stabs elliott with a cardboard fox 20:48:02 Gregor: PHP has goto. 20:48:12 SgeoN1: yeah, but goto in php is bad in that it doesn't work too well 20:48:12 GOTO EVIL :P 20:48:19 What happened to the days when this channel was about the interesting stuff that people weren't doing just because they were paid to do it? 20:48:24 SgeoN1: also it can't do everything that cobol goto can do 20:48:28 elliott: Yarly 20:48:30 Oh right, cheater99 (and others) happened. 20:48:37 * elliott AAAAAANGST 20:48:40 * elliott angus 20:48:58 elliott, well, /ignore cheater99 20:49:01 lets all do it 20:49:01 elliott: that's cute 20:49:02 Well, have your script mechanically copy and paste so that function calls are used where gotos qere 20:49:06 Vorpal: GO!! 20:49:11 Duplication of code, the works 20:49:14 Vorpal: been doing that for ages, oklopol will never do it and the beast will get fed :p 20:49:17 SgeoN1: yeah 20:49:20 Vorpal: but ok, fine, i'll put it on this partition too 20:49:21 elliott, please. 20:49:23 SOLIDARITY 20:49:26 SgeoN1: you too 20:49:27 Should work just fine 20:49:31 MAYBE HE'LL CUT HIMSELF TO DEATH ;_; 20:49:32 elliott, I just added the ignore 20:49:38 with the angst he projects onto others 20:49:38 cheater99, goes on about his job *once*, and the channel is RUINED FOREVER? 20:49:48 Phantom_Hoover: you have terrible memory. 20:49:51 SgeoN1: what does "alter" do actually? 20:49:58 cheater99 has never once discussed anything interesting. i don't recall him mentioning esolangs once. 20:49:59 elliott, other examples? 20:50:00 SgeoN1: it changes the global state, doesn't it? 20:50:00 elliott: i just ignored him today 20:50:06 but then realized maybe it was my fault 20:50:19 Everything I do is becoming decreasingly esoteric :P 20:50:20 Cheater, Google it 20:50:22 Phantom_Hoover: every single thing he says is pointless and stupid. and most of it is blatant trolling. 20:50:26 SgeoN1: ok 20:50:27 Gregor: Yes, but you're not an irritating fuckwit. 20:50:34 I'm not D-8 20:50:36 NOOOOOOOOOOOOOOES 20:50:44 elliott: awwww 20:50:45 <3 20:51:13 Gregor: no your wit is too ugly to be fucked 20:51:20 Ouch :P 20:51:29 Gregor: Try washing your beard sometime. 20:51:35 Then you can start working on your wit. 20:52:33 Gregor has a beard? 20:52:33 -!- Phantom_Hoover has left (?). 20:52:39 I don't :P 20:52:45 He does. 20:52:46 Not even physically capable. 20:52:48 It's unwashed. 20:52:52 -!- Phantom_Hoover has joined. 20:52:56 Wha... 20:53:18 SgeoN1: so alter is basically like a static var containing a callback 20:53:31 Gregor, of course you can't grow a beard. 20:53:39 You're a pansy little Englishman! 20:53:46 *Anglo-Jew 20:53:52 yeah let's all call Gregor names and then ignore him and maybe kill him 20:53:56 by the way 20:53:58 after my dream 20:54:03 i so wanna see Gregor 20:54:06 and drink tea with him 20:54:10 I don't drink tea :P 20:54:20 i'll make ya 20:54:24 i don't either tho 20:54:32 Gregor has no taste of smell 20:54:35 so he just like 20:54:37 tastes and smells only dung 20:54:39 his entire life 20:54:41 i think this is accurate 20:54:45 :o 20:54:54 PLEASE EXPAND GREGOR ON THIS TOPIC 20:55:34 pikhq: Sheesh, I think AMD actually comes out more expensive overall due to the motherboards and shit. 20:55:46 Gregor, you're not a Jew, you're a filthy Englishman! 20:55:54 pikhq: Found a nice motherboard for $49.99. All compatible stuff and the like. AM3. But 95W max cpu. 20:55:59 pikhq: The others are... more expensive. 20:56:04 (ATX.) 20:57:24 Gregor, you can't grow facial hair because you didn't eat PORRIDGE. 20:58:35 SgeoN1: i don't see any other way to understand it than that 20:59:10 That works, I guess 20:59:59 SgeoN1: have you ever done any translation of cobol to something else? 21:00:04 or automatic refactoring of cobol? 21:00:05 Nope 21:00:16 And mope 21:00:19 N 21:00:22 heheh 21:05:04 elliott, you MUST try the vertical waterfall with boat thingy, it is super-fast 21:05:05 wow 21:05:15 very fun 21:05:32 oklopol: btw, how is this topology based cobnti 21:05:44 oklopol: btw, how is this topology based continuity broader defined than the sup based one? 21:06:23 oklopol: i don't think the hypothesis for the sup based one actually ensures there's a topology, does it? 21:06:34 because if you take a certain topology on the poset, then that sup continuity is actually just the usual continuity 21:06:55 err well you add the topology. 21:07:17 but which topology? 21:07:41 because function continuity isn't an invariant of topology, is it? 21:07:47 a topology where open sets are upper sets that you can't reach with sups of directed sets from their complement...... 21:07:51 :P 21:07:57 -!- Phantom_Hoover has quit (Quit: Reboot). 21:08:48 an upper set is (x, oo) ? 21:08:56 well almost 21:09:01 well 21:09:07 that's one, but the definition is 21:09:30 x \in set, y >= x => y in set 21:10:01 if we have infs then yours gives all of these, i think, if you take [x, oo) and (x, oo) 21:10:04 I THINK 21:10:11 nah 21:10:13 i guess it doesn't 21:10:28 even in finite lattices 21:10:42 yah 21:11:08 so anyway i do admit that may not be the most natural topology on an ordered set 21:11:08 it was just a mental shortcut 21:11:17 oh yeah 21:13:34 * cheater99 can't wait till he has all his maths books with him.. 21:13:45 mmmmmmm math books 21:14:18 i was just reading about different topologies for CA today 21:14:34 sexy topologies 21:14:50 CA? 21:14:59 cellular automata 21:15:07 oklopol: i've got like 300 maths and physics books.. 21:15:21 in about 5 languages :) 21:15:23 i have like 1000 on my computer, but very few irl 21:15:29 no that's irl 21:15:36 :o 21:15:38 on my computer that's like 5 digits 21:15:38 yummy 21:15:47 i'd just lick them all day 21:15:55 that's what i do.. 21:15:59 when they're with me :( 21:16:10 hadn't seen most of 'em since i started my moves around europe 21:16:12 why aren't they with you? 21:16:14 ic 21:16:19 moved out of my hometown 21:16:24 to See the World! 21:16:43 in the meantime i acquired some smaller collections :p 21:16:46 do you have kurka's topological and symbolic dynamics 21:16:52 some on britain, some in germany 21:16:58 nope 21:17:05 or do you mean in pdf? 21:17:25 well i meant tree 21:17:35 no 21:17:39 the nearest i have is kuratowski 21:17:44 :D 21:18:07 didn't kuratowski also do some computational stuffs and things? 21:18:15 yeah 21:18:15 maybe he did everything tho 21:18:30 btw 21:18:34 i have a book (not by him) on computable numbers 21:18:50 kurka is actually written with a u with o on top, in case you don't know the guy 21:18:58 kuratowski did mostly stuff around the sets R and N. 21:19:17 don't know him 21:19:27 never heard of that character either 21:19:31 sounds... scandinavian 21:19:40 but kurka means little chicken in polish. 21:19:55 actually i thought it was polish or something 21:20:20 brb tho! 21:22:26 ahh he's czechish 21:22:28 czech. 21:22:51 -!- Sgeo has joined. 21:23:00 www.cts.cuni.cz/~kurka/ 21:23:14 when i saw the first name (petr) it was obvious he's czech 21:24:58 elliott, new screenshots at http://sporksirc.net/~anmaster/minecraft/screenshots/ (2010-11-10_23.45.52.png and onwards). Including a fun attraction 21:25:16 not the loop system yet 21:25:22 was testing out the general idea 21:27:08 -!- pikhq has quit (Ping timeout: 240 seconds). 21:39:18 -!- Sasha2 has quit (Read error: Connection reset by peer). 21:40:33 -!- Sasha has joined. 21:55:03 -!- augur has quit (Remote host closed the connection). 22:01:00 * Sgeo stupidly visits a site he knows will give him nightmares tonight 22:01:57 -!- nooga has joined. 22:02:01 gaaaah 22:02:06 tk looks like motif 22:02:09 uglyy 22:06:15 -!- Phantom_Hoover has joined. 22:08:34 * Sgeo decides not to read the scary parts 22:08:54 Sgeo: what site? 22:09:06 http://exitmundi.nl 22:10:35 Tk !!!!!!!!! 22:10:51 Sgeo, I have the same problem. I hate you. 22:11:01 * Phantom_Hoover exercises SELF-CONTROL. 22:16:12 suddenly i want to make an UI based on boxes and wires 22:16:18 and inlets and outlets 22:16:21 elliott, is the Mysterious Teleport Hacker the guy with the alien science? 22:16:35 but i have no idea for what 22:16:37 yet 22:20:49 haha 22:20:53 2000kB/s 22:21:01 i love my provider 22:35:54 -!- Mathnerd314 has joined. 22:41:23 -!- cheater99 has quit (Ping timeout: 245 seconds). 22:48:22 hmm, the grand plan for today was to not drink and instead work on getting my kernel to relocate itself to a safe area of the address space and put user-accessible pages in a different area 22:54:40 Lemme guess: you drank a lot and made the kernel do a sum total of nothing. 22:55:05 Hey, can we contract you to do OS stuff we can't be bothered to do ourselves? 22:55:31 elliott asked yesterday if I could be contracted to write a compiler in assembly 22:55:36 Why not define "safe area" as "area where the kernel is"? 22:55:56 Sgeo: but the kernel is in the *same* area as the other stuff, so that won't fly 22:55:57 Then again, I guess that could slow down application-level pointer arithmatic 22:56:06 -!- cheater99 has joined. 22:56:25 I need them to live in discontiguous pages, at least discontiguous *virtual* pages 22:56:37 and I need the user-space to not have write access to any kernel stuff 22:56:44 Sgeo, the standard method is IIRC to make the kernel be in the higher-half address space. 22:59:04 Phantom_Hoover: elliott, is the Mysterious Teleport Hacker the guy with the alien science? 22:59:07 who? 22:59:20 and amd64's treatment of addresses (where there's one part growing from the bottom and one growing from the top) pretty much screams "put the kernel HEeeeeEeeRE!" and points to the upper halfspace 22:59:25 The guy who's been screwing with all of the teleport experiments. 23:00:02 oh, another MTH? you do know there are several of them, right? 23:00:10 Phantom_Hoover: name plz 23:00:13 you *do* know, right? 23:00:20 elliott, there *isn't* a name yet. 23:00:35 * Sgeo knows 23:00:37 There's just the two mishaps during the tests of the apparatus. 23:00:41 Phantom_Hoover: oh, the Bad Guy? I know no more than you. And what tells you it's a *person*? 23:00:41 Sgeo, shut up. 23:00:51 Sgeo: If you say anything I will eviscerate your body. 23:00:58 elliott, well, that's the theory I was fielding. 23:01:16 Phantom_Hoover: Where are you up to? 23:01:23 I've read The Story So Far and no further, as it stands. 23:01:25 http://www.youtube.com/watch?v=TywmpMQYojs 23:01:53 elliott, I'm up to 'Two killed in "transporter accident"'. 23:02:00 Sgeo, what's that video? 23:02:24 I think it's probably the song where all the rhymes seem to point to expletives/sexual references but then THEY DON'T 23:02:30 elliott, yes 23:02:32 (The version I saw is rather amusing, at least.) 23:02:34 elliott: THE VERY song 23:02:35 (But I don't have Flash.) 23:02:46 Phantom_Hoover: Gotta love the BBC "quotes" there. 23:03:01 Heh. 23:03:20 Phantom_Hoover: I always get the feeling that the BBC is vaguely sceptical about everything that happens ever. 23:03:30 -!- digimunk has quit (Quit: Leaving.). 23:03:34 Hah, I knew Mitch was hyperspacing. 23:03:43 being sceptical is a very good approach to pretty much everything though 23:03:55 elliott, except nullity 23:04:10 olsner: but not in a condescending way :) 23:04:15 Phantom_Hoover: But not superpositioning... 23:04:16 I don't think scepticism applies to nullity, really. 23:04:22 Just stupidity. 23:05:12 elliott: you're saying condescending scepticism is bad? 23:05:22 olsner: when applied to everything, yes 23:06:00 anything applied to everything is bad, everything should be applied (only) where appropriate 23:06:11 olsner: "Recent scientific studies show strong possibility that [thing] can help significantly alleviate Alzheimer's. [well written, factual, sourced article]" "I'll wait to see the reports on the JOURNALIST'S Alzheimer's first, heh." 23:07:26 (... then again, we should apply anything where inappropriate too - so that we can know how well it works) 23:07:28 elliott, but that's *already* a sceptical headline. 23:07:45 elliott: written by a JOURNALIST? 23:07:46 Phantom_Hoover: Healthy scepticism != condescending "scepticism". 23:07:55 olsner: It takes place in my fantasy world, k. 23:08:01 lol, k 23:08:15 olsner: Now write the ElliottOSLang compiler. 23:08:24 * olsner has very limited knowledge on elliott's fantasy world 23:08:34 olsner, oh, you really must learn. 23:08:35 olsner: EVERYONE HAS A PET BEE 23:08:47 Yes, everyone who learns of my fantasy world gets to LIVE IN IT 23:08:48 Phantom_Hoover: oh, DO I 23:08:50 It's like wonderland, except crazier and less conservative. 23:09:14 At least I don't believe that ElliottOS will ever get significant adoption at all :P 23:09:25 Although I expect Phantom_Hoover to switch, or I'll slit his throat. 23:09:53 Well, I wouldn't switch fully. 23:09:58 killing your potential users means they become guaranteed non-users 23:10:05 IS THIS A GOOD THING??? 23:10:06 Until people who write software started using it. 23:10:39 Phantom_Hoover: It'd have an IRC client and a web browser, what more do you need. 23:10:55 Golly, or an equivalent, 23:11:15 (For jiggery-pokery. 23:11:17 *) 23:11:40 Hey, could you make Emacs a cakewalk by writing an ELISP evaluator in CL? 23:11:53 Nethack. Can't go without Nethack. 23:12:23 Phantom_Hoover: You can play NetHack over VNC or something. 23:12:30 VNC? 23:12:36 Phantom_Hoover: Okay, fine, NetHack too. 23:12:44 VNC = remote display protocol. 23:12:47 Oh. 23:12:59 Like X11 except you don't use it locally either. 23:12:59 Phantom_Hoover: Oh, and a text editor. 23:13:17 Yeah, that's what I was rambling about Emacs for. 23:13:19 Phantom_Hoover: (The UI will probably be something like a souped-up Emacs with a lesser focus on text.) 23:13:43 Add an ELISP interpreter to that, and you basically have Emacs, unless I'm missing something. 23:14:20 -!- poiuy_qwert has quit (Quit: This computer has gone to sleep). 23:14:29 Phantom_Hoover: Well, except not nearly as good. 23:14:29 Phantom_Hoover: I mean, Emacs doesn't have orthogonal persistence. 23:14:55 Or ... any of my OS features. 23:14:55 And its UI is a hopelessly inadequate implementation of what it could be. 23:15:05 Yes, I know that. 23:15:06 -!- poiuy_qwert has joined. 23:15:44 But if you can run ELISP stuff with minimal modification you have a tremendous library of useful stuff. 23:16:07 Phantom_Hoover: (I mean, Genera's interface is basically "Emacs on crack" already.) 23:16:19 Phantom_Hoover: elisp is possibly the worst Lisp ever thought of ever. 23:16:19 Phantom_Hoover: Even disregarding that... they're usually very tied to the idea of "files". 23:16:27 I sort of don't have files. 23:16:30 Simulate files! 23:16:33 Yes, I know ELISP is basically stoneage-lisp. 23:16:38 Also, it's not like I'm not reimplementing everything myself anyway. Compared to what I already have on my plate, what's a bunch of elisp? 23:16:44 It was a crazy idea, remember? 23:16:56 Sgeo, yes, and ruin the whole concept. 23:16:59 What's the oldest lisp? 23:17:07 Sgeo: I have one simple word for that. 23:17:07 Sgeo: No. 23:17:09 LISP, you cretin. 23:17:37 What Phantom_Hoover said :P 23:17:37 Oldest Lisp is LISP. 23:17:39 * Sgeo feels "dolt" woul have been more appropriate 23:18:15 Sgeo: Whatever you say, cretin. 23:18:28 Sgeo, that's the wrong kind of tone, although not having the inflection makes it seem harsher. 23:18:35 "Nixon, you dolt", you dolt! 23:19:08 Sgeo: Sure thing, cunt. 23:19:27 http://www.fark.com/cgi/comments.pl?IDLink=5351459 23:19:39 cretins, the whole lot of you 23:20:22 -!- FireFly has quit (Quit: swatted to death). 23:20:37 hmm, and I may have cut myself on a potato chip 23:21:03 papertato cut! 23:21:05 ...how 23:21:10 I once cut a piece of paper in half with another piece of paper. 23:21:31 * Sgeo vaguely assumes that olsner isn't hemophiliac 23:21:34 is the internet down for everyone else 23:21:45 elliott, no. 23:21:55 elliott, yes. We are all hallucinations 23:22:00 wat. 23:22:05 Sgeo, have you actually read elliott's screeds against filesystems? 23:22:09 Phantom_Hoover: Well it's down for me. Get it back up. 23:22:22 Phantom_Hoover, I ... don't think so 23:22:22 oh the internet is sort of working 23:22:24 slowly. 23:22:40 Phantom_Hoover: Oh, don't make him. 23:22:50 That rant wasn't the best rant I've ever done :P 23:22:52 I want to read it! 23:22:54 (It's all true of course, just expressed badly.) 23:22:55 elliott, wait, you don't like the one on catseye any more, do you? 23:22:57 Ah. 23:23:08 I'll probably agree with it, tbh 23:23:43 http://catseye.tc/ehird/files-suck.html is the essay; but keep in mind what was just said. 23:23:45 Sgeo: http://catseye.tc/ehird/files-suck.html 23:23:45 I think that's the URL. 23:23:59 If you like it, you're as whiny as I was when I wrote that. 23:24:27 Sgeo: assumptions, assumptions 23:24:37 RIP olsner 23:25:12 Although the point has been made elsewhere, like on Stanislav's blog. 23:25:18 -!- zzo38 has joined. 23:25:23 olsner is haemophiliac; blood turns him on. 23:25:28 (SEE ALSO: Twilight.) 23:25:36 OTOH "whiny" is not a word from which Stanislav retracts. 23:25:44 wouldn't that be a haemosexual? 23:26:00 Phantom_Hoover: Ahh, I'd rather my whinging than his incessant pretentious douchebaggery :) 23:26:00 Love the guy, but he's the most pompous fucker you'll ever see. 23:26:15 olsner: Maybe it depends if it is sexual or not? 23:26:19 olsner: "paedosexual" 23:26:21 "Of course not that's simply wasteful, isnt it? Why not store it as the rich representation in the first place, and have functions operate on it directly? That saves computing time and is also much simpler." 23:26:21 -!- wareya_ has quit (Read error: Connection reset by peer). 23:26:27 elliott: right 23:26:32 Uhh... 23:26:35 I think haemophiliac works. 23:26:35 Sgeo: unicode fail 23:26:39 ("?") 23:26:45 zzo38: getting turned on is by definition sexual, I think 23:26:59 olsner: In that case, I guess it is "haemosexual". 23:27:13 -!- SgeoN1 has quit (Ping timeout: 245 seconds). 23:27:23 -!- wareya has joined. 23:27:25 Stanislav Pestov? 23:27:34 elliott,Sgeo: Yes it is unicode failure. (I get shaded blocks after "not " and after "isn") 23:27:36 zzo38: then again, as elliott pointed out, the terms don't necessarily differentiate between philiacs and sexuals 23:27:36 Stanislav Gorgistonnion. 23:27:57 Wow 23:28:03 olsner: Ah. Well, you select the term according to what is properly meant by the terms in the current context, then. 23:28:04 Should I link him to it? 23:28:06 Utterly wrong person I was trying to think of 23:28:15 Sgeo: ? 23:28:23 I had the Factor guy in mind 23:28:26 Phantom_Hoover: What, to the Loper OS blog? Naw. He can find it himself. 23:28:59 Sgeo: That's Slava. 23:29:05 Sgeo: appropos RIP - "64-bit instruction pointer olsner" makes no sense... 23:33:05 elliott, a thing: what would your Lisp OS actually be called once Mitosis came to a sufficiently advanced point? 23:34:52 oh, here I go again... open video in fullscreen, press the fullscreen button, expect video to cover entire field of vision 23:34:56 Hmm, Stanislav lists Mathematica as a non-broken programming system. 23:35:30 -!- elliott has quit (Ping timeout: 245 seconds). 23:35:32 olsner, HTML5 on YouTube? 23:35:46 the real world is a bad window manager, where moving windows means moving monitors aroudn and you can't resize windows without investing money 23:36:34 Wouldnt you have to invest money to move wndows? 23:36:42 Or am I hopelessly lost in your analogy 23:37:06 no, normally you press the mouse button down in the right place then move the mouse and release the button later 23:37:26 that's completely free once you've invested in the computer, pointing device and your first monitor 23:37:38 You're talking about an actual window manager named "the real world"? 23:37:55 yes, I'm talking about the actual real physical world here 23:38:12 I think I get it 23:38:18 i.e. making the full-screen video larger requires getting a bigger screen 23:38:21 which costs money 23:38:22 -!- pikhq has joined. 23:38:26 O... OH 23:38:31 -!- pikhq has quit (Read error: Connection reset by peer). 23:38:37 HO2 23:38:55 -!- pikhq has joined. 23:39:08 * Mathnerd314 drinks some dihydrogen monoxide 23:39:18 * Sgeo gives pikhq some hydrogen dioxide 23:39:20 -!- elliott has joined. 23:39:25 Internet hiccup. 23:39:37 15:34:56 Hmm, Stanislav lists Mathematica as a non-broken programming system. 23:39:37 looked like an elliott hiccup from here 23:39:38 Indeed. 23:39:49 15:33:05 elliott, a thing: what would your Lisp OS actually be called once Mitosis came to a sufficiently advanced point? 23:40:10 Phantom_Hoover: The perfect OS must of course have the perfect name. I cannot possibly trust myself with deciding what the perfect name is without extensive thought; therefore, the actual OS is unnamed. 23:40:23 They make hybrid Bluray/VHS players. 23:40:28 Why god why? 23:40:37 And not DVD? 23:40:38 "Fusion" would be the logical name, but it's too boring. 23:40:49 -!- choochter has quit (Quit: lang may yer lum reek..). 23:40:54 Sgeo: I highly doubt there's a Bluray player that can't handle DVDs... 23:40:56 Phantom_Hoover: Yeahno. 23:40:56 OTOH, it's the only biological term for the phenomenon in question. 23:41:01 Hydrogen Dioxide is the perfect name! 23:41:11 So more cleverness is neaded. 23:41:12 pikhq: imagine making an C/amd 64, 64k and 64-bit in the same box! 23:41:33 Sgeo, naw, dihydrogen dioxide is where it's at. 23:41:46 But that actually exists! 23:41:47 Sgeo: I'm a fan of hydric acid. 23:43:04 pikhq, is that just H_2? 23:43:11 Phantom_Hoover: H2O. 23:43:18 Phantom_Hoover: The phenomenon is NIH. 23:43:26 pikhq: I thought H2O is hydroxic acid, not hydric acid? 23:43:28 http://www.thefreedictionary.com/Hydrogen+dioxide 23:43:30 WTF 23:43:42 Oh, god. 23:43:46 Why can hydrogen's di- be omitted? 23:43:48 s/neaded/needed/ 23:44:40 Uncorrected typos can cause the end of the world. Do be more careful, Phantom_Hoover. 23:44:40 Sgeo: it's ionic 23:44:53 Sgeo: I don't know why? I don't study chemistry 23:45:11 God, my jokes are boring 23:45:49 Sgeo: alternately, people just like calling it that 23:46:34 So what DO you call HO2? 23:46:48 Which I'm pretty sure can't exist, but whatever 23:47:06 Sgeo: You call it "monohydrogen dioxide" 23:47:38 Sgeo: googling suggests HO2 exists 23:48:02 Sgeo: hydroperoxyl? http://en.wikipedia.org/wiki/Hydroperoxyl 23:48:18 I'd suppose one of the Hs get taken away when in solution. 23:50:24 -!- augur has joined. 23:55:30 HO2 would be a free radical... And probably quite reactive one at that... 23:57:18 Probably not quite as bad as FO2... 23:59:28 elliott, what would be the way you'd write software in Lisp86? 23:59:36 -!- MigoMipo has quit (Read error: Connection reset by peer). 23:59:43 Phantom_Hoover: Such an uncouth name for the perfect language!