00:01:09 <ehird> my problem is that I do i = (*r)++; but when we return we try and use that as a primitive when it's not; we need the interpreter, but that's lost :-(
00:08:05 <ehird> "You can tune a file system, but you cannot tune a fish."
00:08:29 <ehird> reads better with s/cannot/can't/ as in bsd
00:29:27 -!- psygnisfive has quit (Remote closed the connection).
00:58:59 <kerlo> Eigen eigen eigen eigen eigen.
03:18:09 -!- coppro has quit ("The only thing I know is that I know nothing").
03:42:58 -!- coppro has joined.
04:01:11 -!- coppro has quit ("The only thing I know is that I know nothing").
04:38:04 -!- Sgeo has joined.
04:38:55 -!- oerjan has quit ("Good night").
05:11:48 -!- pikhq_ has joined.
05:12:58 -!- pikhq has quit (Nick collision from services.).
05:13:03 -!- pikhq_ has changed nick to pikhq.
06:16:55 -!- Sgeo has quit (Read error: 54 (Connection reset by peer)).
07:01:17 <fizzie> "Oh eigen, eigen, eigen, I made it out of clay."
07:05:26 <fizzie> And about instruction dispatchers, I think mooz wrote something in ASM where each instruction-handling block was a fixed power-of-two-size, so it didn't even need a table lookup; it was just "relative jump to playfield[ip] << x".
07:07:28 <fizzie> There's also a well-known addendum to that tunefs quote: "Actually, you indeed can tune a fish; usually at a canning factory."
07:25:02 -!- FireFly has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:02:37 -!- FireFly has quit ("Later").
08:02:41 -!- bsmntbombdood_ has changed nick to bsmntbombdood.
08:06:22 -!- tombom has joined.
08:11:30 -!- neldoreth has joined.
08:22:15 -!- DarkPants has joined.
08:23:31 -!- GreaseMonkey has quit (Nick collision from services.).
08:23:41 -!- DarkPants has changed nick to GreaseMonkey.
08:28:58 -!- pikhq has quit ("leaving").
08:57:45 <fizzie> Tunefs; at least my tune2fs man page is rather boring.
08:58:19 <fizzie> Well, there's this "BUGS: We haven’t found any bugs yet. That doesn’t mean there aren’t any..."
09:00:14 <fizzie> But not the the tunafish pun. That's in at least FreeBSD tunefs(8) man page, probably in many others too.
09:01:20 <fizzie> Not in Solaris 10, even though it has a tunefs tool. Maybe they're too serious for that sort of stuff.
09:07:05 <AnMaster> fizzie, do you use t much in fungot? I forgot.
09:07:06 <fungot> AnMaster: dark one:... but none have found him and his fellow titans were ousted from mount olympus. ( macmillan illustrated animal encyclopedia)
09:07:56 <AnMaster> actually I doubt fungot will gain much from this change, since it actually use so many fingerprints.
09:07:56 <fungot> AnMaster: any small object that is stored in a little squeamish sometimes..." " interesting point, nobby. i rushed to its deadly brink. i think-" ( dining with a unicorn horn rule is if it ain't broke then don't fix it.
09:09:59 <fizzie> I use t never, because any blocking read would anyway block the other threads, and I don't want to busy-loop either.
09:11:07 <AnMaster> fizzie, well this won't help you much, it just avoids allocating much of the fingerprint opcode stacks until it is actually needed.
09:11:28 <fizzie> I guess I could do some sort of "whenever we have something interesting to do, fork a thread for it and let the other IP busy-loop and poll for input; when there's nothing going on, just do a blocking read" logic, but it sounds awfully complicated.
09:11:44 <AnMaster> I'm down to ~310 malloc() and ~60 realloc() calls for a mycology run now.
09:12:04 <AnMaster> and some of those are due to glibc internals
09:12:30 <AnMaster> as in: glibc allocates for IO or network stuff internally
09:12:56 <AnMaster> how could you have missed that
09:13:06 <AnMaster> considering how often it has been discussed in here.,
09:13:34 <lifthrasiir> bsmntbombdood: http://iki.fi/matti.niemenmaa/befunge/mycology.html FYI. it is an extensive funge-98 test suite.
09:13:49 <bsmntbombdood> why do you care about malloc calls in a test suite?
09:14:24 <bsmntbombdood> i assume you are reducing allocation for performace
09:14:41 <lifthrasiir> well, it is also a practical performance test for funge interpreter... and i recommend you to look up previous logs
09:14:47 <AnMaster> I care about reducing memory waste. The time is the same (since glibc malloc() allocates in chunks from the kernel, and the syscall overhead isn't as large)
09:14:59 <fizzie> I would say something to the effect of "Mycology is not very representative of real-world Funge workloads", were there any real-world workloads to speak of.
09:15:34 <lifthrasiir> fizzie: of course not an representative of typical funge program, i agree.
09:15:38 <AnMaster> I just noticed I had an unreasonable number of pointless malloc() calls before for fingerprints
09:16:12 <AnMaster> now I'm down to at most 26 per ip, and only when it is actually needed.
09:16:39 <AnMaster> though they are actually realloc of null pointers
09:17:16 <AnMaster> oh and the code is much less complex now too as a bonus.
09:26:07 <AnMaster> lifthrasiir, tracking exact bounds add almost 0.008 seconds of overhead (wall clock). Of course IO is a way larger overhead. Around 0.03 seconds compared to redirecting to /dev/null
09:26:46 <AnMaster> if -f is used for fully buffered stdout (instead of the default line buffered) the difference to > /dev/null is closed to 0.15, but still...
10:16:08 -!- GreaseMonkey has quit ("X-Chat -> http://xchat.org <- At least when I quit I don't look like a lamer").
12:05:49 -!- Judofyr has joined.
12:28:57 -!- jix has joined.
12:29:57 <AnMaster> Deewiant, hm I have a test program here that CCBI and cfunge prints different things for (rcfunge segfaults). And I forgot why I wrote it, and I forgot to document it...
12:30:02 <AnMaster> "HTRF"4(n 1 0aaaaa****- #v L 'b,a, @
12:30:34 <AnMaster> cfunge prints a followed by a newline and exits, CCBI prints b
12:30:49 <AnMaster> not sure if this signifies anything
12:31:10 <AnMaster> well it means ccbi didn't reflect
12:31:47 <AnMaster> Deewiant, I have no clue what it tests even!
12:32:17 <Deewiant> You're passing a negative number to L
12:32:19 <AnMaster> Deewiant, because the value is negative it seems
12:32:31 <AnMaster> code comment says: /// L - Forth Roll command
12:33:59 <Deewiant> What if you add '0 to the TOS instead of printing a or b
12:34:32 <AnMaster> um? You mean checking what is on stack after?
12:34:43 <Deewiant> My guess is that CCBI just converts it to unsigned and sees it as trying to roll from way below where the stack ends
12:35:27 <AnMaster> last rcfunge2 prints b instead of segfaulting btw
12:35:37 <AnMaster> just noticed I had an outdated version
12:37:07 <Deewiant> To the TOS, to see if it's still 1 or if it's 0 or something else
12:37:22 <AnMaster> wouldn't adding a . be better?
12:38:29 <AnMaster> "HTRF"4(n 1 0aaaaa****- #v L .'b,a, @
12:38:37 <AnMaster> $ other/ccbi/ccbi_linux/ccbi.64 tests/frth-test.b98
12:38:47 <AnMaster> $ build/cfunge tests/frth-test.b98
12:39:11 <Deewiant> If you put another . there you'll get 0 1 b
12:39:30 <AnMaster> well empty stack == 0 so I expect to get that in the end
12:39:45 <AnMaster> $ build/cfunge tests/frth-test.b98
12:39:48 <AnMaster> $ other/ccbi/ccbi_linux/ccbi.64 tests/frth-test.b98
12:40:04 <AnMaster> $ ~/funges/interpreters/rcfunge2/funge tests/frth-test.b98
12:40:48 <AnMaster> Deewiant, anyway I'll just document this properly, and include it, since it did cause older versions of rcfunge to segfault it might be useful to someone else in the future
12:53:53 -!- Sgeo has joined.
12:59:44 -!- KingOfKarlsruhe has joined.
13:03:46 -!- jix has quit (Read error: 113 (No route to host)).
13:12:16 <AnMaster> Deewiant, seems Mike Riley retconned what should happen to "L should act like forth -roll with a negative argument"
13:12:42 <AnMaster> as usual I'm implementing the version that was around when I originally did the implementation.
13:12:52 <Deewiant> Ah, interesting, such a thing exists
13:13:26 <AnMaster> Deewiant, however I consider this undef, it was last time I looked.
13:17:43 <Deewiant> Hmh, it's not in forth-79 nor -83
13:18:30 <Deewiant> Nor in the 1994 ANS Forth standard
13:18:45 <lifthrasiir> Deewiant: i wondered that -roll is even in the standard
13:18:53 <lifthrasiir> since i cannot find any information about it
13:19:09 <fizzie> It's not in GNU Forth either. It seems to be a bit unstandard.
13:20:02 <fizzie> It is in PFE -- http://eckhart.stderr.org/doc/pfe-doc/ -- too.
13:20:22 <Deewiant> Yeah, I saw that, but it says "no info".
13:20:59 <Deewiant> And there's no normal ROLL there.
13:21:28 <fizzie> Uh, yes there is: http://eckhart.stderr.org/doc/pfe-doc/words/w-core-0161.html
13:21:34 <fizzie> It's just under the [ANS] category.
13:21:44 <fizzie> And it is equally undescriptive in the documentation.
13:21:57 <Deewiant> Hmm, wonder how I managed to miss that.
13:22:07 <Deewiant> I F3'd repeatedly and saw only two ROLLs.
13:22:24 <fizzie> You didn't accidentally search for "-roll", did you?
13:22:47 <fizzie> Sadly, even -rot seems not to be in the ANSI standard.
13:23:30 <fizzie> The fact that roll rotates n+1 items came as a bit of a surprise to me.
13:24:44 <Deewiant> And yes, it does, or n, or whatever.
13:24:55 <Deewiant> I can't remember the details but it's not a constant value.
13:26:54 <fizzie> The only example in the current FRTH spec is "Example: n543210a-L will leave a stack of: 2 3 4 5 0 0 0 0 0 0 1" but that refers to the rather vaguely defined -ROLL word.
13:27:39 <Deewiant> Well, it makes sense; it rolls in the opposite direction
13:28:02 <Deewiant> That's what it does in Factor, although for a constant size of 4.
13:28:06 <fizzie> Yes, but "4 roll" rotates the 5 topmost items in Forth.
13:28:42 <Deewiant> I saw that in one of the standards I just browsed through.
13:29:21 <fizzie> Well, I've just been reading this one draft.
13:29:27 <fizzie> Maybe it is not a reliable source.
13:30:34 <fizzie> Draft of X3.215-1994, that is; and in here "1 roll" equals "swap".
13:31:06 <Deewiant> It was in the 79 or 83 where 1 roll was a nop.
13:31:16 <Deewiant> It can have changed, of course.
13:31:43 <fizzie> 1 roll is a swap in GNU Forth, FWIW: "roll x0 x1 .. xn n – x1 .. xn x0 core-ext “roll”"
13:32:08 <fizzie> Strange change, if change it be.
13:32:44 <Deewiant> http://forth.sourceforge.net/standard/fst83/fst83-12.htm#roll says 0 roll is a nop.
13:33:22 <Deewiant> https://mywebspace.wisc.edu/lnmaurer/web/forth/Forth-79.pdf says 1 roll is a nop.
13:36:22 <fizzie> Yes, it certainly seems they've decremented (or incremented, depending on your viewpoint) it between those two standards.
13:39:14 * Sgeo is upset about A!!CT
13:40:01 <Sgeo> I'm also upset about one of the people from A!!CT always talking to be about his issues with the founder of A!!CT
13:40:26 <Sgeo> I think.... I want that last line scrubbed from the log :/
13:40:56 <fizzie> How does one google for something like that? Even quoted, Google just wants to find matches of "a CT" this-or-that.
13:41:39 <Sgeo> Are there search engines that do handle that sort of thing properly?
13:43:49 <fizzie> Yes, I did think about using that.
13:44:10 <fizzie> Your search - "A!!CT" - did not match any documents.
13:45:04 -!- KingOfKarlsruhe has quit (Remote closed the connection).
13:49:34 * Sgeo still can't believe that his map of RSTV is still around
13:50:00 <Sgeo> Actually, geocities is owned by Yahoo!, so I should still be able to get to it
13:51:36 <Sgeo> http://www.geocities.com/sgeo_sgeo/Pool2.JPG a pool I built
13:51:47 <Sgeo> http://www.geocities.com/sgeo_sgeo/Ute.bmp a very broken norn
13:51:53 <fizzie> A!!CT, RSTV... it's like a whole other vocabulary.
13:53:01 <Sgeo> Probably only 6 people or less have heard of RSTV.. or at least what I'm referring to when I say RSTV
13:54:02 <Sgeo> If you're willing to install Active Worlds, I can show you RSTV
13:54:53 <fizzie> Related to the first screenshot, I do remember a magazine article about Active Worlds back in the 1990-something when it was being introduced. I don't think I'm interested enough for installation, though.
13:55:01 <fizzie> Do you have some sort of virtual reality hobby or something?
13:55:11 <Sgeo> I like this sort of thing
13:55:54 <fizzie> There was a VRML model of Helsinki city centre once; I wonder if it still exists.
13:56:21 -!- BeholdMyGlory has joined.
13:56:37 <Sgeo> I've been in and loved Cybertown, Active Worlds, and Second Life
13:57:20 <Sgeo> I've been to There, Worlds.com, vSide, Kaneva, IMVU, and at least one that I remember had a grafitti wall but forgot everything else about
13:57:37 <Sgeo> I _hated_ IMVU
13:59:33 <fizzie> I've seen IMVU banner ads; the rest I don't think I've even heard of.
14:00:03 <Sgeo> You haven't heard of Second Life?
14:01:22 <Sgeo> http://i42.tinypic.com/e5ques.png I'm the reason that sign exists
14:02:55 <fizzie> Well, SL, yes. And I think the Cybertown name rings a bell, but I couldn't say anything what it is.
14:03:22 <Deewiant> Of those, I think I've heard only of Second Life.
14:04:22 <fizzie> Deewiant: Had you been diligently reading MikroBitti, you would've seen a featured article about Active Worlds (I think it might even have been alphaworld at that time) at some point in 1995 or so.
14:04:57 <Sgeo> fizzie, Cybertown is a VRML Blaxxun-based community
14:05:38 <Sgeo> That IVN basically killed when they bought it. When it was free, it was very popular. Then Blaxxun went bankrupt, IVN bought Cybertown, required a $5/month subscription, and now it's practically dead
14:06:19 <Deewiant> Alphaworld rings a bell, though.
14:06:54 <Sgeo> Deewiant, Alphaworld apparently was the old name of Active Worlds. Alphaworld is now the name of the original Active Worlds world.
14:07:55 <Deewiant> "On June 28, 1995, AlphaWorld was renamed Active Worlds" --http://en.wikipedia.org/wiki/Active_Worlds
14:48:52 -!- oklokol has joined.
15:03:11 <oklokol> it was a magnificent glio.
15:16:12 -!- oerjan has joined.
15:21:33 -!- ehird has left (?).
15:23:51 -!- oerjan has set topic: E=mc^2/sqrt(1-v^2/c^2) | http://tunes.org/~nef/logs/esoteric/?C=N;O=D.
15:24:47 -!- jix has joined.
15:25:03 -!- ehird has joined.
15:34:39 -!- okloduk has joined.
15:35:43 -!- KingOfKarlsruhe has joined.
15:46:11 -!- KingOfKarlsruhe has quit (Remote closed the connection).
15:52:29 -!- oklokol has quit (Read error: 110 (Connection timed out)).
16:17:05 -!- okloduk has quit (Read error: 113 (No route to host)).
16:25:02 -!- FireFly has joined.
16:25:58 <ehird> 12:55 fizzie: Do you have some sort of virtual reality hobby or something?
16:26:09 <ehird> clearly he was abused at a young child and this is his only method of escape
16:26:13 <ehird> </pseudopsychologist>
16:26:16 <ehird> 08:07 fungot: AnMaster: any small object that is stored in a little squeamish sometimes..." " interesting point, nobby. i rushed to its deadly brink. i think-" ( dining with a unicorn horn rule is if it ain't broke then don't fix it.
16:26:16 <fungot> ehird: but it feels like...
16:26:20 <ehird> s/sometimes/ossifrage/
16:30:14 <oerjan> fungot: your mind is going. you can feel it.
16:30:15 <fungot> oerjan: buried in the slums are clean. because of my rivals in the big bro' this job before security and the chocobo's personality.
16:30:28 <fungot> Available: agora alice darwin discworld europarl ff7* fisher ic irc lovecraft nethack pa speeches ss wp
16:30:33 <fungot> Selected style: alice (Books by Lewis Carroll)
16:30:39 <fungot> ehird: 24. because they can be marked with certainty: whereas affirmative propositions ( that is, with a tassel, and was fnord him with his umbrella. " who can have patience with a man that's got no more fnord than they, for all you have to hire the fitting on the fnord back, my little dear! it took me a full month to come," he said, with a glass in its eye, dressed in gown, bands, and wig, was fnord a pig on the charge of fnor
16:31:16 <oerjan> no wonder alice in wonderland is so scary, with all those fnords
16:32:38 -!- tombom_ has joined.
16:38:47 -!- pikhq has joined.
16:44:02 <ehird> 16:05:43 <LittleDan> eiz: how did ESR ruin the jargon file?
16:44:02 <ehird> 16:06:10 <eiz> Hang on, I'm eating here
16:44:06 <ehird> UNINTENTIONAL ICE BURN
16:45:33 <pikhq> Anyways, that's an easy question to answer. He's added things to the Jargon File that are true of him and only him.
16:45:51 <pikhq> Rather than making it a, well, Jargon File.
16:46:26 <ehird> Careful pikhq or esr will go GANDHICON ON YOUR ASS
16:46:33 <ehird> You're in for a fisking!
16:47:07 <pikhq> Ah, yes. Amtahma Gandhi -- The anti-Gandhi.
16:47:46 <pikhq> A master of agressive compliance.
16:47:48 <ehird> "Hitler, on the other hand, will be reincarnated at least dozens of times as a sensitive, easily hurt pile of flesh in a realm called Naraka and tortured in specific ways (such as having his bones and organs smashed into thousands of pieces and reassembled) for many trillions of years, but eventually, when all of the bad karma that caused his evil character has been burned away, he'll get a chance to go back to good old Earth."
16:47:51 <ehird> — http://shii.org/afterlife
16:49:02 -!- tombom has quit (Read error: 110 (Connection timed out)).
16:49:02 -!- tombom_ has changed nick to tombom.
17:14:42 <Sgeo> ehird, "(Ranked from worst to best treatment of unbelievers) "
17:15:00 <Sgeo> Why would the Atheism one be put as the most?
17:15:17 <ehird> Sgeo: It's put as the best.
17:15:39 <ehird> Sgeo: There's some bias at work here: Shii is a buddhist, so he views the snuffing out of the soul as a good thing.
17:41:06 <oklopol> this consciousness thing is getting kinda old anyway.
17:41:26 <ehird> I like conciousness :<
17:42:13 <oerjan> BRAINS.. philosophically
17:42:47 <oklopol> okokokokokokokokokokokokokokokokokoko
17:44:08 <oklopol> so what if i was a vegetarian cannibal who only ate other people's hair and nails
17:44:34 <oklopol> you can parse that in two ways
17:47:28 <oerjan> that's not vegetarian, those are still animal products
17:48:24 <oerjan> technically if you _were_ a plant, you could be a vegetarian cannibal...
17:49:00 <oklopol> but it can be vegetarian, or are you saying there's not term for ppl who don't eat meat but drink milk
17:49:54 <ehird> yeah it is vegetarian
17:50:00 <ehird> vegetarian = no meat
17:50:03 <ehird> vegan = no animulz
17:50:09 <ehird> fake vegetarian = no meat unless it's from a fish
17:50:18 <oklopol> vegetarians also often err what you said.
17:50:32 <ehird> FISH AREN'T ANIMALS
17:51:42 <oklopol> but well fish are pretty stupid, unlike most animals people eat
17:52:15 <oklopol> i'm sure most fake vegetarians would eat cockroaches too
17:52:36 <oklopol> and you know braindead cats etc
17:52:38 <oerjan> logically, we should endeavor to kill as few animals as possible per meal. so, eat more whales!
17:52:58 <oklopol> lament: that's very offending, could you kick yourself?
17:53:30 <oerjan> oklopol: would braindead women count as vegetables?
17:54:18 <oklopol> all a vegetarian cannibal would have to do is kill the brain first
17:54:36 <oklopol> ooooooooooooooooooooooooooooooo
17:54:38 <oklopol> ooooooooooooooooooooooooooooooooo
17:54:40 <oklopol> oooooooooooooooooooooooooooooooooo
17:54:40 <Sgeo> oerjan, the only thing wrong with that is that this entire conversation, as well as everything else, is distracting me from doing homework
17:54:50 <oerjan> a mad scientist vegetarian cannibal might put the brain in a jar
17:55:05 <Sgeo> Basic Writing Skills :/
17:55:10 <Sgeo> It's an essay due today
17:55:18 <oklopol> oerjan: i think the brain needs to be somewhat functional and in the body for the human to be a vegetable
17:55:35 <oklopol> Sgeo: and you where from where?
17:55:37 <oerjan> that sounds - backwards.
17:55:48 <Sgeo> oklopol, what?
17:58:14 <oklopol> oerjan: how's it backwards?
17:58:53 <oerjan> clearly the brain has to be dysfunctional
17:59:08 <oklopol> yes, that's a given, but i think it needs to be somewhat functional.
17:59:28 <oklopol> afaik they don't call dead people vegetables.
17:59:33 <oerjan> well, the life-supporting bits
18:00:10 <oklopol> well right i guess you could just have a fake brain for that, and jarrify the actual mergfh
18:00:14 <oerjan> also, a rerun: http://en.wikipedia.org/wiki/Mike_the_Headless_Chicken
18:01:38 <oklopol> this client has automatic correction, it's main use is to teach me new words.
18:02:13 <ehird> "On the first night after the decapitation Mike slept with his severed head under his wing.[1]"
18:03:27 -!- jix has quit (Read error: 110 (Connection timed out)).
18:03:59 <oklopol> lament: ban ehird he's being sentimental
18:04:25 <oerjan> oklopol: that cannot be ehird, must be his good twin
18:04:53 <Sgeo> http://www.youtube.com/watch?v=2FQh1ehArYs
18:06:27 <oklopol> oerjan: actually that's one of ehird's noticeable qualities. for instance there was this one story about a rabbit who died and ehird was all like waaa waaa that was kinda sad
18:06:39 <ehird> i was half joking you know oklopol
18:06:51 <ehird> i'm not really traumatized like 100% of the time :p
18:06:52 <oklopol> and i was like huh must be gay cuz i just laughed at the bunny and peed on it.
18:07:06 <ehird> i was crying at your terrible abuse of the peeing bunny.
18:07:24 <ehird> oklopol: maybe I was crying that we never got to see the bunny's guts
18:07:38 <ehird> exactly. so stop making assumptions, okloPEE.
18:07:40 <oklopol> well, tbh i shed some tears cuzza that
18:08:12 <oklopol> why not show the good parts?
18:10:44 <oklopol> good twin, i'm sure this has been explored somewhere...
18:12:08 <oerjan> it's just rarer than evil ones
18:12:21 <oerjan> http://tvtropes.org/pmwiki/pmwiki.php/Main/EvilTwin
18:13:09 <oklopol> isn't it quite obvious it's rarer considering it's a joke on evil twins?
18:13:18 <oklopol> at least i'm assuming it is
18:13:35 <oklopol> I'M SEEING A TVTROPES LINK.
18:13:55 <oklopol> well it crashed my firefox
18:14:00 <oerjan> i mean, don't worry, it's harmless
18:14:32 <oerjan> and i would have gotten away with it too, if not for that meddling firefox
18:14:52 <oklopol> so can i get IE on ubuntu, i'm getting kinda tired of this crap?
18:15:25 * oerjan wouldn't _know_ but doesn't think so
18:15:57 <ehird> oklopol: yes, but umm
18:17:27 <ehird> oklopol: it's your computer being shitness
18:17:38 <oklopol> ehird: ah yeah probably, i have ubuntu
18:18:28 <oklopol> anyway it's probably because of ram littleness.
18:18:42 <ehird> and cpu old/shitness
18:18:47 <ehird> and components not being frenz
18:19:10 <oklopol> well, you'd think it could like say "i can't open anymore shit" and not like just crash
18:19:13 <oerjan> or something like that, anyway
18:19:30 <ehird> oklopol: i suppose it's because nobody has a computer old enough to test on
18:20:31 <oklopol> right, because on new computers it's really hard to open enough windows to crash the browser.
18:20:52 <oklopol> well okay it is in that the gui actually becomes a bottleneck for easy usage, but anyway.
18:20:54 <ehird> well I'm not saying it's that
18:21:11 <ehird> oklopol: eh, beats windows, prolly. unless its one of the fucked up hw combinations that work well on windows and not on linux
18:21:25 <ehird> but there aren't many of those
18:21:57 <oklopol> i have never had a computer i've tried two different oses on
18:24:06 <oklopol> <ehird> oklopol: eh, beats windows, prolly. unless its one of the fucked up hw combinations that work well on windows and not on linux <<< what exactly was this in reference to?
18:24:27 <ehird> oklopol: well I used to have a computer where linux did the crashy crash and windows sailed like... well it was shit but not like linux
18:24:35 <ehird> oklopol: turned out the hardwares just hated each other
18:24:48 <ehird> and windows happened to step on their toes in a way different to linux
18:24:51 <ehird> that just happened to make them not crash
18:26:06 <oklopol> well anyway, you're absolutely right in that ubuntu and firefox would work much better if i had a more recent computer.
18:26:57 <oklopol> but they work well enough, so i'm going to keep using this thing, and just calling ubuntu a loser
18:27:02 <oklopol> flawless logic if you ask me
18:28:09 <ehird> FLAWLESS LIKE A FLOOR
18:29:39 <oklopol> 95% of teens would cry if they saw the Jonas Brothers at the top of a skyscraper about to jump. Copy and paste this EVERYWHERE if you are in the 5% that would shout 'Jump assholes
18:29:44 <oklopol> i saw this fine thing on youtube
18:29:56 <oklopol> and they were kinda gay on south park so i thought what the heck.
18:30:16 <oklopol> anyway what i was going to say was
18:30:27 <oklopol> youtube comments are kinda stupid, why do i always read them?
18:30:35 <ehird> oklopol: because an !
18:31:32 <oklopol> wait i didn't understand that
18:31:45 <oklopol> wait that's just more nonsense
18:31:50 <oklopol> what's going on in here....
18:33:03 <oerjan> oklopol: probably some unicode thing, we just see !'s but it's really an encoded jpg or something.
18:34:46 <oklopol> nuh-uh you're just wolping me :|
18:36:33 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
18:37:16 <oklopol> i have been informed if i don't attend a shoppe now, it will be too late for me.
18:39:06 * ehird invents nice Forthy control structure
18:39:26 <ehird> : NEED invert IF r> drop THEN ;
18:39:26 <ehird> : ist NEED ." T" ;
18:39:28 <ehird> : isf invert NEED ." F" ;
18:39:30 <ehird> : torf dup ist isf ;
18:39:52 <ehird> "true torf" prints T, "false torf" prints F, "anything torf" prints nuthin'
18:40:47 <ehird> have i mentioned, I love >r and r>
18:57:29 -!- oerjan has quit ("leaving").
19:12:35 <ehird> AnMaster: I just realised: 1d funge is concatenative
19:24:07 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
19:26:03 -!- MizardX has joined.
19:48:46 -!- KingOfKarlsruhe has joined.
19:56:51 <ehird> "For one, I own a netbook, a MSI Wind to be exact. 2GB Ram, 1.6Mhz Atom"
19:56:57 <ehird> ONE POINT SIX MEGAHERTZ
19:59:10 <pikhq> s/Mhz/Ghz/, I assume. Atom's slow, but not *that* slow.
19:59:31 <ehird> pikhq: 2GB of ram seems a bit excessive for 1.6ghz, too.
19:59:54 <ehird> I have 2.5GB RAM and 2.2ghz and when one struggles the other does too, so they're pretty well matched.
20:00:18 <oklopol> GregorR should really fix his client
20:03:02 <ehird> is there a way to mount /tmp in ram unless it gets big and then switch to drive
20:04:06 <GregorR> My intuitive (lame) answer would be tmpfs + mount-bind + unionfs, but there's probably a smarter/better way *shrugs*
20:04:33 <pikhq> tmpfs + unionfs with a script.
20:04:39 <pikhq> Or a custom FUSE filesystem.
20:05:19 <GregorR> Well, tmpfs + unionfs exist as-is :P
20:05:26 <ehird> i'm zipping around http://tombuntu.com/index.php/2008/09/04/four-tweaks-for-using-linux-with-solid-state-drives/ to fig out how to make my upcoming ssd worknicely
20:05:51 <ehird> lessee... noatime, tmpfs, ff cache in /tmp, noop scheduler
20:06:17 <ehird> http://thunk.org/tytso/blog/2009/02/20/aligning-filesystems-to-an-ssds-erase-block-size/ ← now how do I do this with ext3...
20:06:43 <ehird> why do people put /boot in a separate partition
20:08:01 <ehird> ahh... when will intel release TRIM support?
20:08:16 <ehird> those tweaks + TRIM support probably = love
20:08:23 <GregorR> ehird: Some bootloaders can't handle partitions that are too big, or reading data past a certain point on the partitions, or certain filesystems, so /boot is made a separate partition, at the beginning of the drive, in a supported filesystem, of a small size, etc.
20:08:34 <ehird> GregorR: ah. so it's a workaround for lilo.
20:08:53 <pikhq> It's also a workaround for BIOSes that don't support too large partitions.
20:08:54 <GregorR> I doubt highly that LILO has any of those limitations.
20:09:06 <GregorR> Yeah, it's mainly BIOSes with bootloaders that use only BIOS calls.
20:09:18 <pikhq> It's also very handy if you've got root in LVM; bootloaders don't support LVM yet. ;)
20:09:29 <ehird> I wonder how I'll get Factor on the ssd and not on it at the same time, i.e. I want work/ to be on the big hd and the system to be on the ssd
20:09:45 <ehird> ~/local/factor/ since ~/local/ = /aux/ehird-local or whatever
20:09:58 <ehird> then ln -s ~/src/factor-work/ ~/local/factor/work
20:10:27 <ehird> i really want TRIM support though
20:11:13 <AnMaster> ehird, why the noop scheduler?
20:11:21 <AnMaster> also I don't remember seeing that one for ages
20:11:50 <ehird> AnMaster: see http://tombuntu.com/index.php/2008/09/04/four-tweaks-for-using-linux-with-solid-state-drives/
20:11:59 <ehird> deadline is another option but a lkml post sez that deadline can be really slow under heavy load
20:12:52 <ehird> AnMaster: the thing is that cfq optimizes for avoiding moving the head or w/e
20:12:58 <ehird> which is meaningless on an SSD and so just wastes time
20:13:14 <AnMaster> yes cfq is best for disk of course
20:13:18 <ehird> and deadline would work if not for the unusability under load
20:13:36 * ehird examines how big his tmp folders are
20:13:45 <AnMaster> I'm pretty sure deadline was made for normal spinning disks too
20:14:09 <ehird> AnMaster: os x thang.
20:14:10 <ehird> I think /var/folders contains some big stuff
20:14:15 <ehird> So it's not really /tmp
20:14:21 <AnMaster> ehird, /var/tmp can grow big on some *nix
20:15:01 <ehird> AnMaster: apart from /var/folders that's just 2.3MB
20:15:05 <AnMaster> ehird, usually not a good idea to make it tempfs either, since fhs iirc declares it not to be temporary temporary, rather persistent temporary (not cleaned out on boot automatically)
20:15:08 <ehird> and I have no problem with losing 2.3MB RAM for performance
20:15:16 <ehird> AnMaster: I'm sure there's an fstab thing to fix that
20:15:24 <ehird> I assume the 5g doesn't happen on the linux stuff
20:15:33 <ehird> and anything under 100MB is just fine to lose :P
20:15:38 <ehird> so that'll be great
20:15:57 <AnMaster> ehird, depends. Gentoo use /var/tmp/portage for compiling packages in. If you decide to use openoffice instead of openoffice-bin I guess it could hit the 5 GB mark easily :P
20:15:57 <ehird> noatime helps a lot sez ted ts'o
20:16:10 <ehird> AnMaster: well then I'd reconfigure that
20:16:11 <AnMaster> ehird, yes noatime is good for normal drives too
20:16:27 <ehird> AnMaster: to use my big hd -- since writing 5GB to an ssd will have a quite big impact until TRIM
20:16:38 <ehird> but it'll accellerate it a bit
20:16:47 <ehird> AnMaster: ATA TRIM; do you know how ssds work?
20:16:55 <ehird> to overwrite data in a block they have to copy it, modify it, then put it all back
20:17:00 <ehird> so if you write a lot of data, it slows down
20:17:04 <AnMaster> ehird, I don't know how ssds work, but I do know how flash works. I have used cf cards
20:17:12 <ehird> with TRIM, they'd mark blocks that are just that file and nothing else
20:17:31 <ehird> thus making it go fast for way longer
20:17:37 <ehird> (even tho they're still really fast worst-case anyway)
20:17:39 <AnMaster> as in writing to a cache line without fetching it first?
20:17:51 <AnMaster> like is done for write-combining memory
20:18:09 <ehird> AnMaster: what i mean is that when you use an ssd for a while it runs out of free blocks
20:18:20 <ehird> and so has to copy a block, rewrite it and put it back most times you overwrite
20:18:24 <ehird> which is fast, but slower than just writing a new block
20:18:34 <AnMaster> hm ssd and journaled file systems wouldn't work nicely together
20:18:35 <ehird> since TRIM would mean more blocks are marked free on delete time, it'd help a lot
20:18:57 <ehird> AnMaster: the only usable SSDs are intel's and the ocz vertex
20:19:09 <ehird> AnMaster: the intel X25-M (best one): http://thunk.org/tytso/blog/2009/03/01/ssds-journaling-and-noatimerelatime/
20:19:14 <ehird> tl;dr basically no journaling overhead
20:19:27 <ehird> AnMaster: he uses ext4 there
20:19:32 <ehird> so ext3 would be more or less the same there
20:20:04 <AnMaster> ehird, anyway unless you need the access time to be accurate for files/directories I recommend noatime for *all* partitions
20:20:30 <ehird> AnMaster: I never really trust fs timestamps
20:20:38 <ehird> AnMaster: Changing them is just one 'touch' away
20:20:48 <ehird> I prefer in-file metadata when I can control the format
20:21:00 <ehird> since that needs more deliberate rewriting and can be version controlled
20:21:06 <AnMaster> well true. But unless you assume someone is malicious it can be useful for trivial stuff like auto deleting unused files in /tmp
20:21:44 <AnMaster> which is more or less needed on servers where you will have uptime in the range of half a year - a year
20:22:05 <AnMaster> depending on how often you reboot your desktop it may be needed there too
20:22:26 <AnMaster> usually init scripts clear out /tmp on reboot, and since you plan to use tmpfs for it anyway
20:22:43 <ehird> AnMaster: any idea about http://thunk.org/tytso/blog/2009/02/20/aligning-filesystems-to-an-ssds-erase-block-size/; it tells you how to do an alignment thing with ext4, can ext3 do that?
20:23:25 <ehird> AnMaster: oh, and, what fs would you recommend for /home? plain text (inc. code), document files, music, long high-definition video files (multi-gb ofc)... so really a whole range of stuff
20:23:37 <ehird> AnMaster: but not really any OS stuff; apart from e.g. binaries of programs I'm working on and what not
20:23:53 <ehird> for the os i'm pretty sure I'm going with ext3, since it's popular
20:26:28 <ehird> AnMaster: what's lvm btw
20:27:18 <AnMaster> ehird, an outdated technology, 2.6 kernels use LVM2
20:27:39 <AnMaster> ehird, it is a flexible Logical Volume Manager 2
20:27:49 <AnMaster> meaning you can resize partitions on the fly and such
20:28:00 <ehird> AnMaster: ah. so, most distros have that then?
20:28:05 <AnMaster> depending on fs you may need to unmount or not first
20:28:11 <ehird> "As it turns out, if you are using ext4, there is a way to tell the file system that it should try to align files so they match up with the RAID stripe width. (These techniques can be used for RAID disks as well)."
20:28:14 <ehird> I don't understand that
20:28:19 <AnMaster> ehird, I have yet to see a GUI frontend to LVM
20:28:22 <ehird> does it mean that aligning to raid stripe width is useful on non-raids?
20:28:28 <ehird> AnMaster: right but do distros ship with lvm
20:28:45 <AnMaster> ehird, usually the kernels support it yet. But usually it isn't the default way to install
20:28:58 <AnMaster> ehird, also you need at least /boot on a non-lvm partiton
20:29:35 <ehird> AnMaster: what about the raid-y question thing
20:29:51 <AnMaster> if you wish to skip initramfs/initrd then /sbin and /etc must be on non-lvm too
20:30:13 <AnMaster> anyway lvm (which implies lvm2 99% of the time, I was just nitpicking...) is very useful
20:30:39 <AnMaster> ehird, it doesn't make sense to me
20:31:05 <ehird> i can tell it's obviously right, I just dunno what i'm meant to do to get a simple ext3 ssd setup that's all aligned and shit :)
20:31:25 <AnMaster> why would you need aligning for non-RAID?
20:31:43 <ehird> AnMaster: 'cuz of how ssds' blocks work, see http://thunk.org/tytso/blog/2009/02/20/aligning-filesystems-to-an-ssds-erase-block-size/
20:31:55 <AnMaster> also I wouldn't go for ext3 personally, but either ext4 or xfs. But I don't know how well xfs works on SSD.
20:32:21 <AnMaster> My experience is exclusively based on rotating storage technology usage.
20:32:41 <ehird> AnMaster: ext4 blowing up on badly written apps scares me :)
20:32:51 <ehird> and I don't wanna trust that shit to an experimental fs
20:32:58 <ehird> whutz wrong with ext3 AnMaster?
20:33:05 <AnMaster> ehird, only if you crashed in the middle of that app running
20:33:34 <AnMaster> ehird, well it is a bit slow and over time it does get fragmented, and there is no defragmentation tool for it either.
20:33:46 <ehird> AnMaster: Fragmentation means nothing on SSDs,
20:33:52 <ehird> accessing any point is as fast as accessing any other.
20:34:18 <ehird> bit slow? I'd think SSD speed offsets that, but with the noop scheduler I'd expect ext3 to be... well, not slow.
20:34:27 <ehird> AnMaster: Unless you mean cfunge-optimization kinds of slow. :-)
20:34:31 <AnMaster> ehird, well you might want to tune the journal thing to only journal metadata (unless that is the default, I forgot)
20:34:47 <AnMaster> ehird, no I meant slow in a more usual sense
20:35:00 <AnMaster> ehird, yes but writing all data twice is going to wear out the disk twice as fast.
20:35:16 <ehird> AnMaster: see http://thunk.org/tytso/blog/2009/03/01/ssds-journaling-and-noatimerelatime/
20:35:21 <ehird> it seems not to matter much
20:35:26 <ehird> also, OS drives generally don't see big files
20:35:47 <ehird> "Yes, default ext3 is painfully slow, but journal_data_writeback and dir_indexes give it a steroid pump"
20:36:32 <ehird> AnMaster: is there a way to bind mount ~/.*?
20:36:35 <ehird> that is, all dotfiles
20:36:48 <ehird> that'd be useful as config files are tiny and apps wouldn't have to hit the slow mechanical HD
20:36:53 <ehird> AnMaster: can it match on something like that though
20:36:56 <ehird> instead of just a directory
20:37:11 <ehird> union mechanical and ssd to ~
20:37:15 <ehird> and put dotfiles in /aux/ehird/dotfiles
20:37:17 <AnMaster> ehird, yes something like that
20:37:24 <ehird> AnMaster: that would work if not for apps managing their own dotfiles
20:37:27 <ehird> putting them on the wrong drive
20:37:27 <AnMaster> but I don't know if it will work
20:37:48 <AnMaster> iirc unionfs can do some stuff to decide where the files will go. But I don't know to what degree.
20:38:17 <AnMaster> ehird, anyway ext4 issues were precisely because apps were changing their dotfiles a lot
20:38:28 <ehird> it'd just be nice to have apps start up without bothering the noisy, slow (yet big) mechanical monster
20:38:52 <ehird> i should only hear that thing if I open a document/media file :-P
20:39:58 <AnMaster> ehird, just get enough ram to put most stuff in cache?
20:40:09 <AnMaster> anyway you can get quiet spinning disks
20:40:16 <ehird> AnMaster: Uh, do you think 12GB is enough? :-P
20:40:23 <ehird> And yes, you can, but at the cost of speed.
20:40:49 <ehird> The drive I'm getting is 7500rpm... and it's going in a copper enclosure + the case is lined with acoustipak to dampen the noise; I really doubt I'll hear anything.
20:40:49 <AnMaster> ehird, um. You know about "acoustic management" that most modern disks have?
20:41:01 <AnMaster> use hdparm to put it in quiet mode
20:41:12 <AnMaster> since you don't care about cfunge performance level anyway
20:41:13 <ehird> AnMaster: a 5400 rpm will always be quieter than a higher one
20:41:19 <ehird> but 5400 rpm is just too slow for me
20:41:30 <ehird> AnMaster: think I should spend $10k to get 7 x 160GB SSDs in a RAID? >:D
20:41:57 <AnMaster> how large will that SSD you decided to buy be?
20:42:16 <ehird> AnMaster: It is not the cheapest drive.
20:42:27 -!- ais523 has joined.
20:43:07 <ehird> AnMaster: The prices for the 160GB Intel X25-M are like $500-$900. Really pricey drive, but the CPU, the RAM and the drive are really what matter...
20:44:14 * ehird examines his current drive.
20:44:21 <ehird> Hey, it's an Intel.
20:44:26 <pikhq> Bah. I'm planning on just spending $200 to double to quadruple almost all the stats of my system...
20:44:42 <AnMaster> ehird, "Why 250k and not 256k? I can’t tell you — sometimes the LVM tools aren’t terribly intuitive." <-- That captures the essence of LVM tools very well.
20:44:43 <ehird> pikhq: It does not help that I have an unhealthy obsession with getting this thing silent.
20:45:01 <AnMaster> but LVM is very good. Just the tools are a bit "baroque" as it says.
20:45:04 <pikhq> ehird: Oh. Well, that's going to be very expensive.
20:45:12 <ehird> pikhq: You have no idea. Take a guess.
20:45:16 <ehird> Oh, it's an i7 machine too.
20:45:23 <AnMaster> ehird, if you plan to use lvm you _will_ need to read a tutorial
20:45:42 <ehird> Say, you know you guys said you've seen fans on RAM? When would that be neccessary?
20:45:42 <AnMaster> ehird, and bookmark some reference docs!
20:45:56 <AnMaster> ehird, fast enough ram I guess.
20:46:00 <ehird> What about, say, 6 of these... http://www.mushkin.com/doc/products/memory_detail.asp?id=745
20:46:22 <AnMaster> also I don't claim to be an expert on it
20:46:26 <ehird> Would be a bitch if that'd need a fan.
20:46:37 <ehird> pikhq: Anyhoo, take a guess.
20:46:52 <AnMaster> ehird, the fact that they have some sort of heat-sinky thing (the red metal thing...
20:46:59 <AnMaster> make sure there is enough space between the ram slots
20:47:05 <ehird> AnMaster: Yeah, but that's just for 3 of them.
20:47:12 <ehird> 6 might get a lil' hot.
20:47:20 <AnMaster> ehird, read what I said instead?
20:47:34 <ehird> I'm looking at the mobo now
20:47:35 <AnMaster> and three could get hot too. I don't know.
20:47:50 <ehird> The moboroborama: http://www.endpcnoise.com/e/images/asus_p6t_deluxe.jpg
20:48:15 <ais523> stange... I've got something that looks very like a phishing email
20:48:16 <AnMaster> ehird, you need physical measurements of the RAM and the mobo and some math
20:48:21 <pikhq> ehird: $1,000 and up.
20:48:27 <ais523> but it's from and reply-to a correct address at Birmingham Universit
20:48:36 <ais523> and it's telling me a username/password, rather than asking for it
20:48:41 <ehird> ais523: forged from
20:48:45 <ais523> the rest looks very suspicious, though
20:48:53 <ehird> pikhq: That's very vague.
20:48:55 <ais523> ehird: reply-to the same address? they'd never get my reply...
20:49:01 <ehird> ais523: true enough
20:49:04 <pikhq> The point being that it's more than you really need.
20:49:14 <ais523> I know from can be forged, after all I've done it myself in the past
20:49:35 <pikhq> Given $1,000, I'd probably end up with a couple terabytes and a minimum of 8 cores...
20:49:38 <ais523> most disturbing is that they put my email in initcaps
20:49:40 <ehird> pikhq: Getting even a mid-end machine silent is expensive enough.
20:49:42 <ais523> seeing Bham.Ac.Uk is just weird
20:49:52 <ehird> pikhq: So, might as well go upwards-compatible while I'm at it.
20:49:55 <ais523> nobody writes top-level domain names like that!
20:49:56 <ehird> pikhq: Also, terabytes of what? RAM? :P
20:50:06 <AnMaster> ais523, what about the received from headers and such?
20:50:17 <pikhq> No, though I'd also have quite a bit of RAM.
20:50:25 <pikhq> I don't use hard drive space *that* much.
20:50:44 <ehird> pikhq: Mm; I'm going for a 1TB main hd. As I said, upwards compatibility; since, as I also said, silencing a mid-end machine is already up there price-wise.
20:50:55 <pikhq> (even with my habit of ripping DVDs, it's hard to fill my 500G HD)
20:51:01 <ais523> AnMaster: I suspect it'll all be internal University
20:51:16 <ehird> pikhq: Ever seen the size of a blu-ray rip?
20:51:18 <pikhq> But, hell, I guess if you've got the money to spend, might as well...
20:51:30 <AnMaster> ehird, about stripe width I don't think ext3 can do that.
20:51:32 <pikhq> Oh, right. Blu-ray.
20:51:33 <ais523> yep, it never left bham.ac.uk
20:51:43 <pikhq> That'd make it easy to fill the hard drive.
20:51:57 <ehird> pikhq: You also need ridiculous CPU to play those things.
20:52:04 <ehird> pikhq: 3GHz for a single-core machine, I read.
20:52:09 <pikhq> Or a mid-range graphics card.
20:52:11 <ais523> my guess is that it's legit, but very incompetent
20:52:21 <pikhq> ... Single-core? They still make single-core chips?
20:52:27 <ais523> luckily, it's the sort of message that I'd ignore both ways round
20:52:32 <AnMaster> <pikhq> But, hell, I guess if you've got the money to spend, might as well... <-- ... go for IBM RoadRunner?
20:52:43 <ehird> pikhq: I'm going for the Radeon HD 4850; which is probably mid-high-end or something. Only cuz it's the best one that I could find passively cooled.
20:52:47 <ais523> and wow, the HTML mart is bad
20:53:05 <ais523> <P class=MsoNormal style="MARGIN: 0cm 0cm 0pt"><SPAN style="FONT-SIZE: 10pt; FONT-FAMILY: Georgia"><?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o:p> </o:p></SPAN></P>
20:53:06 <ehird> IIRC, Deewiant said the only faster cards ATI make are <weird thing beginning with X> and the 48{70,90}.
20:53:11 <ais523> <--- a representative sample
20:53:11 <pikhq> So, you're far pickier than I am with computers.
20:53:11 <ehird> ais523: That's Word that is.
20:53:28 <pikhq> I prefer to be a generation behind. Cheaper. :p
20:53:29 <ais523> ehird: I know, I've spent hours dewording html before now
20:53:51 <ais523> the trick's to first use Tidy's auto-deworder to remove the worse bits, then a whole load of Emacs regexps
20:53:52 <AnMaster> ais523, that is out of order too. Unless <?xml> can be put at any place except the start.
20:53:59 <ehird> pikhq: More like "my current computer mostly fulfils my needs, but I get really irritated when it doesn't, and hw more powerful than you need runs high loads smoothly"
20:54:01 <ais523> well, it looks very out of order to me
20:54:12 <ais523> and that's quite a lot of effort just for one nbsp
20:54:14 <ehird> pikhq: A workload that'd max out a 2-core 2.2ghz system will be just a blip for a 4-core 3.2ghz system.
20:54:25 <ehird> pikhq: So if you max out your system quite a lot...
20:54:34 <pikhq> And I should specify, my main reason for being a generation behind is budgetary concerns.
20:54:43 <AnMaster> pikhq, man I'd love an IBM Roadrunner though. Think of how fast you could compile openoffice on it. On and how fast cfunge would run.
20:54:55 <AnMaster> actually cfunge wouldn't currently gain anything from it :/
20:54:58 <pikhq> I could get use out of much more expensive hardware if I had it, I just don't have the cash.
20:55:09 <ehird> pikhq: I'm considering hiring the guy who handled Disaster Area's finances for this.
20:55:29 <AnMaster> but some bits could be made parallel
20:55:46 <ehird> AnMaster: you could make t multi-threaded, just sync it up
20:55:53 <pikhq> I'm falling rather far behind, though.
20:56:05 <pikhq> I've got a single-core chip.
20:56:06 <ehird> * 14 USB 2.0 ports (6 ports at mid-board, 8 ports at back panel)
20:56:12 <AnMaster> ehird, that wouldn't gain anything, since it would be lock step, you couldn't even run two ips in the same tick at once.
20:56:12 <ehird> I'm trying to think of a use-case for 14 USB ports.
20:56:15 <pikhq> And my AMD processor isn't with AM2.
20:56:38 <AnMaster> pikhq, I have a Single core Sempron 3300+
20:56:43 <ehird> *Multi-language BIOS
20:56:50 <pikhq> Single core Sempron 2800+.
20:56:51 <ehird> I assume mylogo lets you replace the BIOS hugeimage, right?
20:56:54 <ehird> So set it as a totally black one.
20:57:04 <oklopol> is there a nice and simple way to get this computer's stats?
20:57:06 <ehird> You'd just get the "Press DEL" etc at the bottom of the screen.
20:57:11 <ehird> That would be the nice.
20:57:14 <AnMaster> ehird, I was using a Dell recently. And entered BIOS setup. And it asked what language I wanted
20:57:23 <ehird> AnMaster: Klingon.
20:57:26 <pikhq> And I max out my CPU regularly.
20:57:33 <pikhq> I use Gentoo, after all.
20:57:45 <AnMaster> pikhq, I use gentoo too. But I use openoffice-bin
20:57:54 <AnMaster> but with IBM Roadrunner I could compile it myself :D
20:58:13 <ehird> I'm currently using 35% of my CPU. (Up to 200%)
20:58:23 <ehird> And I've used 2.03GB out of 2.5.
20:58:35 <ehird> And I have 28GB free of my 221GB HD.
20:58:46 <ehird> Although the main culprit atm for CPU usage is Safari.
20:58:51 <ehird> It always uses like 70% of CPU.
20:59:12 <ehird> Anti-virus software (OEM version)
20:59:14 <ehird> Image-Editing Suite
20:59:18 <ehird> I've always wished my mobo came with an image-editing suite and an antivirus
20:59:20 <ehird> And a suite for artificial intelligence!
20:59:23 <pikhq> I don't have a major job running, so I'm currently using 1% of my CPU. Gleee.
21:06:23 <oklopol> anyone know a nice chess ai library?
21:11:28 -!- jix has joined.
21:14:04 <AnMaster> 22:14:00 up 5 days, 2:08, 15 users, load average: 0.25, 0.10, 0.02
21:14:59 <pikhq> 15:14:53 up 2 days, 19:49, 11 users, load average: 1.20, 1.03, 0.62
21:15:09 <pikhq> I've got lzma -9 running.
21:15:29 <pikhq> And Ratpoison & Xulrunner.
21:15:32 <pikhq> Single slower core.
21:18:31 <lifthrasiir> 5:18 up 27 days, 2:22, 7 users, load averages: 0.75 0.72 0.67
21:19:15 <lifthrasiir> 2 GHz dual core here. but i don't know why LA is somewhat high.
21:23:54 -!- sebbu2 has joined.
21:28:57 <ais523> yay, they ruled that turnitin is fair use
21:28:57 -!- sebbu3 has joined.
21:29:13 <ais523> which gives me a very good reason not to assign copyright on the things I hand in to them like the University want
21:36:10 -!- sebbu3 has quit (Read error: 60 (Operation timed out)).
21:36:40 -!- sebbu3 has joined.
21:38:21 <oklopol> no idea what you're talking about, but good for you!
21:39:13 <pikhq> TurnItIn is an anti-plagery site.
21:39:32 <pikhq> And I guess his uni wants students to assign copyright to the university.
21:39:48 -!- sebbu has quit (Connection timed out).
21:40:10 <pikhq> Which seems like something that couldn't really stand -- contracts under duress don't really stand up in court all that well.
21:40:14 -!- Judofyr has quit (Remote closed the connection).
21:40:27 <ehird> pikhq: what's xulrunner doing
21:40:34 <pikhq> Being a web browser.
21:40:49 <pikhq> Currently, I'm using it for Conkeror.
21:41:01 <pikhq> If I'm feeling more GUI-inclined, it'll be Firefox instead.
21:41:12 <ehird> why ratpoison instead of stumpwm?
21:41:25 <pikhq> Just not bothered switching to stumpwm yet.
21:43:40 -!- Judofyr has joined.
21:44:17 <pikhq> Seems that Ratpoison isn't really actively developed any more, so I'll probably switch over to stumpwm sometime soon.
21:44:59 <ehird> pikhq: You're just too faux-hardcore for floating windows. :P
21:45:40 <pikhq> Floating windows?...
21:46:01 <ehird> pikhq: Ratpoison is a tiling window manager.
21:46:10 <ehird> → No floating windows
21:46:15 -!- sebbu2 has quit (Connection timed out).
21:48:46 <ehird> http://www.endpcnoise.com/e/images/psm-5000.jpg This cooler looks like bread.
21:49:22 <pikhq> And probably costs some bread.
21:49:42 <ehird> pikhq: Well, $24 _is_ an amount of money...
21:49:53 <ehird> It's just for wimpy mini-ITX cases.
21:49:57 <ehird> "The Nexus PSM-5000 is a passive full copper SkiveTek heatsink for Intel Pentium M processor and it is perfect if you want to create a single fan solution in your psile case. The interiour design of your Psile allows you to cool your Pentium M using only the internal case fan that is inlcuded in the package of your psile case. "
21:56:12 <ehird> "Woman rams someone's car in road rage incident, floors the accelerator and spins her tires until they light her car ablaze, then refuses to get out of the car and tells everyone to "fuck off" as flames engulf the car. She was burned alive."
21:57:29 <ehird> A modern day Thích Quảng Đức!
21:58:43 <Deewiant> Or at least, I can't infer what exactly she was protesting against, if anything.
21:59:21 <ehird> Deewiant: Other cars!
22:00:03 -!- jix has quit ("Lost terminal").
22:00:11 -!- oerjan has joined.
22:01:32 -!- Judofyr has quit (Read error: 110 (Connection timed out)).
22:15:51 -!- tombom has quit ("Peace and Protection 4.22.2").
22:16:45 <ais523> yay, you can mount zip files in Ubuntu Jaunty
22:16:55 <ais523> you were supposed to be able to in earlier versions, but I never got it working
22:17:32 <ehird> ais523: do you know anything about filesystems?
22:18:49 <ais523> I assume there's some sort of zipfs or whatever on the loopback device, that's the obvious way to implement it
22:19:30 <ais523> I know a bit of the details of how ext3, FAT, and tar work
22:20:14 <ehird> just trying to figure out how to do http://thunk.org/tytso/blog/2009/02/20/aligning-filesystems-to-an-ssds-erase-block-size/ without LVM or ext4
22:20:33 <pikhq> LVM is very, very nice.
22:20:43 <ehird> pikhq: it looks scary and I don't want a separate /boot partition
22:21:09 <pikhq> It's not all that scary, and if you *insist* on not having a seperate /boot partition, check out GRUB 2.
22:21:23 <ehird> GRUB 2? AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA!!!!
22:22:32 <ehird> pikhq: That doesn't solve ext4, anyway.
22:22:54 <pikhq> Also, -E stripe-width is valid for ext2 and ext3.
22:23:04 <ehird> pikhq: really? yaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaay
22:23:09 <ehird> pikhq: but does it align to the width?
22:23:19 <ehird> ais523: what kernel is ubuntu 9.04?
22:24:01 <ais523> Linux dell 2.6.28-11-generic #42-Ubuntu SMP Fri Apr 17 01:57:59 UTC 2009 i686 GNU/Linux
22:24:17 <ais523> pretty recent kernel, actually, just 4 days old
22:24:48 <pikhq> ehird: Pretty sure.
22:25:00 <pikhq> That option is meant to make ext2 on RAID faster.
22:25:14 <ehird> ais523: ATA TRIM was committed on april 6th
22:25:18 <ehird> ais523: will that kernel have it?
22:25:54 <ehird> will a patch that recent be in that kernel
22:25:57 <Deewiant> 4 days old but it's based on a Linux much older than that!
22:26:07 <ais523> it could be, it might not be
22:26:13 <ehird> Deewiant: so, no then
22:26:14 <ais523> different patches go through the system at different speeds
22:26:18 <Deewiant> Well, the 2.6.28 would indicate that
22:26:25 <ais523> a fix for an = via == typo would go through pretty quickly, for instance
22:26:30 <ais523> a major feature addition probably wouldn't
22:26:38 <ehird> apparently the ocz vertex supports trim
22:26:55 <ehird> OCZ YOU MAY GET $$$ FROM ME IF INTEL DO NOT GET THEIR SHIT TOGETHER
22:27:53 <ehird> All the articles I've read say that the hardware could do it but the firmware doesn't support it, Deewiant
22:29:06 <ehird> 20:16 ehird: AnMaster: ATA TRIM; do you know how ssds work?
22:29:07 <ehird> 20:16 ehird: to overwrite data in a block they have to copy it, modify it, then put it all back
22:29:09 <ehird> 20:16 ehird: so if you write a lot of data, it slows down
22:29:11 <ehird> 20:17 ehird: with TRIM, they'd mark blocks that are just that file and nothing else
22:29:13 <ehird> 20:17 ehird: as free
22:29:15 <ehird> 20:17 ehird: at delete-time
22:29:32 <Deewiant> ehird: http://communities.intel.com/message/11911;jsessionid=C561787359D1A280372F19A61E8EE881
22:29:35 <ehird> 20:18 ehird: AnMaster: what i mean is that when you use an ssd for a while it runs out of free blocks
22:29:38 <ehird> 20:18 ehird: and so has to copy a block, rewrite it and put it back most times you overwrite
22:29:40 <ehird> 20:18 ehird: which is fast, but slower than just writing a new block
22:29:42 <ehird> 20:18 ehird: since TRIM would mean more blocks are marked free on delete time, it'd help a lot
22:29:50 <Deewiant> ehird: Maybe it's new, since that was posted only two days ago
22:29:51 <ais523> ehird: do you have an SSD hard drive, then?
22:29:59 <ehird> ais523: future tense
22:30:00 <Deewiant> I think he's just planning on getting one
22:30:08 <ais523> I'm pretty sure ext4 does that no matter what the filesystem
22:30:12 <ehird> Deewiant: well, that's awesome
22:30:20 <ehird> ais523: ext4 does that no matter what the filesystem?
22:30:22 <Deewiant> ehird: Intel's SSDs are awesome
22:30:25 <ais523> although obviously the optimisation isn't nearly as useful on other filesystems
22:30:33 <ais523> I mean, optimising deletion of large files
22:30:44 <ehird> ais523: Oh, no, TRIM is basically unused
22:31:08 <ehird> Deewiant: OK, do you know if you can get the kernel to do TRIM shit without compiling your own?
22:31:24 <ais523> is it a kernel module, or in main kernel, I wonder?
22:31:26 <Deewiant> No, I don't, and I don't care, for two reasons
22:31:38 <Deewiant> Firstly, I don't have such a drive AFAIK
22:31:43 <Deewiant> Secondly, I always compile my own
22:31:43 <ais523> strange, filesystems are normally in modules
22:31:53 <ehird> ais523: IT'S NOT A FILESYSTEM!
22:31:55 <ehird> It's a drive issue
22:32:03 <ais523> it's vaguely fs-related, thoguh
22:34:42 <ehird> Btw, is /tmp = /var/tmp?
22:35:11 <ais523> /tmp is wiped every reboot
22:35:16 <ais523> /var/tmp is wiped less often
22:35:16 <Deewiant> /tmp is temporary, /var/tmp is not
22:35:17 <ehird> What's /var/tmp for
22:35:24 <ehird> Untemporary temporary
22:35:25 <ehird> HURRRRRRRRRRRRRRRRR
22:35:35 <ais523> temporary stuff that you might want to persist a bit, but you don't mind the system deleting if it runs out of disk space or whatever
22:35:39 <Deewiant> It's for less temporary temporary stuff
22:35:42 <ais523> sort of like /tmp on Windows
22:35:56 <ehird> ais523: ok, so /tmp should be in tmpfs, but /var/tmp should not.
22:36:13 * ehird asks #linux about TRIM.
22:37:25 <ehird> 22:37 ivantis: What is a PHB?
22:38:27 -!- KingOfKarlsruhe has quit (Remote closed the connection).
22:41:41 <ais523> was that PHB as in pointy-haired boss, or some other expansion of the acronym?
22:42:22 <ehird> Following the lead of other open source projects, the Clojure project requires contributors to jointly assign their copyright on contributed code. The Contributor Agreement (CA) gives Rich Hickey and the contributor joint copyright interests in the code: the contributor retains copyrights while also granting those rights to Rich Hickey as the open source project sponsor.
22:42:26 <ehird> Download and print out the Contributor Agreement
22:42:29 <ehird> If you hope to contribute via Clojure's Google Code projects (clojure and clojure-contrib), specify your Google Code username on the agreement. Please specify the name/email you use on the Google Group as well.
22:42:32 <ehird> Sign the agreement
22:42:34 <ehird> Send your signed agreement via postal mail to:
22:42:40 <ehird> Pleasantville, NY 10570-0316
22:42:42 <ehird> So to contribute to Clojure...
22:42:45 <ehird> ...you have to fax or mail a copyright assignment form. To the USA.
22:43:00 <ehird> 22:42 Guest60476: how do you run bin files?
22:43:09 <ehird> It seems ##linux isn't very intelligent.
22:43:15 <ais523> chmod a+x them and type the filename into a prompt
22:43:18 <ais523> you never know, it /might/ work
22:43:32 <ais523> and if it doesn't, probably nothing else will, unless you happen to have an emulator handy
22:43:50 <ehird> ais523: in this case, it would have
22:43:56 <ehird> 22:43 Guest60476: Acd201: qt-sdk-linux-x86_64-opensource-2009.01.bin
22:44:08 <ehird> OTOH he should use a package manager
22:44:24 <AnMaster> <ehird> Untemporary temporary <ehird> HURRRRRRRRRRRRRRRRR <-- as I said before
22:44:47 <ehird> AnMaster: I was mocking the concept of untemporary temporary.
22:45:29 <AnMaster> <ehird> AnMaster: apart from /var/folders that's just 2.3MB
22:45:29 <AnMaster> <AnMaster> ehird, usually not a good idea to make it tempfs either, since fhs iirc declares it not to be temporary temporary, rather persistent temporary (not cleaned out on boot automatically)
22:45:29 <AnMaster> <ehird> and I have no problem with losing 2.3MB RAM for performance
22:45:29 <AnMaster> <ehird> AnMaster: I'm sure there's an fstab thing to fix that
22:45:43 <AnMaster> not sure what you meant with that reply though
22:45:47 <ehird> you never said /var/tmp
22:45:53 <ehird> I thought you meant setting /tmp broke that
22:46:16 <AnMaster> <AnMaster> ehird, /var/tmp can grow big on some *nix
22:46:16 <AnMaster> <ehird> AnMaster: apart from /var/folders that's just 2.3MB
22:46:16 <AnMaster> <AnMaster> ehird, usually not a good idea to make it tempfs either, since fhs iirc declares it not to be temporary temporary, rather persistent temporary (not cleaned out on boot automatically)
22:46:35 <ehird> AnMaster: You were remarkably unclear; and stop whining about it, honestly.
22:46:54 <AnMaster> ehird, I was crystal clear, but it is pointless to disucss that with you
22:47:06 <ehird> AnMaster: I was so crystal clear that I couldn't understand you.
22:48:36 <Deewiant> Hmm, I think the DMD bug I want fixed is in the 9643-line expression.c.
22:48:52 <Deewiant> That doesn't narrow it down as much as I'd hoped.
22:48:54 -!- BeholdMyGlory has quit (Remote closed the connection).
22:48:56 <AnMaster> Deewiant, I hope that file is auto generated.
22:49:03 <ehird> nothing wrong with big files.
22:49:10 <AnMaster> ehird, well it is harder to manager
22:49:19 <ehird> AnMaster: Splitting up atomic units makes no sense.
22:49:29 <AnMaster> ehird, true. But then the design is wrong.
22:49:38 <Deewiant> FWIW this isn't atomic in any sense of the word.
22:49:55 <ehird> Deewiant: Sure, but AnMaster meant it's inherently too large.
22:49:55 <AnMaster> ehird, anything over 2000 lines in a single file is hard to manage in my experience when it comes to C
22:49:56 <Deewiant> This file, as one might guess from the name, implements the semantics of every type of expression in D.
22:50:03 <AnMaster> other languages has different inherent limits
22:50:12 <ehird> AnMaster: Bullshit.
22:50:41 <ehird> I wonder what recent nvidia cards are like.
22:50:53 <AnMaster> Deewiant, it could be split into expression/foo.c expersion/bar.c and so on I guess.
22:51:41 <Deewiant> Of which 61830 are in the 46 .c files.
22:52:10 <Deewiant> Makes for a mean of 1344 lines per C file.
22:56:17 <ehird> 22:52 manpoole: i used a weird cpu temp to display 17 C would the be close to the normal tmep of a processor?
22:56:55 <ais523> interesting, amongst my updates for today is a suggested update to Finnish hyphenation in OpenOffice
22:57:03 <ais523> normally when it updates languages, it updates them all at once
22:57:42 <Deewiant> ehird: My guess is he can't read Celsius
22:58:06 <ais523> even more interesting, I'm going to have to run updates twice in a row
22:58:13 <ehird> Deewiant: Maybe he's using a pentium 386 or something
22:58:18 <ais523> one of the updates was for update-manager, which apparently isn't updating something it should
22:58:26 <ehird> Those could probably run at 17c
22:59:06 <ehird> Deewiant: What about http://upload.wikimedia.org/wikipedia/commons/0/04/KL_Intel_i386DX.jpg
22:59:59 <Deewiant> And? Doesn't mean it runs at 17 C :-P
23:00:07 <oerjan> ais523: this is finnish we're talking about. hyphenation is actually ai-complete there, so it's a separate project.
23:00:43 <ehird> **SETTING: It is 2038. A stark room filled with thousands of computers. Eliezer Yudkowsky sits at a terminal.**
23:00:58 <pikhq> I highly doubt that it runs at lower than room temperature.
23:01:01 <Deewiant> A word can be hyphenated differently depending on its meaning, without changing its spelling.
23:01:12 <pikhq> If it does, that's either impressive cooling or a cold freaking room.
23:01:18 <ehird> <AI> I am an AI more intelligent than any human. What do you demand of me?
23:01:21 <oerjan> Deewiant: you mean i was accidentally right?
23:01:28 <ehird> <Yudkowsky> We humans are in a dark age.
23:01:32 <Deewiant> oerjan: Oh, that was accidental? :-P
23:01:39 <ehird> <Yudkowsky> There is one problem we cannot solve -- important to millions of people.
23:01:46 <oerjan> i thought you knew me that well
23:01:46 <ais523> I guessed it was just oerjan oerjanising
23:01:52 <ehird> <Yudkowsky> Please help us hyphenate Finnish text.
23:01:56 <ais523> although I'm not surprised he was right by accident
23:02:22 <Deewiant> I'm struggling to come up with an example although I know I'm right.
23:02:55 <ehird> I wonder what the best way to copy multiple gbs of data really quickly between local computers is.
23:04:09 <ehird> OCZ Vertex firmware upgrades require that you wipe your drive.
23:04:15 <ehird> I wonder if this is so for the X25-M?
23:04:28 <ehird> If it is, no worries. Before upgrading I can always just copy it all to the 1TB HD.
23:04:28 <ais523> that's a bit of a ridiculous restriction, having to wipe your drive to upgrade its firmware
23:04:32 <ehird> 's only 160GB max.
23:04:59 <ais523> and I'd imagine physically putting both the HDs inside the same computer and using dd would be pretty fast
23:04:59 <oerjan> Deewiant: theoretically there's probably a norwegian example as well
23:05:02 <oklopol> <oerjan> Deewiant: you mean i was accidentally right? <<< you didn't know ä and a are a separate character with nothing in common in finnish?
23:05:04 <ais523> it's what dd was designed for as well
23:05:16 <pikhq> ehird, if you don't mind disconnecting the drive, SATA and rsync.
23:05:23 <Deewiant> Ah, "kaivosaukko". Can be either "kai-vos-auk-ko" ("opening of a mine") or "kai-vo-sauk-ko" ("well otter")
23:05:29 <ehird> pikhq: Note: it does not have to be realtime
23:05:36 <Deewiant> oerjan: Theoretically probably? :-P
23:05:38 <ehird> One option I was considering is a huge flashdrive.
23:05:44 <ehird> Copy it over bit by bit.
23:05:45 <oklopol> or wait what was being talked about
23:05:47 <ehird> USB flashdrive that is
23:06:00 <pikhq> Dumb. The network is probably faster.
23:06:26 <pikhq> Especially if you use rsync instead of dd. ;)
23:06:31 <ais523> a USB cable will almost certainly be faster than a USB cabel
23:06:34 <ehird> pikhq: Last time I copied a lot of data over ethernet, it took something like an hour to copy 11GB.
23:06:45 <ehird> ais523: usb cable = infinitiest infinity?
23:06:55 <ais523> as in the cable will be faster than the pendrive
23:06:58 <ehird> 23:05 Deewiant: Ah, "kaivosaukko". Can be either "kai-vos-auk-ko" ("opening of a mine") or "kai-vo-sauk-ko" ("well otter")
23:07:01 <ais523> because you only have to copy over USB once, not twice
23:07:12 <pikhq> And a USB flash drive is slower than that...
23:07:13 <ehird> ais523: Firewire would be faster
23:07:15 <ais523> do otters often live in wells?
23:07:20 <ehird> TOO BAD NOTHING DOES FIREWIRE
23:07:24 <Deewiant> I admit that that's a bit of a ludicrous word. :-P
23:07:26 <ehird> ais523: I thought it meant like, a healthy otter
23:07:32 <pikhq> Also, might I recommend trying to get gigabit Ethernet?
23:07:34 <ehird> But an otter living in a well is fine too!
23:07:37 <Deewiant> ehird: No, specifically the water kind of well.
23:07:38 <ais523> well, I suspect they're different words in Finnish
23:08:06 <ehird> pikhq: Too lazy to get gigabit ethernet.
23:08:24 <ehird> I mean, my 'net can't benefit from it and this is just a one-time copy.
23:08:24 <ais523> get terabit ethernet! Or preferably, petabit!
23:08:31 <ehird> (Migration of crap from imac to new box)
23:08:40 <ehird> (I don't really want to try and remove the HD for obvious reasons)
23:08:47 <ehird> (i.e. it's really fucking hard to get into an imac)
23:08:59 <ehird> JUST LIKE OS X AMIRITE GUYS
23:09:05 <pikhq> At a bare minimum, use rsync instead of dd.
23:09:10 <pikhq> It's significantly faster.
23:09:14 <ais523> pikhq: for a one-time copy?
23:09:26 <ais523> certainly it's faster the second time
23:09:32 <ehird> dd is more raw, no?
23:09:45 <pikhq> ais523: rsync's the easiest way to copy data while preserving permissions, modification time, etc.
23:09:51 <ehird> "Fortunately for us, flash based storage has access times nearly as fast as RAM" ← I thought this said nearly twice as fast as ram and was like WAT
23:10:00 <ais523> pikhq: actually, the old intended program for that was cpio
23:10:14 <ais523> but it lost a war against tar, which does something entirely different
23:10:20 <ais523> just because cpio can also do the same as tar
23:10:52 <pikhq> ehird: dd copies every fucking byte of the filesystem. Unless you've got a full filesystem, dd is one of the slowest ways to get stuff off the drive.
23:11:12 <ehird> I meant dding files, silly pikhq.
23:11:13 <pikhq> ... Or if you have a particular reason for keeping the exact same filesystem.
23:11:20 <pikhq> Oh. Well, that's just dumb.
23:11:29 <ais523> depends on how big the files are!
23:11:49 <ehird> Largest file on here... probably a few hundred meg.
23:11:50 <ais523> I wonder how fast cp -rp as root would be?
23:11:56 <pikhq> Especially if you want to preserve permissions.
23:11:59 <ehird> ais523: cp -R is slow as fuck
23:12:05 <oerjan> Deewiant: i've trouble thinking of a norwegian example which isn't equally ridiculous
23:12:08 <ais523> that doesn't surprise me
23:12:14 <pikhq> Or modification times, or ownership, or...
23:12:27 <ais523> the famous example in English is "coworker"
23:12:36 <ais523> even though cow-orker doesn't make a whole lot of sense
23:12:41 <ais523> people like saying it anyway
23:12:45 <ehird> someone who orks cows, duh.
23:12:55 <ais523> Deewiant: someone who orks
23:13:02 <oerjan> kor-sanger: choir singer, kors-anger: cross regret
23:13:03 <ehird> orking4 up, 3 down
23:13:04 <ehird> fucking. esp. in the manner and style of orcs (rough, brutal, noisy & crude)
23:13:06 <Deewiant> cow orker [Usenet] n. fortuitous typo for co-worker, widely used in Usenet, with perhaps a hint that orking cows is illegal.
23:13:07 <ehird> — Urban Dictionary
23:13:10 <ais523> I'm not entirely sure the phrase is defined, but the joke is common enough that it tends to come up a lot anyway
23:13:23 <ehird> So, cow orkers are people who fuck cows in the manner and style of orcs.
23:14:41 <ehird> http://images.anandtech.com/graphs/oczvertexfollowup_033009020937/18688.png ← Damn, the X25-M is fast.
23:14:45 <ehird> Like reaaaaaaaaaaaaaaaaaally fast.
23:15:03 -!- FireFly has quit ("Later").
23:15:20 <ehird> Wuzzat supposed to mean
23:15:30 <ehird> Kiddies aren't allowed X25-M's? :P
23:15:38 <Deewiant> It's what you said when I first pointed you to those graphs :-P
23:15:59 <ehird> Deewiant: that was when I was still blinded by reason and account balances.
23:16:10 <ehird> I have since seen the light. The platinum-plated light.
23:16:41 <ehird> "If you’ve read anything about Windows 7 Starter Edition, your first reaction was probably the same as mine: Is Microsoft nuts? This ultra-cheap edition is intended for use on netbooks, but its biggest restriction sounds like a complete deal-breaker: it only runs three applications at once. "
23:17:01 <ais523> what you didn't know the 3-app-at-once limit?
23:17:02 <Deewiant> The point of that article was, of course, that it isn't a deal-breaker.
23:17:11 <AnMaster> ais523, btw how much do you interact with fingerprint manager in IFFI?
23:17:11 <ais523> it's been discussed for years on the Interenet
23:17:11 <ehird> Deewiant: it's wrong
23:17:18 <ais523> AnMaster: not a lot, I don't think
23:17:20 <ehird> ais523: err, for windows 7?
23:17:21 <ais523> all I replace is the main loop
23:17:24 <AnMaster> I *might* have broken IFFI again. I don't think so though
23:17:29 <ais523> ehird: pretty much ever since 7 was announced
23:17:39 <ais523> XP Starter had the same restriction, but it wasn't intended for netbooks
23:17:42 <ais523> as they didn't exist at the time
23:17:56 <ehird> Netbooks are becoming more powerful, really.
23:18:01 <ais523> it was the really cheap version for selling to countries that couldn't afford anything more powerful
23:18:09 <ehird> Soon s/(notebook|laptop)/netbook/ in the lexicon.
23:18:10 <ais523> and netbooks becoming more powerful is missing the point
23:18:37 <ehird> ais523: the really cheap thing misses the point, since they just pirated it instead.
23:18:51 <ehird> "and have no trouble opening three full-featured programs as well."
23:18:58 <ehird> Why do people let Microsoft tell them this shit?
23:19:46 <ehird> 23:19 rob327: [root@hahainyourface ~]# python
23:19:46 <ehird> 23:19 rob327: Segmentation fault
23:19:50 <ehird> Yeah well, ha ha in your face.
23:21:45 <AnMaster> pathetic attempt at being funny there ehird.
23:21:53 <AnMaster> and yes I understood this happened
23:22:16 <ehird> I don't know what you're talking about, but the general structure seems to imply shit, so I'm just going to ignore it.
23:23:02 <ais523> also, why is he trying to start python as root?
23:24:26 <oklopol> mooooore extra for yooooou
23:27:05 <oerjan> happy earth day, says google
23:27:56 <ehird> 23:24 bmc: apt isnt a package format
23:27:57 <ehird> 23:24 bmc: and it does support rpm
23:27:58 <ehird> 23:25 bmc: apt is just a wrapper/frontend for dpkg or rpm
23:29:23 <oklopol> oerjan: imo earth day should be established only once there is an alternative for it.
23:29:34 <oerjan> ehird: just pretend bmc is AnMaster, then you should immediately see that it's a laughable idea that no one in their right mind would think of
23:29:54 <ehird> oerjan: hey now I use statistical correlation like all the best AIs
23:30:27 <oklopol> you can only celebrate that which is a minority
23:30:42 <AnMaster> ehird, it is a frontend for dpkg
23:30:45 <ehird> oklopol: BEING ALIVE DAY
23:30:49 <AnMaster> and the package format is *.deb
23:30:58 <oerjan> oklopol: http://familycrafts.about.com/library/spdays/bljuly20th.htm
23:31:49 <oklopol> oerjan: err what about that
23:31:58 <oerjan> oklopol: an alternative :D
23:32:14 <AnMaster> earth day? I heard about earth hour, but not earth day
23:32:16 <oklopol> by alternative i meant an alternative for earth
23:32:27 <oerjan> although no one likes the place really, it has no atmosphere
23:32:38 <oklopol> moon isn't one yet, well, will never be
23:32:56 <oklopol> because there are better ones at hand
23:33:41 * ehird twiddles thumbs while waiting for adsl modem to arrive
23:33:48 <ehird> i have a doorstop router until then :D
23:34:06 <oklopol> also even if there's an alternative for earth, earth would have to contain less than 50% of humanity in order to be celebrateable.
23:34:27 <ehird> oklopol: what about easter
23:34:39 <ehird> like 99.314159% of humanity are christians
23:34:49 <oklopol> they're not celebrating chriatianity.
23:34:52 <ehird> oklopol: jesus dying then resurrecting and going "lol j/k guys"
23:34:58 <ehird> which is not very... minority?
23:35:03 <oerjan> ehird: no, it just feels that way
23:35:08 <oklopol> has it happened many times?
23:35:13 <ehird> oklopol: good point :D
23:35:30 <oklopol> anyway you could have an earth day if it was formally about some event related to earth
23:36:01 <oklopol> but you cannot just assign a day for earth per se, until earth contains less than 50% of humanity.
23:36:29 <ehird> oklopol: what about people living on mars
23:36:31 <ehird> can they celebrate earth day
23:36:32 <AnMaster> <ehird> like 99.314159% of humanity are christians <-- No. It is is more like 25-30% or such iirc
23:36:51 <ehird> pretty sure it's obvious I was hyperbolerizing
23:37:11 <ehird> how about ‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽
23:37:19 <ehird> like 99.314159%‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽‽ of humanity are christians
23:37:20 <oklopol> ehird: hard to say, i have given this less thought than i should've.
23:37:45 <ehird> The Christian share of the world's population has stood at around 33 per cent for the last hundred years.
23:38:04 <oklopol> maybe that means they have one third right?
23:38:18 <AnMaster> ehird, ok 33% then I was slightly off
23:38:27 <oklopol> maybe god assigns portions of people to religions based on the accuracy of their beliefs.
23:38:39 <AnMaster> what a pitty it wasn't 66.6666666666666666666%
23:38:52 <ehird> oklopol: what if the majority of people become atheists
23:38:55 <ehird> is that god offing himself
23:39:14 <AnMaster> ehird, god of atheists take care of it.
23:39:16 <oklopol> why are your questions so hard
23:39:23 <oklopol> ask something simple like you know bunnies.
23:39:47 <ehird> i can attest to that, being one.
23:39:54 <AnMaster> oklopol, how are "bun" and "bunnie" related?
23:39:54 <oklopol> i was just illustrating what i meant by simple.
23:39:58 <ehird> did i just say that
23:40:21 <ais523> -y becomes -ies in most English plurals, bunny -> bunnies is one of them
23:40:21 <ehird> clearly bunnies are tasty in buns
23:40:27 <oklopol> AnMaster: did you miss all 4 of them? :D
23:40:28 <ais523> maybe it's an Easter reference?
23:40:38 <AnMaster> bunnie<half pressed down s key>
23:40:39 * ais523 wonders how rabbits ended up associated with Easter anyway
23:40:50 <oklopol> "bunnies." "bunnies?" "bunnies." "ah, bunnies." "bunnie?"
23:40:57 <AnMaster> ais523, yes, and what about eggs...
23:41:08 -!- coppro has joined.
23:41:14 <AnMaster> ais523, eggs are even more connected with easter, at least around here.
23:41:19 <oklopol> ais523: there's a south park episode about it
23:41:21 <AnMaster> doesn't make a lot of sense either.
23:41:23 <ais523> but that's less complicated
23:41:28 <ais523> oklopol: doesn't mean it's true
23:41:36 <oklopol> ais523: it's better than the truth.
23:41:40 <AnMaster> ais523, oh and witches of course. But I guess that is pretty much specific to Sweden.
23:41:51 <AnMaster> and go google it if you are interested in it,
23:41:55 <ais523> I've never heard of easter witches before
23:42:01 <ais523> witches are associated with halloween over here
23:42:36 <AnMaster> ais523, first hit on google (haven't checked accuracy with more than a glance): http://www.time.com/time/specials/packages/article/0,28804,1889922_1890008,00.html
23:42:57 <AnMaster> ais523, there is no halooween here. Well it has been imported in recent years.
23:43:39 <AnMaster> ais523, rather we have "All saints" at the same point. Same origin iirc(?)
23:43:49 <ais523> apparently it's a lot bigger in the US than in the UK
23:43:55 <ais523> and isn't all saint's the day before or after/
23:44:51 <AnMaster> ais523, maybe, I'm not religious.
23:45:08 * ehird reasks ssd q in ##linux again
23:45:24 <AnMaster> actually that times article is wrong
23:46:05 <AnMaster> "The tradition is said to come from the old belief that witches would fly to a German mountain the Thursday before Easter to cavort with Satan." <-- Not a German one. One on an island in the Baltic sea, slightly north of the much larger island "Gotland"
23:46:06 <ehird> the times is a uk newspaper
23:46:35 <ehird> "The Times was founded by John Walter on 1 January 1785 as The Daily Universal Register"
23:46:36 <ehird> that's an awesome name
23:46:40 <ehird> THE DAILY UNIVERSAL REGISTER
23:47:06 <AnMaster> but how is that related to um a newspaper...
23:47:18 <AnMaster> I mean "universal register" sounds like something that is a part of a UTM
23:47:39 <AnMaster> wait, Universal register... Would be a turing complete register
23:47:40 <ehird> AnMaster: in the old sense, i assume
23:47:46 <ehird> a registry of going-ons
23:47:53 <ehird> universal meaning that it doesn't just cover a town, I assume
23:47:57 <AnMaster> still I like my interpretation better
23:48:59 <ehird> yay, the april 20th intel x25-m firmware update doesn't wip
23:49:01 <ehird> unlike the old one
23:49:28 * AnMaster wonders how computable a single (bignum) register for both code and data would be
23:49:30 <ehird> "Bizarrely, according to the synthetic HD Tach benchmark, the new firmware slows the read and write speeds by ten per cent."
23:49:31 -!- GreaseMonkey has joined.
23:49:31 <ehird> "On the other hand, CrystalDiskMark shows that the write speed doubled from 35MB/s to 70MB/s. It was a similar story in ATTO Disk Benchmark, where the write speed also jumped forward."
23:49:43 <ehird> AnMaster: a bignum is equivalent to infinite memory
23:49:58 <AnMaster> ehird, yes but you could only operate on the whole register at once
23:50:24 <AnMaster> thus to modify data you would have to modify the code too
23:50:48 <AnMaster> I'm not sure how this would work (if at all) but I don't think I ever saw this idea before.
23:50:53 <ehird> everyone's testing the 80gb x25m and not the 160gb one
23:51:06 <AnMaster> ais523, I need your help here. ^
23:51:22 <ais523> AnMaster: there are self-modifying programming languages around
23:51:30 <ais523> it really just depends on how powerful your modification mechanisms are
23:51:36 <AnMaster> ais523, yes befunge for example...
23:51:55 <ais523> if it's just increment/decrement, and you don't have an IP or anything else separate, it's obviously non-TC
23:52:08 <AnMaster> ais523, it bit be bitwise operations only!
23:52:20 <ais523> yes, must be with just increment/decrement
23:52:26 <ais523> because you can only have one program that's an infinite loop
23:52:44 <AnMaster> ais523, err depends on how program is read
23:52:45 <ais523> and there are obviously at least two different non-terminating Turing-compatible programs
23:52:57 <AnMaster> example: stored in MSB downwards
23:53:05 <ais523> AnMaster: this is bignums
23:53:08 <ais523> they don't have an MSB
23:53:20 <AnMaster> ais523, they have a highest currently set bit though
23:53:22 <ais523> but either you either have a separate instruction pointer to record where in your program you are
23:53:40 <ais523> or the bignum needs to store the entire state, including the program location
23:54:04 <AnMaster> ais523, I think there should be no separate IP and always the 8 lowest bits will be used for selecting instruction to execute.
23:54:25 <AnMaster> and only operations are the INTERCAL bitwise ones :D
23:54:38 <AnMaster> possibly with some other absurd one added to make it TC if it isn't
23:55:00 <AnMaster> issue is I can't think of any good ones
23:55:16 <ais523> but some of the INTERCAL operations take two arguments
23:55:21 <ais523> and I seriously doubt it would be TC without mingle
23:55:37 <ehird> One thing I take consolation in:
23:55:48 <ehird> There are a few hobbies more expensive than my upgrade.
23:56:01 <ais523> why use powerful massively-upgraded computers?
23:56:07 <ais523> I've been trying to stay low-end as much as possible
23:56:10 <AnMaster> ais523, well maybe the binary operations could be implicitly using bit-interleaved numbers
23:56:11 <ais523> this laptop is more or less bottom of the range
23:56:17 <ehird> ais523: I'm upgrading because this system limits me
23:56:22 <ais523> AnMaster: that would be utterly pointless on a mingle!
23:56:27 <ais523> ehird: this system doesn't limit me!
23:56:30 <ehird> ais523: Going a bit higher in spec to make the upgrade worth it, then making it silent, costs a lot anyway.
23:56:36 <ais523> in what ways are you limited?
23:56:37 <ehird> ais523: So...; also, it may not limit you.
23:56:46 <ais523> I'm interested, rather than critical
23:56:59 <ais523> incidentally, the only resource I've really found has caused me problems in the past is disk space
23:57:18 <ehird> I run many applications at once; you tend to stay minimal and use things like w3m, I don't; I mess with large amounts of data a lot; I hate waiting for a computer to realise I've done something
23:57:23 <AnMaster> ais523, so lets say it takes the top half and the lower half. half here meaning "half of the range to the currently most significant set bit"
23:57:24 <ais523> Filesystem 1K-blocks Used Available Use% Mounted on
23:57:26 <ais523> /dev/sda6 111713482 64038606 41998712 61% /
23:57:31 <ehird> and I may as well buy something futureproof
23:57:45 <ais523> Filesystem Size Used Avail Use% Mounted on
23:57:46 <ais523> /dev/sda6 107G 62G 41G 61% /
23:58:17 <ais523> I don't really have a 100G hard-drive, do I?
23:58:19 <ais523> I thought it was only about 20
23:58:30 <AnMaster> ehird, my / is 0.7 GB and contains 230 MB data.
23:58:44 <ehird> ais523: I had an 80GB HD in the early 2000s.
23:58:51 <AnMaster> ehird, I have an old Dell with a 20 GB harddrive.
23:58:54 <ehird> ais523: Getting a new 20GB drive would be a marvel.
23:59:01 <ais523> I wonder what's taking up all the space?
23:59:03 <ehird> AnMaster: your HD is 700mb?
23:59:04 <ais523> maybe backups, I suppose
23:59:08 <AnMaster> ehird, it has a "designed for windows 98 sticker"
23:59:08 <ais523> or entire Linux distributions
23:59:17 <ehird> ais523: backups on the drive?
23:59:36 <ais523> ehird: backups of other things
23:59:43 <ais523> I have backups of things here elsewhere, obviously