00:02:09 -!- grndlvlbombdood has changed nick to bsmntbombdood.
00:04:22 <ehird> I still say that there should be an ordered directory structure in filesystems.
00:06:45 -!- jix has quit ("...").
00:15:40 <oklopol> ordered directory structure?
00:15:53 <oklopol> ehird: how did you generate it?
00:15:54 -!- Corun has quit ("This computer has gone to sleep").
00:16:01 <ehird> i didnt generate it
00:16:36 <oklopol> err right, would've been quite fast
00:18:50 -!- Corun has joined.
00:19:59 <oerjan> bah seems not to converge
00:23:54 <kerlo> This pangram contains zero a's, zero b's, zero c's, zero d's, zero e's, zero f's, zero g's, zero h's, zero i's, zero k's, zero l's, zero m's, zero n's, zero o's, zero p's, zero q's, zero s's, zero t's, zero u's, zero v's, zero w's, zero x's, zero y's, zero z's, zero apostrophes, zero commas, zero spaces, and one period.
00:25:32 <ehird> That's not very accurate.
00:40:53 <oerjan> oklopol: i didn't check for longer cycles, let me see
00:41:22 <oerjan> (btw i started with the string "abc...z")
00:41:50 <oerjan> also i used "no" rather than "zero", not that it was ever used of course
00:42:39 -!- kerlo has left (?).
01:00:35 <oerjan> "This Pangram contains four a's, one b, two c's, one d, thirty-one e's, five f's, seven g's, nine h's, thirteen i's, one j, one k, one l, two m's, nineteen n's, fourteen o's, two p's, one q, six r's, twenty-seven s's, twenty t's, three u's, five v's, seven w's, three x's, three y's, & one z." repeats after 126 steps
01:04:29 <oerjan> and does not appear until about 1008 steps after a..z
01:05:08 <oerjan> i just did not wait long enought the first times i tried
01:42:16 -!- GregorR has quit (Remote closed the connection).
01:45:31 -!- Dewio has joined.
01:57:40 -!- Dewi has quit (Read error: 101 (Network is unreachable)).
03:49:04 <lament> This pangram contains your mom.
03:51:51 <oerjan> That is not a pangram.
04:10:57 <oerjan> SPAM SPAM WONDERFUL SPAM
04:12:10 <psygnisf_> there's an algo for pangrams. what is it?
04:13:05 -!- bsmntbombdood has quit (Read error: 113 (No route to host)).
04:14:04 <oerjan> you mean self-describing pangrams? pangram alone just means something containing every letter
04:15:54 <oerjan> that would be constraint solving i think. or maybe something evolutionary...
04:15:58 <psygnisf_> i suspect its solvable with a linear equation
04:17:12 <oerjan> the number -> vector of letters in numeral mapping is complicated
04:19:07 <oerjan> by evolutionary i mean, you could use an iteration like i did, but then do a random change when you got stuck
04:20:44 <oerjan> only 16 letters actually appear in the numbers 1-99, that reduces search space
04:25:49 <oerjan> http://en.wikipedia.org/wiki/Pangram
04:26:06 <oerjan> i wonder if the example posted above was pasted from there
04:26:37 <oerjan> if so, it was computed with dedicated hardware
04:27:01 <oerjan> mind you that was in 1984 so obviously an ordinary computer should be able to do it now
04:27:27 -!- Corun has quit ("This computer has gone to sleep").
04:27:46 <oerjan> looks identical on a glance
04:28:11 <psygnisf_> its merely coincidental that some autograms can also be pangrams
04:28:24 <oerjan> yes but i guess making them also pangrams probably is not that much worse
04:30:43 <oerjan> the page speaks of Binary Decision Diagrams
04:31:00 <oerjan> now if wikipedia wasn't constantly locking up...
04:36:38 <oerjan> which probably means that they are essentially solving it by reducing to a case of an NP-complete problem...
04:37:34 <psygnisf_> i enjoy understanding what P and NP mean
04:37:57 <oerjan> but not so much hitting into them, eh?
04:38:51 <oerjan> but you don't enjoy quite as much having to solve NP-complete problems...
04:41:18 <oerjan> actually most interesting puzzles are probably human-sized (small!) versions of NP-complete problems
04:46:13 <oerjan> sudoku for example, becomes NP-complete if you have arbitrary board size and non-unique solutions
04:47:34 -!- CakeProphet has quit ("lol").
05:27:55 -!- Sgeo has joined.
05:51:46 -!- moozilla has joined.
06:04:06 -!- bsmntbombdood has joined.
06:34:00 -!- kar8nga has joined.
06:38:29 -!- CakeProphet has joined.
07:08:09 -!- kar8nga has quit (Read error: 145 (Connection timed out)).
07:12:53 -!- psygnisf_ has quit (Read error: 110 (Connection timed out)).
07:31:30 -!- Asztal has quit (Read error: 60 (Operation timed out)).
07:32:04 -!- Asztal has joined.
07:58:23 <AnMaster> <flexo> hm. looking at that mess... i'm really proud of it :) <-- optimizing?
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:07:14 -!- Sgeo has quit (Remote closed the connection).
08:14:05 -!- MizardX has quit (kornbluth.freenode.net irc.freenode.net).
08:14:05 -!- oerjan has quit (kornbluth.freenode.net irc.freenode.net).
08:14:11 -!- oerjan has joined.
08:14:11 -!- MizardX has joined.
08:19:54 -!- oerjan has quit ("Good night").
08:26:32 -!- moozilla has quit (Read error: 54 (Connection reset by peer)).
08:26:40 -!- moozilla has joined.
08:33:26 -!- Dewio has changed nick to Dewi.
08:42:53 -!- moozilla has quit (Read error: 104 (Connection reset by peer)).
08:42:54 -!- metazilla has joined.
08:56:21 -!- GreaseMonkey has quit ("e disasterous.").
10:54:35 <oklopol> that pangram problem would be quite hard to show np-complete given english numbers aren't exactly that simple
10:55:09 <oklopol> and of course it would trivially be in p if you just had the english numbers and the english alphabet, because a solution is known
10:55:20 <oklopol> a problem instance can't really be np-complete
10:56:49 <oklopol> but you could probably include subset sum somehow given the ability to specify the alphabet and the way to represent numerals
10:57:11 <oklopol> not that i can think very coherently this early
11:16:22 <Ilari> At least it becomes meaningful to talk about its algoritmic complexity if alphabet and numerics used are variable.
11:17:27 <oklopol> but i can't find a simple way to encode anything in it
11:18:01 <oklopol> i think i'm close to getting it, but i don't really have the time atm
11:18:52 <Ilari> But is that problem even in NP (i.e. it does not blow up superpolynomially)?
11:19:09 <Ilari> In solution size that is...
11:19:14 <oklopol> it's trivially in np, you can just count characters and their amounts.
11:20:04 <oklopol> given numerics, an alphabet and a solution, just check the solution contains only pieces of the alphabet, and for all characters, their amount is represented with the correct numeric
11:21:12 <oklopol> well we need to be polynomial on the size of the set of numerics
11:21:27 <oklopol> i can't see a way to get a superpolynomial solution
11:21:57 <Ilari> What could happen is similar to what happens in Sokoban: The solution chains can blow up exponetially -> PSPACE.
11:22:06 <oklopol> the solution is essentially an n-tuple of numerics, one representing each character's count in the whole tuple's numerics + 1
11:22:21 <oklopol> you don't need a solution chain
11:22:30 <oklopol> are we talking about the same thing?
11:22:31 <Ilari> Actually, the problem is not well-defined. How you encode numerics used?
11:22:50 <oklopol> i'm talking about generalizing "this pangram contains x1 a's, x2 b's..."
11:23:28 <oklopol> well just give a list of all the numerics you're allowed to use
11:23:39 <oklopol> if you need a bigger number, there's no solution
11:24:02 <Ilari> That 'solution chain' referred to Sokoban. You can encode explicit solution, but it can be exponential in lenght, so Sokoban is not in NP (its in PSPACE).
11:24:36 <oklopol> well i don't know sokoban, i can look it up
11:24:46 <oklopol> yeah what does that have to do with this?
11:25:02 <oklopol> there's no chain here, you just calculate amounts, and see if the right numeric is in place
11:25:25 <oklopol> any solution will just be an n-tuple from the given set of numerics, where n is the size of the alphabet
11:25:49 <oklopol> there cannot be anything superpolynomial about that, because you have a polynomial-size tuple of polynomial-size objects
11:26:23 <oklopol> i mean i did understand what you meant by the solution chain thing
11:27:10 <Ilari> Bound for certificate space required is n*ceil(log2(k)), where n is number of alphabet and k is number of numerics.
11:27:12 <oklopol> but here we're just checking a solution, there's no computation involved in the solution checking, as is usual for np-completeness
11:27:40 <oklopol> no actually k should be the size of the largest numeric
11:28:04 <oklopol> right, yeah, that's okay too
11:28:34 <oklopol> you're just exploiting the fact the list is in the problem descriptino
11:29:58 <Ilari> Even if numeric description size is logaritmic in numbers representable, then its still O(n*k)...
11:31:46 <oklopol> yeah it's the same as sokoban, counting the characters is the computation that can last an exponential amount
11:32:34 <oklopol> i was thinking there would be only a polynomial amount of polynomial numerics
11:32:55 <oklopol> err. then again, if you want the *number* of numerics to be polynomial
11:33:02 <oklopol> then the actual list of numerics is exponential
11:33:29 <oklopol> well. i guess it doesn't have to be
11:34:14 <oklopol> blargh, if you keep talking to me, i'm going to have to start thinking at some poitn
11:35:13 <oklopol> blah, yes you're right, might be pspace, should probably define it better
11:35:28 <Ilari> If there is some upper limit to numerics used, and representation of numerics in problem is at least logaritmic, then its in NP.
11:36:07 <oklopol> well i assumed a polynomial amount of numerics initially
11:37:18 <oklopol> basically you need the counting of characters to be in np, which that should guarantee, yes
11:38:31 <oklopol> but i never really even considered that, i assumed if there was a beautiful solution to encode something in it, you'd easily see if the result was checkable in polynomial time, and that i could define the problem after that
11:39:21 <oklopol> ah representation at least logarithmic so the character amounts can't get infinite, yes, good point
11:40:07 <oklopol> Ilari: approximately, where do you live? i'm such a patriot that i like to know that about finns
11:41:44 <Ilari> Actually, I think with logspace numerics, its not in NP.
11:41:48 <Ilari> oklopol: Helsinki.
11:42:31 <Ilari> The reason for that is that by increasing the alphabet size, you can make solution size blow up superpolynomially.
11:43:30 <Ilari> And the size of problem encoding is only logaritmic in alphabet size...
11:43:56 <Ilari> But the size of solution is linear in alphabet size -> exponential blowup.
11:44:30 <oklopol> well err, if the number of numerics is polynomial, and all their sizes are polynomial, i don't see what could go wrong. i mean there couldn't, then, be an exponential size solution because that would require an exponential input, right?
11:45:40 <oklopol> err the problem encoding also contains the numerics, so even if the solution is linear in alphabet size, it will only contain a small subset of the given set of numerics
11:46:20 <oklopol> what i mean is, you can't need an exponential amount of time to count the characters, because then you'd have an exponential amount of numerics in the input as well
11:48:22 <oklopol> i mean the crucial problem with getting an exponential blowup is we want the set of numerics as an explicit list
11:48:49 <oklopol> if we just encoded it like, say, english does, in a logarithmic amount of rules, then we'd hit the linear in alphabet size problem
11:49:24 <Ilari> The problem comes from the fact that solution size is necressarily linear in number of alphabet, but the problem description size is linear in number of bits in alphabet size.
11:52:42 <oklopol> i don't think so, if we need the numerics to be given explicitly. we can only use time O(|set of numerics|) to count characters
11:52:52 <oklopol> and if that needs to be polynomial, i don't see a problem
11:54:37 <oklopol> just like, if you're given a graph with n connected components, and you need to find the biggest subset of those components whose union size is smaller than a given number
11:55:22 <oklopol> you can just use dynamic programming, because the input size contains the numbers in "unary", just like in here
11:56:17 <Ilari> It takes exponential time just to iterate through the alphabet...
11:56:50 <oklopol> the alphabet size is polynomial, so what do you mean?
11:57:37 <Ilari> What in problem input is linear in size of alphabet?
11:58:36 <oklopol> the tuple that is the solution is linear in the size of the alphabet, it has an element for each character
11:59:20 <Ilari> Exactly. But I don't see that the input has to be linear in size of alphabet...
11:59:50 <oklopol> the checking problem requires you to, for each character (linear in size of alphabet), count the number of those characters in the numbers (linear in size of set of numberics and their representation)
12:00:04 <oklopol> Ilari: umm, so you can have an exponential input?
12:00:17 <oklopol> then, it takes exponential time to check the numbers are even in the set of numerics.
12:00:24 <oklopol> thus clearly it isn't in np.
12:00:45 <oklopol> the crucial point is how we encode the numerics, i assumed they were given explicitly, in a list.
12:03:00 <oklopol> are you perhaps implicitly assuming the set of numerics can be represented in logarithmic space?
12:03:33 <oklopol> at least if it could, you would be correct, it would not be in np
12:04:26 <oklopol> anyway, i need to go read my book, we're not really getting anywhere arguing about... well not sure what :P
13:04:46 -!- Slereah has joined.
13:13:47 -!- Slereah_ has quit (Read error: 110 (Connection timed out)).
13:38:45 -!- Corun has joined.
13:45:15 -!- jix has joined.
14:01:04 -!- jix has quit (Read error: 104 (Connection reset by peer)).
14:07:22 -!- jix has joined.
14:18:07 -!- jix has quit ("...").
14:18:28 <Slereah> http://img.secretareaofvipquality.net/src/1230134229081.jpg
14:23:10 <Slereah> It's usually an universal quantifier.
14:23:18 <Slereah> It are supposed to be a mona kittun
14:24:51 <Slereah> ASCII art stuff on the chans.
14:32:07 -!- Corun has quit ("This computer has gone to sleep").
15:09:16 -!- Corun has joined.
16:04:15 -!- ehird has set topic: i don't eat your face | man | http://tunes.org/~nef/logs/esoteric/.
16:07:57 <ehird> 06:18:28 <Slereah> http://img.secretareaofvipquality.net/src/1230134229081.jpg <-- lol
16:56:38 <ehird> I don't eat your face. Man.
16:56:42 * ehird is possibly tired.
16:57:22 <AnMaster> also what do you think about irregular webcomic today?
16:57:32 <AnMaster> ehird, I think the annotation rocked
16:58:36 <ehird> I skimmed the annotation, mostly.
16:58:40 <ehird> But I got the gist :P
16:59:02 <ehird> "You are my reader. I hate you."
16:59:28 <AnMaster> hm I would say: "black body radiation and how it can be used to figure out temperature or stars and distance to them"
16:59:41 <ehird> The intentions are what matters :-P
16:59:45 <AnMaster> also a bit about general relativity
16:59:57 <ehird> Then he has failed :P
17:00:05 * ehird predicts that the comic will now fade to octarine
17:00:10 <AnMaster> ehird, he has done such annotations a few times before
17:00:31 <AnMaster> ehird, err that isn't inside the sRGB space is it?
17:00:41 <ehird> sRGB is for squares.
17:01:11 <AnMaster> err no, triangles, though square colour spaces would be possible I think
17:01:52 <AnMaster> ehird, what colour space are apple monitors?
17:02:19 <ehird> System Preferences says "iMac".
17:02:27 <ehird> It also has "sRGB Profile" as an alternative.
17:02:30 <AnMaster> I would have assumed some wider one like AdobeRGB or such
17:02:36 <ehird> Which mainly seems to be darker.
17:02:44 <ehird> And it makes the text anti-aliasing look bad.
17:02:53 <AnMaster> well no monitor matches a colour space exactly
17:02:57 <ehird> AnMaster: it also has adobe rgb.
17:03:01 <ehird> and ntsc, and PAL.
17:03:03 <ehird> and god knows what.
17:03:10 <AnMaster> well but what can the monitor actually show?
17:03:22 <ehird> AnMaster: I'm not sure.
17:05:04 <AnMaster> I assume they exist over in UK too?
17:05:27 <ehird> I think we've only ever got one her. I want more so I can waste their time :(
17:06:34 <AnMaster> well I wonder why they keep bothering with our house after my father once (3rd time same year or so) decided to start to try to convert them about Buddhism. No visits for a few years after that, but seems they started again last year or so
17:07:27 <AnMaster> I saw it happen btw, was quite funny
17:08:06 <ehird> AnMaster: there's a way to get them stop coming
17:08:10 <ehird> i forget what the term is
17:08:19 <ehird> if you tell them you used to be one, but got excommunicated(?I think?)
17:08:30 <ehird> their strict rules are that they CANNOT come there ever again
17:08:34 <ehird> (because you're a devil or something)
17:13:57 <ehird> AnMaster: "Disfellowshipping" it seems
17:14:09 <ehird> There are over 30 violations for which a member can be disfellowshipped [21], including: Abortion, adultery, apostasy, bestiality, voluntary blood transfusions, drug abuse, drunkenness, extortion, fornication, fraud, gambling, heresy, homosexual activity, idolatry, incest, interfaith activity, loose conduct[22], manslaughter, murder, perverted sex relations[23], polygamy, pornography[24], sexual abuse, spiritism, theft, and use of tobacco.
17:14:14 <ehird> OH NO, PORNOGRAPHY
17:14:19 <AnMaster> better to try to convert them to Buddhism or such
17:14:49 <ehird> AnMaster: apparently blood is sacred.
17:14:59 <ehird> blood transfusions are strictly forbidden for jehova's witnesses
17:15:02 <ehird> even in case of emergency
17:15:59 <AnMaster> gambling huh..., so they can't play chess or anything?
17:16:30 <ehird> Arguably that's not gambling.
17:16:44 <AnMaster> ah so gambling is just when it is about money or?
17:17:14 <ehird> Gambling is the wagering of money or something of material value on an event with an uncertain outcome with the primary intent of winning additional money and/or material goods. Typically, the outcome of the wager is evident within a short period.
17:17:18 <ehird> -http://en.wikipedia.org/wiki/Gambling
17:17:19 <AnMaster> anyway trying to convert them to Buddhism really get them confused, which is great fun.
17:17:40 <AnMaster> ehird, so they don't like stock markets?
17:18:04 <AnMaster> because you could describe that as close to that
17:18:05 <ehird> Stock markets aren't gambling in public opinion, for some unfathomable reason.
17:18:22 <AnMaster> agreed you would need to bend the definition of "wager", but not much
17:19:25 <ehird> I want a new colour. :-P
17:20:14 <ehird> If a saturated green is viewed until the green receptors are fatigued and then a saturated red is viewed, a perception of red more intense than pure spectral red can be experienced. This is due to the fatigue of the green receptors and the resulting lack of their ability to desaturate the perceptual response to the output of the red receptors.[2] Kodak engineered Epcot's pavement to be a certain hue of pink so that the grass would look greener through th
17:20:17 <ehird> e reverse of this effect.[citation needed]
17:22:12 <ehird> I should compile something to llvm.
17:22:23 <ehird> Maybe a nice lil' OOP language or something.
17:22:29 <AnMaster> well why not just use the Lab colour space
17:22:41 <ehird> Is it possible to write an LLVM thing in non-C++?
17:22:50 <AnMaster> ehird, yes iirc, they have a C interface
17:23:14 <AnMaster> ehird, well ocaml, since someone used the C interface to make an ocaml interface iirc
17:23:22 <AnMaster> or was it making ocaml compile using llvm?
17:23:31 <ehird> AnMaster: Why isn't it language agnostic?
17:23:44 <AnMaster> ehird, well it could be, but you need to make an interface
17:23:56 <ehird> That's not language agnostic.
17:23:59 <AnMaster> iirc both python and perl can interface C libraries
17:24:07 <AnMaster> ehird, the bytecode spec format is open
17:24:12 <AnMaster> you could write your own generator
17:24:34 <AnMaster> ehird, but what do you suggest instead of a library with a standard interface?
17:25:18 <ehird> AnMaster: A simple plaintext format that you can just pipe to llvm?
17:25:57 <ehird> So why not use that?
17:26:17 <ehird> Well, is it more complex than the API?
17:26:36 <AnMaster> ehird, iirc the llvm website have info somewhere on it comparing pros and cons of those different ways
17:27:53 <ehird> http://llvm.org/docs/FAQ.html#langirgen
17:27:58 <ehird> against: the .ll parser is slower than the bitcode reader when interfacing to the middle end
17:28:01 <ehird> that would be especially bad
17:28:08 <ehird> as i'd like it to be fast enough for e.g. using it for a REPL
17:28:14 <ehird> eval() and suchlike.
17:28:23 <AnMaster> ehird, yes could be a problem then
17:28:34 <AnMaster> if you want to JIT stuff you probably need to use the API
17:28:50 <ehird> I guess I might have to bite the bullet and use C.
17:28:55 <ehird> Ugh, I'll have to use effing flex and yacc.
17:29:15 <ehird> Well, I could use lemon instead of yacc.
17:29:33 <AnMaster> ehird, hm can't languages like python, ruby and perl interface C libraries?
17:29:41 <AnMaster> I'm pretty sure at least python and perl can
17:30:10 <AnMaster> ehird, maybe someone already wrote a wrapper? Might be worth checking
17:30:10 <ehird> Of course, it won't fit in with the rest of the code, oh well.
17:30:59 <AnMaster> <ehird> Of course, it won't fit in with the rest of the code, oh well. <--?
17:31:12 <ehird> AnMaster: it'll take structs, etc
17:31:22 <ehird> which I'll have to construct manually from the objects
17:31:29 <AnMaster> well yes, I guess it will be easier to wrap the C API rather than the C++ one
17:31:39 <AnMaster> at least you'll only have to deal with POD then
17:32:06 <ehird> You know, writing this in a HLL might be a bit stupid.
17:32:17 <ehird> And, uh, s-expressions so I don't have to use flex/lemon.
17:32:28 <ehird> ... even though I don't really want to use s-exprs, but oh well
17:32:57 <ehird> High level language.
17:33:09 <AnMaster> what language did you plan to use to begin with?
17:33:25 <ehird> Eh, just one of the typical scripting languages. But tying a HLL to another HLL would be silly.
17:33:25 <AnMaster> iirc there is some llvm thing for haskell for example, but I may remember wrong
17:34:12 <ehird> Ooh. That could be very nice.
17:34:18 <ehird> Yes, I believe there is such a thing
17:34:30 <ehird> That is so, so tempting.
17:34:49 <AnMaster> ehird, well I don't know details, I think I was randomly browsing on ohloh or such
17:34:55 <AnMaster> or viewing suggestions for the stack there
17:35:03 <ehird> I saw it a few days ago on reddit, I think, actually!
17:35:18 <AnMaster> 24 Dec 2008 ... http://hackage.haskell.org/cgi-bin/hackage-scripts/package/llvm ... http:// hackage.haskell.org/packages/archive/llvm/0.4.1.0/llvm-0.4.1.0. ...
17:35:18 <AnMaster> aur.archlinux.org/packages.php?ID=18018 - 7k - Cached - Similar pages -
17:35:20 <ehird> http://augustss.blogspot.com/2009/01/llvm-llvm-low-level-virtual-machine-is.html
17:35:29 <ehird> Also, yeah, it's on hackage: http://hackage.haskell.org/cgi-bin/hackage-scripts/package/llvm
17:36:10 <ehird> Thing about compiling to llvm is that I have to learn someone _else's_ arcane assembly, instead of inventing my own ;-)
17:36:16 <AnMaster> ehird, oh also if you plan to mess with x86_64 and llvm use svn version
17:36:22 <AnMaster> I hit some bugs in last release
17:36:29 <ehird> I don't think I need 64-bit for this.
17:36:41 <ehird> I'm basically inspired by this: http://github.com/why/potion/tree/master
17:37:11 <AnMaster> ehird, on last release you can get llvm-gcc to ICE in stage2 of bootstrapping itself
17:37:20 <ehird> http://github.com/why/potion/commits/master <- Wow, that's some activity
17:38:08 <AnMaster> ehird, interesting, and that uses llvm?
17:38:28 <ehird> but works on non-x86, just slower
17:38:40 <AnMaster> so it uses a custom written JIT for x86 and x86_64 (it says so in README)
17:39:02 <ehird> I think, um, a few weeks.
17:39:07 <AnMaster> considering the activity level I would assume fairly new
17:39:07 <ehird> Maybe a month or two.
17:39:16 <ehird> why created repository potion 1 day ago
17:39:21 <ehird> but he's worked on it for longer, presumably
17:39:40 <AnMaster> writing a good JIT takes time, lots of time
17:40:57 <AnMaster> " * No error handling. I'm wary of just tossing
17:40:57 <AnMaster> in exceptions and feeling rather uninspired
17:40:57 <AnMaster> on the matter. Let's hear from you."
17:41:03 <ehird> * core: okay, first checkin. the parser is coming together. started this
17:41:26 <ehird> so he's developed a prototype with a working jit in less than a month.
17:41:38 <ehird> of an object-oriented language with also a VM and interesting object model.
17:41:44 <ehird> i swear that guy is paid to sit around all day and be awesome
17:42:05 <AnMaster> ehird, well I can believe someone managed to write a well working JIT in less than a month if he was dedicated, and a OO lang... But both? No way
17:42:20 <ehird> Well, there's your evidence.
17:42:23 <ehird> Browse through the commits if you want.
17:43:00 <AnMaster> well, what about work? Either rich enough to not need it or the company sponsors this, or something else
17:43:12 <ehird> See: "i swear that guy is paid to sit around all day and be awesome"
17:43:30 <ehird> AnMaster: I think what you have to realise is that he's crazy as all hell.
17:43:48 <ehird> Evidence: http://poignantguide.net/ruby/chapter-4.html (<-- this is from a _programming_ tutorial. Needs images.)
17:43:54 <AnMaster> ehird, or he is a student, one of those who back in the 1970s would have rewritten unix over the weekend
17:44:20 <ehird> More evidence: handwritten code, with coloured pencils. http://redhanded.hobix.com/cult/ahaNoticeTheExpandoWhichPrecludes.html
17:44:28 <AnMaster> ehird, well he isn't too bad at drawing
17:44:33 <AnMaster> certainly draws better than me
17:44:51 <AnMaster> ehird, WHY ARE THERE NO PUNCH CARDS!?
17:45:00 <ehird> AnMaster: don't give him ideas, man
17:45:11 <AnMaster> ehird, interesting nick btw, "why"
17:45:34 <ehird> his full name is actually why the lucky stiff.
17:45:38 <ehird> Also he has a GUI toolkit called Shoes, singular.
17:45:51 <ehird> why's new toolkit, Shoes, is great. why the lucky stiff is so awesome.
17:46:15 <AnMaster> wikipedia... "why the lucky stiff (often known simply as why or _why) is the persona of a prolific writer, cartoonist, musician, artist, and computer programmer notable for his work with the Ruby programming language."
17:47:17 <AnMaster> also the wp article seems like a link dump
17:48:37 <AnMaster> ehird, hm are you sure this is one person?
17:48:51 <AnMaster> not several performing a practical joke of some weird sort
17:48:57 <ehird> AnMaster: the theory that he is a collective has been put forward. The personality is a bit too consistent for that, though :P
17:49:48 * ehird decides to write language as custom vm language in Haskell. Maybe.
17:50:11 <AnMaster> ehird, well I'd assume only one would do the drawing bit, to make it consistent, and so on. If not he is some sort of super-productive multi-skilled genius...
17:50:43 <ehird> Alternative theory: He is a regular person who codes Java by day in a regular, messed-up javacorp. Wears a tie.
17:50:49 <ehird> Has a huge stash of LSD at home.
17:50:52 <ehird> And an internet connection.
17:51:00 <ehird> And coloured crayons.
17:51:27 <AnMaster> would LSD do that? I don't know any details of the effects caused by that drug
17:52:34 <ehird> http://blog.nihilogic.dk/2009/01/genetic-mona-lisa.html <- the javascript demoscener takes up the genetic algorithms meme
17:53:20 <AnMaster> genetic algorithms meme? When did genetic algorithms become a meme?
17:53:34 <ehird> since everyone and his dog generated the mona lisa from polygons with it :P
17:54:03 <AnMaster> ehird, didn't that one miss recombinating and so on?
17:55:46 <AnMaster> I'm pretty sure I read about the original and it seemed to do a, .... forgot the word
17:56:02 <ehird> stochastic hill climb
17:56:55 * ehird names language poke
17:57:05 <AnMaster> think I seen that used already
17:57:20 * ehird decides between C and Haskell while wondering why that's even a choice
17:57:50 <AnMaster> ehird, well I never tried the haskell interface for llvm, #llvm can be found on OFTC (not freenode)
17:58:10 <ehird> I decided against llvm :P
17:58:27 <ehird> Probably just a regular VM to start with.
17:58:31 <ehird> Y'know, sanity and all that.
17:58:47 <AnMaster> well I guess _why doesn't have any of that
17:59:33 <AnMaster> ehird, and/or he could be somewhat like those famous composers, Mozart and such, brainchilds
18:01:10 <AnMaster> maybe in 200 years people will talk about famous programmers from the twenty-first century(sp?), like we today talk about famous composers, writes and painters of previous centuries
18:01:34 <AnMaster> assuming spelling is correct for the last word
18:03:10 <ehird> "Phamus pr0gramrz aften used ''fakneames'' to hyd3 deir tru identiti. Forexam, ``why the lucky stiff`` (inth oldspeak, "wot a lucki blighter")"
18:03:24 <ehird> I disclaim all liability if English ends up like that in the future.
18:03:43 <AnMaster> ehird, I certainly hope AOL speak doesn't take over the world
18:07:31 <AnMaster> ehird, hm is there a linux C version of this genetic meme?
18:08:03 <ehird> AnMaster: http://github.com/mackstann/mona/tree
18:08:09 * AnMaster doesn't know his way around this strange blogosphere thing
18:08:09 <ehird> written by the kind of person who writes everything in C :-P
18:08:50 <AnMaster> "velociraptor.png"? Is this the xkcd author?
18:10:03 <ehird> Badger: but but you come from #haskell
18:12:30 <Badger> can't use the language though
18:12:49 <Badger> my puny attempts failed, whereas C is more comprehendable
18:13:08 <ehird> only because you're used to it
18:14:04 * ehird plots to have a component called mon in this language, so he can make horrible puns about pokemon
18:15:04 <AnMaster> Badger, what other functional languages do you know?
18:15:46 <ehird> GOTTA INTERPRET THEM ALL! umm, no.
18:16:19 <AnMaster> ehird, about that "http://redhanded.hobix.com/cult/ahaNoticeTheExpandoWhichPrecludes.html" you linked, wtf is "expando"?
18:16:44 <ehird> and the linked article
18:16:44 <ehird> http://diveintomark.org/archives/2007/04/01/document-dot-wacko
18:19:32 <ehird> i promise that poke will have an expando method on all objects.
18:19:34 <AnMaster> "“Precludes the functionality.” Awesome." <-- eh? what is odd with that? Sure, it isn't informal everyday language. So what? Yes I read the whole post, and I agree with the poster that the thing is very wtf.
18:20:23 <ehird> AnMaster: it is wtfy because it's worded ridiculously and sounds funny.
18:20:49 <AnMaster> ridiculously? You mean bureaucratically?
18:21:01 <ehird> i think you'd have to be a native speaker :P
18:21:43 <AnMaster> when I translate it mentally it ends up as highly bureaucratic, but not ridiculous
18:21:54 <AnMaster> sure sometimes bureaucratic == ridiculous
18:24:30 <AnMaster> ehird, like (real world example, not urban myth, I know the person first hand who read this in a report from someone on the local city council, a journalist btw), translated from Swedish: "green fodder converting milk production unit"
18:26:21 <AnMaster> ehird, oh and I have heard (from another journalist, my mother in fact, and no jokes about that) another person at the same city council call "windows" "light inlets"
18:26:48 * ehird thinks about writing a program that takes text and pounds it through a thesaurus to make it more beaurocratic
18:28:11 <AnMaster> about building some new building, library or museum or something, and that there should be many "light inlets" to create a nice environment and such, and when the journalists asked what he meant with "light inlets" he said something like "oh, um... you know, light inlets, um.... um... windows!"
18:28:22 -!- pikhq has joined.
18:28:38 <AnMaster> ehird, I guess it is the same in UK?
18:29:59 <AnMaster> hm and this was back in the 1980s, I wonder what they sound like now (said person no longer works as a journalist, she is teaching journalism instead)
18:31:36 * ehird attempts to hack up object.h into something that looks right
18:31:44 <ehird> (from the null string)
18:34:14 <AnMaster> ehird, oh about the "green fodder converting..." thing, the journalist in question called the person who had written it and asked, why he didn't use a normal word. And the person said that if you used "cow" it could mean either "milk cow" or "meat cow".... Yes both of those exists as direct translations to Swedish and yes he used them, no I don't think he was ever able to explain why he didn't just us
18:35:17 * ehird wonders if state is actuallyneeded in objects, with sufficient magic
18:36:38 -!- pikhq has quit (Read error: 60 (Operation timed out)).
18:37:06 <ehird> you could do it as a method definiton like
18:37:39 <ehird> hello <- { Slots at(5461651) }
18:37:44 <ehird> where { ... } is a function
18:37:49 <ehird> and the number is generated on newSlot
18:37:57 <ehird> then hello actually returns a value
18:37:59 <AnMaster> ehird, oh and something I saw myself... "park bench" as "sitting function". Sadly the translation in this case makes it even more messed up, because "function" in this case have more than one meaning
18:37:59 <ehird> problem is assigning it
18:44:24 -!- pikhq has joined.
18:44:42 -!- Hiato has joined.
18:48:48 <Hiato> Ok, let me ask a seemingly worthless question: Which distro: GoboLinux, Zenwalk, Lunar, SliTaz or Puppy (or something light/small too)?
18:49:26 <AnMaster> it all depends on what you want
18:49:41 <AnMaster> I use Arch Linux and Gentoo Linux. ehird here said GoboLinux was good iirc
18:49:48 <AnMaster> and I like both Gentoo and Arch
18:50:05 <ehird> Hiato: slackware? i dunno.
18:50:07 <ehird> gobolinux is nice.
18:50:16 <ehird> but not that lightweight
18:50:22 <Hiato> AnMaster: Well, a stable learning/messing platform that is extensible, but doesn't go bandwidth crazy like Ubuntu. This has to fit in <700mb's
18:50:50 <Hiato> Oh, and it has to be something different, from the usuals, FC, RHL, SuSE, Deb etc
18:51:08 <Hiato> Ok, I'll take a look at Arch
18:51:22 <Deewiant> If installing the core system from the internet instead of the CD, half of that
18:51:51 <ehird> don't look at arch :-P
18:52:02 <Hiato> That sounds awesome. Thing is, I'm not really even sure what I use an OS for anyway. I have a Windoze partition, mainly for Fallout 3, and then that's it. I do python and ruby, etc, so nothing constricting there
18:52:12 <ehird> Hiato: if you want to learn how linux works
18:52:16 <Hiato> Heh, well, I meant *look* at, as in read :P
18:52:33 <ehird> Hiato: i mean, slackware is like one level up from Linux From Scratch
18:52:50 <Hiato> I have a vague idea, but I just wanna be able to mess around. Yeah, so I've heard, hence I included zenwalk in the list
18:52:56 <ehird> the installer is basically an interface over a program that copies over a kernel and runs the (barely-extant) package manager
18:53:03 <ehird> see, that can be a good thing
18:53:08 <ehird> if you want to get proficient with linux, that's great
18:53:14 <ehird> because you learn it without a glossing over
18:53:23 <ehird> it's also very small and fast sort of feeling, in general
18:53:40 <Hiato> Hrmm, so, BlueWhit64? I have an AMD
18:54:05 <ehird> I wouldn't go for a really obscure distro, it'd just be painful
18:54:19 <ehird> Hiato: what do you want to do with this?
18:54:43 <Hiato> ehird: not sure, and that's 90% of the problem. I would say use it, but I'm not even sure what that means.
18:54:52 <ehird> what primary uses? :-)
18:55:14 <Hiato> On windoze, gaming, on linux/misc: surfing, programming and stuffing about
18:55:38 <ehird> Hiato: is learning linux to a deeper extent one of your goals?
18:56:18 <Hiato> ehird: Most certainly, but, not necessarily a priority. (PS: You should make a web-wizard distro selector, you have all the right questions)
18:56:50 <ehird> Hiato: lol, there are plenty of distro selectors to go around... But, if you have a lot of spare time and don't mind messing about a lot to get things going, Slackware could be a good learning experience.
18:57:02 <ehird> You might also want to check out one of the *BSDs.
18:57:26 <ehird> For something more "everyday" but still not a mainstream distro, not sure. Damn Small Linux?
18:58:04 <Deewiant> I'd say Arch is fine as an easier Slackware
18:58:24 <ehird> Deewiant: except that's a misguided goal
18:58:41 <ehird> either you want to get familiar with linux, i.e. slackware
18:58:43 <Hiato> Meh, I'm skeptical about BSD, as I am about solaris/derivatives and other funny kernerls/os's. I tried DSL, and although it was Debian, it failed. Not my kind of thing.. leaves you way in the cold - I felt. I mean, I've never really felt comfortable in Linux, ever, so Arch is looking good and so is slackware. GoboLinux is out because it's too non-standard
18:58:47 <ehird> or you want something for every-day usage, i.e. something else entirely
18:59:07 <Deewiant> Why couldn't one become familiar with linux with an every-day distro
18:59:09 <ehird> Hiato: BSD actually predates linux, by a long shot
18:59:17 <ehird> Deewiant: You could. But if it's one of your specific goals...
18:59:32 <Deewiant> I think the familiarity comes naturally with use
18:59:51 <ehird> slackware can still be valuable, it really depends on what you want
18:59:52 <Hiato> Deewiant: Could, but where's the fun? ehird: Yep :P Read up a bit (a lot) in the past
18:59:54 <Deewiant> Hell, I think I get /too/ familiar just trying to get rid of warnings and errors in dmesg and then in Xorg.0.log ;-)
19:00:27 <ehird> Hiato: Slackware might "leave you out in the cold" to start with -- though its interactive installer is helpful.
19:00:36 <Deewiant> Hiato: Well, there's the offset in that an everyday distro is more fun to use when you're not specifically tinkering and messing about in order to learn linux ;-)
19:00:40 <ehird> Well, interactive is a stretch: Basically, you run it and it tells you what to do next and you have to figure out how ;-)
19:00:46 <ehird> Treasurehunt installation!
19:00:54 <ehird> Well, it does give you pointers.
19:01:19 <Hiato> ehird: :P Deewiant: yeah, I realise, which is why I'm leaning towards Arch now... no offense ehird
19:01:36 <ehird> i still think arch is misguided
19:01:58 <ehird> there's two poles, down-dirty and high-level... arch tries to be the latter by starting from the former
19:02:03 <ehird> which is kind of silly
19:03:01 <Hiato> Hrmm... ok, so fundamentally, technically, it's not cool. Deewiant, do you use it? PS: it does have a much catchier description: a lightweight and flexible Linux® distribution that tries to Keep It Simple.
19:03:29 <ehird> disregard all of Deewiant's opinions, plz
19:03:50 <Deewiant> Less effort to keep running Gentoo than to reinstall :-P
19:03:53 <Hiato> I see, well, I'm not in the mood for massive source compiling/downloading, so no gentoo/bintoo/saboyan etc
19:04:56 <Deewiant> I still like its USE flags idea, being able to set features on a per-package basis, leaving out what you don't need... but being able to install stuff in seconds instead of minutes or hours does outweigh that benefit :-P
19:05:42 <ehird> Deewiant: macports has variants, essentially use flags
19:05:46 <ehird> but less fine-grained, which is fine by me
19:06:03 <ehird> of course, that's uncool, i mean.
19:06:07 <Deewiant> I did flip between Arch and Gentoo on the laptop, though... something didn't work in Arch (I think it was the frame buffer settings) but did in Gentoo for some reason so I went with Gentoo
19:06:15 <Deewiant> I wasn't in the mood for figuring stuff out
19:06:53 <Deewiant> ehird: How does that work? Basically multiple different packages which provide the same virtual package or something like that?
19:07:07 <Hiato> Well, not that it really makes a difference, but according to DW: Slackware:14 Arch:16
19:07:24 <ehird> if you want popularity, why are you looking for something obscure :P
19:07:32 <ehird> slackware is one of the oldest distros fwiw
19:07:46 <Deewiant> I believe it's the second-oldest still active one, after Debian
19:07:48 <ehird> few months older than debian
19:07:59 <Hiato> ehird: :P I know, just wanted to see what was on the web :P
19:08:00 <ehird> slackware = 16 July 1993
19:08:06 <ehird> debian = 16 August 1993
19:08:19 <Deewiant> I distinctly seem to remember being annoyed at the fact that slackware wasn't older than debian
19:08:44 <Deewiant> Maybe it was just at debian's age then
19:08:53 <ehird> well, slackware was just a barely-modified SLS to start with it seems
19:08:57 <ehird> so arguably it originated in 1992
19:09:17 <Deewiant> Well yeah, arguably Ubuntu originated in 1993 ;-P
19:09:39 <Hiato> er, what is going on here, where's the frinedly iso? ftp://ftp.is.co.za/mirror/ftp.slackware.com/pub/slackware/
19:09:52 <Deewiant> Slackware, friendly? Surely you jest
19:09:54 <ehird> you didn't specify friendly as a requirement.
19:10:03 <ehird> maybe wrong directory
19:10:22 <fizzie> Yes; go one directory level upwards.
19:10:29 <fizzie> And take slackware-12.2-iso or something.
19:10:36 <fizzie> "isolinux" is just the boot system thingie.
19:10:52 <Deewiant> Haven't used slackware in a long while, and never installed it myself
19:11:01 <ehird> Hiato: http://www.slackware.com/install/
19:11:10 <ehird> might be worth printing that out / keeping it open on a separate machine.
19:11:38 <Hiato> er, ah, now that's a problem. Slackware>700mb's
19:12:13 <fizzie> You can make a small slackware installation, probably with the new versions too.
19:12:17 <fizzie> Just don't install everything.
19:12:28 <Deewiant> You might not need all the isos
19:12:30 <ehird> Hiato: why is that a problem? The actual install won't be that bg
19:12:38 <ehird> if you disable most things
19:13:00 <Hiato> It's the download that's the problem. I can only use 700mb's before the 20th or bust
19:13:17 <fizzie> I installed slackware 3.2 back in 1997 or so when it came with a Finnish computer magazine; and their installation CD was borken. Package descriptions were missing for a couple of categories, had to choose what to install based on package names only. And it was my first Linux experience. That wasn't friendly.
19:13:18 <Deewiant> Hmh, why doesn't irssi send a message when I press keypad enter
19:13:29 <ehird> Hiato: want me to send you a disk in the post? XD
19:13:43 <ehird> fizzie: I'm sure you said that in 2005 or something
19:13:48 * ehird = obsessive log reader
19:13:52 <fizzie> ehird: Probably; I tend to repeat myself.
19:13:58 <Hiato> ehird: I would love it, here's a mackerel in return
19:14:07 <ehird> Hiato: ooh! Legal B Nomic tender!
19:16:07 <Hiato> back to my [non] life-altering decision, I can comprimise and take Zenwalk, or do as the Deewaint does and take Arch. Is it worth the compromise?
19:16:40 <ehird> Slackware, download it via carrier pigeon
19:16:56 <Hiato> ehird: we have bad reception
19:17:07 <ehird> Hiato: thus carrier pigeon. Here's some documentation: http://www.rfc-editor.org/rfc/rfc1149.txt
19:17:54 <Hiato> XD :D and any combination of the letters "l", "o", "r" and "f"
19:18:13 <ehird> Deewiant: re - macports variants
19:18:13 <ehird> vim @7.2.065 (editors)
19:18:14 <ehird> Variants: athena, big, cscope, gtk1, gtk2, huge, motif, nls, perl,
19:18:16 <ehird> puredarwin, python, ruby, small, tcl, tiny, universal,
19:18:23 <Hiato> with each octet separated by whitestuff and blackstuff. ?
19:18:27 <ehird> the tcl portfile defining it just has "variant foo { ... }" blocks that override the global definition
19:20:48 <Deewiant> ehird: Wait, macports installs from source?
19:20:58 <ehird> Deewiant: yes, irritatingly
19:21:02 <ehird> but, it's fast, generally
19:21:17 <ehird> Deewiant: it could be done in a binary package manager
19:21:18 <Deewiant> Why bitch at me for using Gentoo then :-P
19:21:26 <ehird> instead of tweaking cflags, it'd just point to a different binary tarball
19:21:29 <Deewiant> ehird: Yes, it could, which is why I was asking for details on how it works
19:21:41 <ehird> combining variants would be non-trivial
19:21:57 <ehird> Deewiant: gentoo has more of an attitude problem regarding it than macports ;-)
19:22:26 <Deewiant> Meh, I don't humanize distros, I don't care about their attitude :-P
19:22:45 <ehird> hey, who wants to buy me a pc that I can write an x11 WM on for no reason>???????
19:22:47 <Hiato> ok, after checking forums, Zenwalk it is *conversation has taken carrier pidgen over my head*
19:23:03 <ehird> Hiato: a pox be upon you
19:23:19 <fizzie> MacPorts isn't even that fast on my one-gigahurtz G4 PPC; I remember doing a "port upgrade outdated" after a year or two of inactivity and it took several hours.
19:23:47 <ehird> 2ghz intel core 2 duo >>> 1ghz G4 PPC, I'd assume
19:24:11 <Hiato> Deewiant: well, it's the fact that Zenwalk forums respond nicely to questions like this:
19:24:12 <Hiato> Hello, I'm COMPLETELY new to Linux. So I have a few questions about it xD
19:24:12 <Hiato> If I install Zenwalk, will I lose the programs I had installed on Windows?
19:24:12 <Hiato> Also will I lose my files? :S sorry, rather dumb questions ^_^;
19:24:12 <Hiato> Ehird: ... :( I tried
19:24:12 <fizzie> Probably, but I don't have such EVIL CAPITALIST MONEY-HUGGER hardware.
19:24:22 <ehird> Hiato: yes, you will.
19:24:27 <ehird> unless you have them on a separate partition.
19:24:31 <ehird> or, you know. back them up
19:24:41 <Hiato> ehird: that's not my question...
19:24:54 <Hiato> O_o I can't believe you believed that was me
19:24:57 <ehird> 19:24 <Hiato> Also will I lose my files? :S sorry, rather dumb questions ^_^;
19:25:06 <Hiato> http://support.zenwalk.org/viewtopic.php?f=46&t=9346, somewhere there
19:25:18 <Deewiant> Hiato: Arch forums don't, then?
19:25:30 <ehird> i think i just indirectly insulted Hiato's intelligence by mistake. ouch
19:25:45 <Hiato> haven't checked them
19:25:49 <Hiato> all is forgiven :P
19:26:01 <ehird> Hiato: thanks, idiot. Wait, disregard that <_<
19:26:02 <fizzie> Are you sure you want to associate with people who aren't all "AH HA HA YOU STUPEF STUPIDO" on the Interwebs?
19:26:31 <fizzie> They sound... abnormal.
19:26:31 <Hiato> fizzie: very true, and highly valid
19:26:43 <Hiato> let me ask you then: Zenwalk vs Arch
19:26:54 <ehird> " Also, spaces and symbols like '*' can confuse the Linux inner workings ('kernel'). "
19:27:00 <ehird> Issue with zenwalk: they're all retards.
19:27:12 <Hiato> .. fizzie: I'll tryu again :P ehird: lol, heh
19:28:00 <Hiato> THINK..AND READ BEFORE UPGRADING - amuch better topic in the Newb forums
19:28:40 <fizzie> Anyway, I don't know anything about either distro, so I can't really comment. (But Debian's netinst .iso is 145 megs, and contains the base system; then you can choose what other packages you want to download; so at least it fulfills that one arbitrary criterion.)
19:28:50 -!- Corun has quit ("This computer has gone to sleep").
19:29:24 * Hiato whips out his big bag o distros
19:30:16 <ehird> Hiato: try linux from scratch
19:30:23 <ehird> in case you're suicidal
19:30:36 <ehird> and, like, want to perform one last senseless, pointless act before leaving this earth.
19:30:41 <ehird> then you should try lfs.
19:31:07 <fizzie> Oh, and if you feel like it, you can do Slackware too; just avoid the full .isos, pick up a suitable boot disk and the network.dsk image, then install over the network. It will not be friendly at all, though.
19:31:38 <Hiato> I have tried, and not really liked: Ubuntu x.xx, DSL, Austrami, PC-Linux {Mini-me}, Mach boot, SLED 10, SliTaz, TinyMe, Mepis, FC, RHL, Mandrake, Mandirva and a bunch of others
19:31:56 <ehird> what's your main issue with them?
19:33:05 <Hiato> I just felt, well, I couldn't mess around, or that I wasn't supposed to. When I broke things, they magically healed and I was left none the wiser. Everything's a wizard (not necessarily bad), but get's frustrating.
19:33:26 <Hiato> fizze: will keep that in mind
19:33:51 <fizzie> Well, LFS is certainly something where you're supposed to mess around.
19:34:12 <Hiato> If it's any indicator, I liked Puppy - and still do. It's simple, stuff works, you fix what you break, the essentials are all there, but the rest can be done (with relative ease)
19:34:14 <ehird> Hiato: slackware would be pretty much perfect then
19:34:25 <ehird> when you mess up slackware, it doesn't work until you find out what's wrong and you fix it :-P
19:34:35 <ehird> note: that is simultaneously very annoying
19:34:58 <Hiato> fizzie: understood, on both notes. ehird: yeah, but it's better in a snese
19:35:18 <ehird> Deewiant: "i did something, this something was incorrect, the results are not correct"
19:35:26 <Hiato> Deewaint: change device0 to device1 or mess with other stuff in xorg.conf, just to see what it does
19:35:27 <Deewiant> I've yet to use a system which wasn't like that :-P
19:35:53 <Hiato> for example. Or change/stop start-up scipts
19:35:53 <ehird> Deewiant: yes, but it's usually at a higher level
19:36:06 <Hiato> mess with the HAL, etc etc
19:36:53 <Deewiant> How can stuff possibly work automatically if you break a start-up script :-P
19:37:04 <Hiato> that's how I (and most of humanity) learn. Break stuff, look at the pieces, do it again and then work out what it is supposed to do and why it works
19:38:04 <fizzie> Deewiant: A package manager might automatically fix it when upgrading stuff.
19:38:04 <ehird> Hiato: which would you prefer - everything is a wizard... or nothing is a wizard?
19:38:12 <ehird> that's Mandrake vs Slackware :-p
19:38:30 <Hiato> ehird: can't I float in between, with a slow drift to the nothing side
19:38:44 <Hiato> But yes, that sounds like a fair comparison
19:38:49 <ehird> that is not an answer :-P
19:38:51 <fizzie> I'd prefer an IRC Wizard. All this free-form writing is so tiresome; with a wizard interface I'd just have to click "Next" every now and then.
19:38:52 <Hiato> and I'd go with the nothing one
19:39:28 <Deewiant> Hmm, it seems that switching to Marvell's sk98lin driver and setting Moderation=Static as an option has made my machine work even whilst downloading torrents, hoorays \o/
19:39:30 <ehird> fizzie: forming grammatical sentences is so HARD, i'd prefer to just be given options... whenever I get just one phrase wrong i get asked what i meant
19:39:37 <Hiato> blarg, well, it looks like with my ridiculous restrictions, I'm going to have to go for Arch or bust
19:40:04 <Hiato> good distro name there
19:40:08 <fizzie> Sounds not entirely work-safe.
19:40:35 <Deewiant> Symptoms previously included: all processes using about 6000-10000% the CPU time even after the torrent client was closed
19:40:51 <Hiato> 'We guarantee to break something, or your sourcecode back'
19:43:56 <fizzie> There's a curious network driver thing, also: the r8169 in an older kernel used to generate a ginormous number of "dropped" packets.
19:44:03 <fizzie> Ifconfig output: "RX packets:67776111 errors:0 dropped:423893525723034 overruns:0 frame:0"
19:44:18 <fizzie> And that number increased by something like 100 million packets every second.
19:44:27 <fizzie> This 2.6.28 doesn't seem to do that any more.
19:44:58 <Deewiant> The annoying thing about this is that sk98lin is a third-party driver I have to get from marvell.com :-/
19:49:19 <fizzie> Oh well; I have to use the horrible nvidia binary blob driver too, since the second monitor I have is pivotable but the open-source 'nv' driver disables all (2D too) acceleration if I want a rotated image.
19:50:12 <Deewiant> This one's not a binary blob; it comes with an install.sh which patches the linux source tree (and unsuccessfully—had to manually correct a broken Kconfig file for 'make menuconfig' to run)
19:50:43 <Deewiant> Which is in some ways even more annoying because it's a bit trickier to install than just a blob.
19:51:26 <ehird> did Deewiant just use a — on IRC?
19:51:48 <Deewiant> I was using the fglrx driver but now I've switched to radeonhd (no acceleration supported for my card model!) because fglrx would lock up the whole machine if I accidentally restarted X
19:51:50 <fizzie> Oh, the nvidia driver also has a horrible installation script which touches all kinds of places. I just use the Debian-packaged version, I'm not so interested in having the newest released ones.
19:52:30 <Deewiant> sk98lin doesn't have an Arch PKGBUILD, I was thinking of making one but I guess that'd require me to duplicate much of the install.sh logic myself and it'd probably be too much work to keep up-to-date
19:52:41 <fizzie> And Lexmark's Linux drivers also had a horrible installation shell script. And so did some vmware product. It seems to be a habit. I usually just read them and do the steps manually, since I don't feel comfortable running them.
19:53:41 <Deewiant> Yay, about 90 minutes of torrenting and 'time man man >/dev/null' still completes in less than 0.1 seconds
19:54:07 <Deewiant> (As opposed to 5-7 seconds previously.)
19:55:38 <Deewiant> Hmm, what the hell happened to google's favicon
19:56:00 <Deewiant> http://googlesystem.blogspot.com/2009/01/new-google-favicon.html
19:56:52 <ehird> fire whoever did this
19:57:22 <Deewiant> I was fine with the previous change though the old one is the best, but this one sucks :-P
19:59:07 <fizzie> Lexmark's linux drivers also include a full copy of Sun's Java 1.4.2 runtime. Probably some configuration dialog or something was done as a Java application.
20:00:24 -!- alex89ru has joined.
20:00:50 <fizzie> An exclamation of disgust.
20:00:52 <fizzie> I may need to start running a local proxy here just so I can override the google.com/favicon.ico URL.
20:01:25 <ehird> just use greasemonkey
20:01:37 <Deewiant> Or https://addons.mozilla.org/en-US/firefox/addon/3176
20:01:42 <fizzie> Then I'd need to do that for all installations.
20:01:50 <Deewiant> Or no, it seems that's only for bookmarks
20:02:12 <ehird> Or use w3m and forget about retarded graphics
20:02:53 <fizzie> There's an about:config entry called browser.chrome.favicons in Firefox.
20:03:05 <fizzie> Based on the name, toggling that true -> false might make them all go away.
20:03:33 <ehird> I will do that right after I switch to linux and use a 1000 sloc WM.
20:03:34 <Deewiant> See also browser.chrome.site_icons
20:03:51 <fizzie> Yes, I'm seeing the kb.mozillazone.org description of that right now.
20:05:11 <ehird> doesn't hide the icon tho
20:05:13 <ehird> jusut always default
20:05:26 <fizzie> The site_icons preference will hide them.
20:06:38 <fizzie> Oh. Documentation says it will. The favicons one is defined to force the favicon of all sites to the default, but not affect the displaying. (Or the icon displayed for image files.)
20:08:15 <fizzie> The documentation seems to be a filthy liar.
20:10:43 <ehird> mice are so ergonomically terrible
20:12:03 -!- psygnisfive has joined.
20:18:39 -!- alex89ru has quit (Remote closed the connection).
20:22:47 -!- Hiato has quit ("Leaving.").
21:05:34 -!- Corun has joined.
21:59:18 -!- oerjan has joined.
22:26:27 <AnMaster> <ehird> don't look at arch :-P <-- why not?
22:26:56 <AnMaster> <Hiato> ehird: Most certainly, but, not necessarily a priority. (PS: You should make a web-wizard distro selector, you have all the right questions) <-- that exists
22:27:18 <AnMaster> http://www.zegeniestudios.net/ldc/
22:27:35 <AnMaster> it did suggest arch and gentoo for me :)
22:27:57 * ehird notes that AnMaster buffers his output and never reads input while calculating & sending output.
22:28:09 <AnMaster> also slackware is painful, I know, I used it
22:28:13 <fizzie> One of those "which OS are you?" questionnaires also had at least a couple of Linux distributions in there.
22:28:23 <ehird> Depends on your definition of worst.
22:28:33 <ehird> Well, what Deewiant said.
22:28:34 <Deewiant> it cannot be the worst because it doesn't exist
22:28:38 <ehird> It doesn't manage packages, it just installs them :P
22:28:39 <AnMaster> at least linux from scratch doesn't pretend to have a working package manager
22:29:03 <AnMaster> Deewiant, well it has packages
22:29:08 <ehird> Deewiant: it has a script that takes a tarball and runs ./configure && make && make install
22:29:13 <fizzie> I used slackware for a couple of years without touching the package manager. It wasn't very much advertised.
22:29:19 <ehird> not a package manager, it doesn't purport to be one
22:29:20 <AnMaster> ehird, not slackware no, I remember installing it
22:29:24 <ehird> and it's not exactly advertised
22:29:41 <AnMaster> same for some other packages I downloaded
22:29:59 <Deewiant> AnMaster: btw, if you've got a Befunge-93 interpreter handy run on latest (november) mycology and let me know if any BADs ensue
22:30:05 <ehird> All it does is unpack a tarball to /
22:30:10 <ehird> then run doinst.sh
22:30:14 <ehird> well, install/doinst.sh
22:30:29 <ehird> and it can tell you what's in ionstall/slack-desc
22:30:50 <ehird> Then it can remove a "package" that it's installed. And upgrade, which is really a special case of remove&install.
22:30:52 <AnMaster> Deewiant, I don't have any verified one no, cfunge has a 93 mode, which changes space rules and so on, but won't actually forbid 98 instructions in general
22:31:03 <ehird> Deewiant: i'm talking to AnMaster
22:31:08 <AnMaster> I have been thinking of making that compile time mode for extra speed ;P
22:31:09 <ehird> i.e., it's not a package manager.
22:31:17 <AnMaster> with possibly greater emulation then
22:31:18 <Deewiant> ehird: You directed a line at me
22:31:29 <ehird> yeah that was ages ago :P
22:31:45 <Deewiant> ehird: Yeah, and I responded to it when I noticed it :-P
22:31:46 <AnMaster> ehird, also that counts as package manager without dependency handling
22:31:54 <ehird> AnMaster: it doesn't, though
22:32:06 <ehird> your definition of package manager is wrong
22:32:16 <AnMaster> and I wouldn't consider dep tracking to be important
22:32:20 <ehird> it's only the worst package manager because your brain has the definition package manager = dependency-tracking etc
22:32:24 <flexo> working towards the ballmer peak now
22:32:35 <ehird> minimalism isn't for everyone
22:32:48 <fizzie> I did that web-test and it suggested me Debian and Ubuntu. I currently use Debian and Ubuntu. Clearly it must be infallible, based on my sample of N=1 persons.
22:33:06 <ehird> AnMaster: A package manager. Your definition is just wrong.
22:33:22 <AnMaster> ehird, hm? I said I didn't include dep tracking in package manager
22:33:34 <ehird> Fine, then it isn't for you.
22:34:03 <AnMaster> ehird, indeed, so I stick with something slightly more advanced while still no silly gui config stuff
22:34:14 <AnMaster> ehird, also I have installed LFS
22:34:18 <ehird> So why is it the worst package manager you've ever used?
22:34:21 <ehird> You haven't explained that yet.
22:35:23 <Deewiant> That webtest gave me gentoo and, interestingly enough, slackware.
22:35:23 <AnMaster> ehird, 1) it lacks the features that are RECOMMENDED, optional but silly to exclude (IMO) 2) it managed to loose track of files, this was around 5 years ago, slackware 9 or 10 I think, it may be better now
22:36:18 <ehird> the distribution chooser is _heavily_ biased to the mainstream distros.
22:36:26 <ehird> But, umm, I seem to get every single one.
22:36:40 <ehird> Arch, Kubuntu, Fedora, Mandriva, Slackware, OpenSuSE.
22:36:52 <ehird> I guess "just show all distros" always gets the right one, but it's not exactly clever...
22:37:18 <fizzie> Well, I only got two out of it.
22:37:32 <AnMaster> ehird, that differs, I get two, Arch and Gentoo
22:37:59 <AnMaster> actually I don't get Arch if I select that I have a 64-bit CPU, which is odd, Arch have x86_64 support nowdays
22:38:05 <ehird> I think it's because I answer "I don't care" to quite a few questions because, well, I don't.
22:38:10 <Deewiant> Ah, 64-bit might make a difference.
22:38:22 <fizzie> I have a feeling I sort-of answered the questions so that I'd get Debian (and coincidentally Ubuntu) because I happen to be happy with this stuff.
22:38:52 <Deewiant> Silly if it assumes I want a 64-bit system (which is the case, but still)
22:39:11 * ehird tweaks answers to try and get better results.
22:39:25 <fizzie> Uh... is there a way to move backwards in the test?
22:39:29 <AnMaster> for x86_64 it suggests gentoo and slackware
22:39:33 <Deewiant> Now it gave me Slackware+Mandriva+Kubuntu+OpenSUSE+Ubuntu+Arch
22:39:37 <AnMaster> fizzie, seems no, you could just restart it
22:39:59 <Deewiant> I think I added one more "don't care" than last time
22:40:04 <fizzie> That's a lot of questions to answer if I just want to change one.
22:40:06 <Deewiant> To the development packages one
22:40:11 <ehird> Okay, by lying through my teeth I get another huge pile. XD
22:40:22 <Deewiant> No, I don't /need/ development packages
22:40:33 <Deewiant> I don't /need/ X packages for all X because I'm smart enough to get em elsewhere
22:40:47 * ehird tweaks answers further
22:40:48 <AnMaster> Deewiant, anyway on x86/x86_64 you generally want a 64-bit distro, because a) multilib means you can run 32-bit just fine b) more registers more than compensates for increased pointer size
22:40:50 <Deewiant> But it probably thinks "aha you don't /want/ them"
22:40:59 <Deewiant> AnMaster: I don't run multilib, I run a 32-bit chroot
22:41:15 <AnMaster> Deewiant, well both works the same way as far as the kernel cares
22:41:16 * ehird lies through some more teeth.
22:41:16 <Deewiant> b) is a tradeoff; CPU for memory+disk
22:41:26 <ehird> C'mon, gimme slackware.
22:41:29 <ehird> Dammit you stupid machine.
22:41:33 <ehird> A huge bag is not a good result
22:41:33 <Deewiant> And it depends on the program anyway.
22:41:41 <AnMaster> Deewiant, double register count more than compensates for double pointer size
22:41:49 <AnMaster> in more than half of the cases
22:42:05 <ehird> One thing I really like about slackware is that it doesn't mess with 3rd-party apps.
22:42:06 <Deewiant> Yes, you are emphatically agreeing with what I said. :-P
22:42:13 <ehird> Debian is an especially bad offender there
22:42:17 <fizzie> Heh, now I get Fedora, Ubuntu, Gentoo, Arch, Debian, Slackware if I tell it I don't care about the package management (that question is a bit silly, too, since a couple of answers are specific package formats and not about the abilities of the package management) and don't choose the 64-bit option.
22:42:24 <AnMaster> Deewiant, an -m32 build of cfunge with 32-bit cells is about 5-10% slower on this computer, nothing else differs in build
22:42:33 <AnMaster> for 64-bit cells the difference is around 70% faster
22:43:00 <AnMaster> that is 70% faster for 64-bit build
22:43:03 <Deewiant> Yeah, the package management question was really stupid as well.
22:43:36 <ehird> I should write my own chooser that always gives either Slackware or Ubuntu.
22:43:37 <AnMaster> it was just one of the *least* stupid wizard style linux selector thingies I knew
22:44:05 <AnMaster> and it did give the right answers for me, that is the ones I ended up with after testing lots and lots of distros before
22:44:38 <Deewiant> http://www.tuxs.org/chooser/ for the win
22:44:46 * ehird answers hardcore-style
22:44:59 <ehird> I bet answering "Expert" to "How would you rate your technical skills?" makes it mark you as lower than "Beginner"
22:45:11 * AnMaster used red hat (back when it wasn't fedora, red hat 6.0...), OpenSuSE (don't remember version), Slackware, Debian before ending up with Gentoo and Arch
22:45:16 <AnMaster> Ubuntu was hardly around back then
22:45:21 <Deewiant> I answered "expert" simply because I know many professionals that don't know jack shit
22:45:30 <ehird> If your PC is fairly new and you are looking for a more technical distribution to install on your hard drive then try
22:45:31 <ehird> Debian or Slackware
22:45:40 <ehird> If your PC is a few years old and you are looking for a more technical distribution to install on your hard drive then try
22:45:41 <ehird> Debian or Slackware
22:45:52 <ehird> Umm, gee, that's some variation.
22:45:57 <Deewiant> ehird: Note that it says "more info on Gentoo or Slackware"
22:46:04 <AnMaster> <ehird> I bet answering "Expert" to "How would you rate your technical skills?" makes it mark you as lower than "Beginner" <-- well I *know* I'm expert, since I'm able to help even with the tricker questions in ##linux a lot of the time
22:46:31 <ehird> AnMaster: that's pretty arrogant...
22:46:35 <Deewiant> http://distrogue.awardspace.com/ was nice
22:46:41 <Deewiant> It gave me "Perfect match!" but nothing else
22:46:46 <ehird> Deewiant: awardspace? srsly? XD
22:46:50 <AnMaster> ehird, and I successfully cross compiled a hardened LFS, for which there is no guide :P
22:47:12 <Deewiant> Wow, you did something for which a guide did not exist
22:47:12 <ehird> [[AnMaster's ego shoots off the charts]]
22:47:12 <AnMaster> there is a hardened guide and a cross compiling one
22:47:18 <AnMaster> compiling them isn't that easy
22:47:26 <ehird> Deewiant: i know, I need guides _all the time_
22:47:33 <ehird> that shit's hard, man
22:47:49 <Deewiant> Breathing guide: http://www.breathingmatters.com/bodybm.htm
22:48:10 <ehird> AnMaster: ok, but how do I combine this with the eating guide?????
22:48:18 <ehird> combining guides is _really hard_
22:48:25 <AnMaster> ehird, depends on what version of the eating guide
22:48:40 <AnMaster> ehird, also a hardened cross compiling toolchain is quite complex in fact
22:48:48 <ehird> 4. Do you need a 3D desktop? <- Fuck off, distro chooser XD
22:48:56 <AnMaster> ehird, it never asked me that?
22:48:58 <ehird> 5. Will you be using your system for gaming? <- excuse me, isn't this for Linux?
22:49:04 <ehird> AnMaster: deewiant linked to http://distrogue.awardspace.com/
22:49:06 <fizzie> "Here are the results: Perfect match!" Heh, that's funny.
22:49:22 <AnMaster> <ehird> 4. Do you need a 3D desktop? <- Fuck off, distro chooser XD <-- heck I need to *NOT* have oen
22:49:23 <Deewiant> http://desktoplinuxathome.com/distro.html seems okay
22:49:31 <ehird> fizzie: also "This page has been viewed times."
22:49:44 <ehird> I got a lot of results
22:49:51 <Deewiant> fizzie: Do you have cookies enabled?
22:50:12 <fizzie> Deewiant: Nope. That might be it.
22:50:26 <Deewiant> I was hoping "Yes" so I wouldn't have to do it again :-P
22:51:01 <ehird> http://desktoplinuxathome.com/distro.html, my results in order:
22:51:02 <AnMaster> <ehird> AnMaster: deewiant linked to http://distrogue.awardspace.com/ <-- it suggests FreeBSD, PC-BSD, SabayonLinux, Pardus, Gentoo, Arch, "Frugalware", Debian, Fedora, openSUSE, mandriva and more
22:51:14 <ehird> debian, slackware, vector(???????????????????????????????????????????????????), gentoo
22:51:16 <ehird> then it goes on to 3 stars
22:51:20 <ehird> see the lines beklow
22:51:24 <ehird> it lists the problem
22:51:29 <ehird> (it lists every distro and then whether it matched)
22:51:38 <ehird> http://www.vectorlinux.com/ wtf.
22:51:40 <Deewiant> Gives me Gentoo,Arch,Frugalware,FreeBSD-Stable,PC-BSD as "perfect matches"
22:52:34 <AnMaster> <ehird> http://www.vectorlinux.com/ wtf. <-- shiny!
22:52:46 <ehird> http://polishlinux.org/choose/quiz/
22:52:52 <ehird> good questions, good results
22:53:00 <ehird> I got {Gentoo,{Free,Open,Net}BSD,Slackware}
22:53:19 <Deewiant> http://desktoplinuxathome.com/distro.html gave me Ubuntu,Mandriva,Vector,Fedora,Gentoo,Mepis as ones with more than 3 smileys
22:53:20 <ehird> out of which I'd consider using {NetBSD,Slackware}, but I can understand why it gave the others
22:53:27 <AnMaster> 3. How important stability and maturity is for you?
22:53:36 <ehird> it's from a polish site
22:53:38 <ehird> give em some slack
22:53:45 <ehird> AnMaster: it's a continuum from a lot to not at al
22:53:46 <Deewiant> I hate the way saying that I have a new computer causes them to give me bloatware
22:53:57 <ehird> 22:53 <AnMaster> ehird, ...ware <-- Wut>?
22:54:05 <AnMaster> <ehird> give em some slack <AnMaster> ehird, ...ware
22:54:13 <Deewiant> ,,,,,,,,,,,,,
22:54:20 <ehird> someone else try the polish one
22:54:22 <Deewiant> polishlinux.org times out for me
22:54:37 <fizzie> Polish says: FreeBSD, OpenBSD, Debian, Gentoo, NetBSD.
22:54:38 <Deewiant> http://polishlinux.org.nyud.net/choose/quiz/ seems to work
22:54:41 <fizzie> It seems very BSD-friendly.
22:54:53 <ehird> I'd just ignore the BSD results, tbh
22:55:20 <AnMaster> it seems to contrast like this:
22:55:41 <AnMaster> I have seen binary and hard to use
22:55:42 <ehird> if you read it, it doesn't :P
22:55:55 <AnMaster> it's a crucial thing for me - I often test different software and don't have the time to play with compiling from sources, looking for dependencies and so on
22:55:56 <AnMaster> quite much - easy installation is a big plus; still I can compile and build a package once in a while
22:55:56 <AnMaster> not much - nice package management system temps, but I usually prefer to prepare packages myself so I know what exact functionality I get
22:56:04 <AnMaster> I want easy and source, with fine level of control
22:56:10 <ehird> you can just as easily read "build" as "make own package"
22:56:18 <AnMaster> some system to set features I want, then have the system to build them for me
22:56:38 <ehird> that's an irrelevant detail
22:56:53 <AnMaster> and then it should be easy to use, track all deps
22:57:07 <ehird> you seem to have a penchant for ignoring the bigger picture and concentrating on minutae...
22:57:15 <Deewiant> Oh darn, the nyud.net one sends the results to the non-cached one anyway
22:57:26 <ehird> nyud.net couldn't handle POST
22:57:31 <ehird> it caches per-URL...
22:57:45 <AnMaster> is what the polish one suggested
22:57:50 <Deewiant> And the main site still times out for me
22:57:58 <ehird> lol i wonder if you can get it to NOT suggest bsd
22:58:04 <ehird> Deewiant: use a proxy?
22:58:33 <AnMaster> "Why are you going to try Linux? " (http://desktoplinuxathome.com/distro.html)
22:58:42 <ehird> AnMaster: leave it blank
22:58:45 <AnMaster> "I'm already a Linux user" is missing
22:59:06 <Asztal> Debian, Fedora, KateOS, Gentoo, Ubuntu
22:59:17 <Deewiant> Wonder why that site doesn't work from here
22:59:18 <Asztal> I like that it has Gentoo and Ubuntu in the same list
22:59:20 <ehird> kateos looks polish
22:59:25 <ehird> which would make sense considering the site
22:59:32 <ehird> http://www.sabayonlinux.org/ <-- worst distro ever or worst distro ever
22:59:33 <AnMaster> ok it suggests ubuntu first, then gentoo, mandriva, vector, Mepis, Slackware, Fedora
22:59:41 <ehird> AnMaster: yeah it sucks :P
22:59:49 <ehird> there are distros worse than ubuntu
22:59:51 <fizzie> ehird: "for desktop, not a programmer" + "newbie" + "no time for learning" + nothing else concerns me => no BSDs.
22:59:56 <ehird> case in point: mandrake
22:59:58 <AnMaster> ehird, well opensuse comes close
23:00:06 <AnMaster> ehird, I haven't used mandriva/mandrake
23:00:11 <ehird> opensuse isn't bad
23:00:16 <ehird> not my cup of tea, but not bad
23:00:33 <fizzie> (Actually that sort of settings give Mandriva, openSUSE, Aurox, Xandros, Fedora.)
23:00:39 <AnMaster> anyway I used quite a few of them, and that chooser is the worst so far
23:00:57 <ehird> polishlinux has been the best for me
23:01:00 <ehird> the questions aren't stupid
23:01:25 <fizzie> AnMaster: The Polish one.
23:01:28 <AnMaster> oh I never heard of this vectorlinux before
23:01:39 <AnMaster> fizzie, the polish one wasn't too bad, except the language
23:01:55 <AnMaster> maybe it isn't in it's database
23:03:26 <AnMaster> anyway personally I have found the distros I like, gentoo (primary desktop); arch (low end [Pentium 3 or lower] headless development computer); FreeBSD (remote server in datacenter); OpenBSD (dedicated firewalls)
23:03:30 <Deewiant> At last: Gentoo, FreeBSD, Arch, Debian, PLD
23:03:44 <Deewiant> http://polishlinux.org/linux/pld/
23:03:59 <ehird> PLD took some best features from couple of other distributions: RPM format from Red Hat (but PLD uses RPM-s capabilities much more efficiently),
23:04:00 <AnMaster> it didn't recommend that to me
23:04:13 <AnMaster> ehird, I thought RPM was a bug
23:04:18 <ehird> I was about to say that
23:04:22 <ehird> to hell with both of you
23:04:34 * ehird considers filing a bug report for every RPM-based distro
23:04:35 <fizzie> There seem to be quite many Polish distributions in there; surprisingly.
23:04:38 <AnMaster> ehird, at least it was back when I used red hat 6.0... Oh the nostalgia! RPM hell!
23:04:42 <ehird> fizzie: or, rather, not surprisingly
23:04:49 <fizzie> I hadn't heard of Aurox either.
23:04:53 <ehird> It's a polish site :-P
23:05:16 <AnMaster> anyway that vectorlinux thing, anyone ever heard of it before?
23:05:29 <AnMaster> yes that is the right word, shiny
23:05:35 <AnMaster> Deewiant, you heard of it before
23:05:37 <fizzie> It's the best; the site says so.
23:05:40 <AnMaster> ehird, yes that was what I said :P
23:05:42 <ehird> AnMaster: "eye candy", more like
23:06:16 <AnMaster> for example for some reason I don't know the buttons on my mobile phone are shiny
23:06:18 <Deewiant> I recall it being said that Vector was lightweight, but it might have been compared to Ubuntu or something
23:06:21 <AnMaster> you see all the fingerprints on it
23:06:37 <fizzie> But VectorLinux has a DELUXE version!
23:06:39 * ehird considers doing LFS then killing himself
23:07:00 <AnMaster> ehird, there is some package manager somewhat like checkinstall
23:07:12 <AnMaster> anyway made for recording packages for LFS
23:07:33 <AnMaster> it is in their "hints" section
23:08:10 <ehird> I'm pretty sure that "Go into chroot, run make install, record the differences to a file, then extract from chroot" is pretty trivial to write.
23:08:23 <ehird> I mean, as these things go.
23:08:26 <AnMaster> ehird, I think it was using LD_PRELOAD trick
23:09:12 <ehird> I'd prefer a chroot :-P
23:09:17 <ehird> 's what macports uses
23:09:31 <AnMaster> ehird, well how do you know the tools you need to install are in there
23:09:50 <AnMaster> some packages might need obscure tools to install
23:09:59 <ehird> AnMaster: that's why you copy the whole system to a chroot.
23:10:00 <psygnisfive> besides a british dialects way of saying truth :p
23:10:11 <AnMaster> ehird, well that will take a while. 4 GB or whatever
23:10:21 <ehird> AnMaster: do whatever macports does, then.
23:10:29 <AnMaster> ehird, I would use chroot + unionfs
23:10:31 <ehird> AnMaster: also, stop trying to teach psygnisfive. hopeless case.
23:10:42 <ehird> also, that fails when the package modifies existing files
23:10:51 <ehird> oh, copy-on-write?
23:11:05 <ehird> okay, that would work splendidly then
23:11:11 <ehird> and be easy to implement
23:11:31 <AnMaster> ehird, and then you check what was changed in the overlaying one, need to handle deleted files (unionfs write specially "whiteout" files for that iirc)
23:11:50 <AnMaster> unionfs isn't in vanilla kernel
23:11:53 -!- oerjan has quit ("leaving").
23:11:54 <psygnisfive> ehird, just telling me it was a shell command makes was good enough :P
23:12:02 <AnMaster> this means you may need to wait for the patchset to be ported
23:12:07 <ehird> who cares about a vanilla kernel when you're hypothesising your own distro
23:12:12 <fizzie> You can also use a LVM snapshot volume mounted in the chroot; that's copy-on-write. Although looking at changes is then a bit more difficult, with unionfs you nicely get the changed/deleted files easily.
23:12:23 <AnMaster> psygnisfive, chroot isn't a shell command, it is a system call
23:12:39 <AnMaster> ehird, because you can't track last sources, you need to wait for it to be ported to last
23:12:48 <AnMaster> which means you can't follow bleeding edge
23:12:59 <fizzie> "man chroot" does give here the 'shell command' -- well, it's not a built-in, but anyway -- instead of the system call, though.
23:12:59 <AnMaster> fizzie, yes that would work nicely
23:13:04 <ehird> who honestly cares about a bleeding edge kernel?
23:13:30 <AnMaster> ehird, me? at least on my development box, on my desktop I run more stable
23:13:34 <fizzie> Anyone with bleeding edge hardware. :p
23:13:37 <ehird> does it really matter though?
23:13:47 <AnMaster> ehird, yes because I want to test new features
23:13:47 <ehird> as long as it's relatively new, does it bother you?
23:13:57 <psygnisfive> why are we talking about hardware and kernels
23:14:07 <ehird> psygnisfive: because we want to, go away if you don't like it.
23:14:16 <AnMaster> ehird, which is why it ran gcc 4.3.2 one week after it was released
23:14:17 <fizzie> And 2.6.28 did really add useful things. Someone xtables-ized the iptables "recent" module, for example, so my SSH knocking thing works for IPv6 now, too.
23:14:21 <ehird> AnMaster: i guess it falls under the "all programmers are fashion-obsessed nerds with ADHD" axiom
23:14:26 <psygnisfive> you should be talking about a kernel written in BF!
23:14:34 <AnMaster> ehird, you are a programmer too?
23:14:48 <AnMaster> ehird, you have ADHD? Would explain a lot :P
23:14:56 <ehird> That was a metaphor. :P
23:15:04 <ehird> Just like how by fashion I didn't mean clothes.
23:15:18 <ehird> if all 13 year olds have ADHD it's hardly ADHD, is it
23:15:28 <ehird> i mean, for something to be an actual... thing it has to be non-norm
23:15:54 <psygnisfive> attention-deficit hyperactivity disposition. :P
23:15:59 <AnMaster> ehird, I never care for clothes. Simple classical model blue jeans that aren't pre-worn-out, and t-shirt, and due to climate something thicker over
23:16:12 <ehird> and AnMaster successfully derails the conversation to an irrelevant topic
23:16:13 <AnMaster> like this thing made of fleece
23:16:20 <ehird> only to have everyone else pull it back again
23:17:23 <AnMaster> ehird, anyway you derailed it by starting talking about clothes!
23:17:36 <ehird> no, I didn't talk about clothes
23:17:47 <AnMaster> <ehird> Just like how by fashion I didn't mean clothes.
23:17:53 <ehird> that's not talking about clothes
23:18:01 <psygnisfive> i googled lumberjack for a humorous picture
23:18:04 <AnMaster> ehird, that is meta-talking about not talking about clothes
23:18:04 <psygnisfive> and this is what i find: http://i37.photobucket.com/albums/e98/bolson51/lumberjack.jpg
23:18:14 <ehird> AnMaster: you are confusing mentioning and changing the topic to
23:18:37 <ehird> 23:15 <EvilTerran> @@ @run words "sleep eat haskell idle" !! (pred.read.last.words$(@show @dice 1d4))
23:18:37 <ehird> 23:15 <lambdabot> "haskell"
23:19:14 <AnMaster> bbl, going to listen to an one hour radio program :)
23:22:31 <AnMaster> psygnisfive, an (one hour) radio program
23:22:52 <AnMaster> and "one" begins with vowel sound, thus "an"
23:23:14 <psygnisfive> but its not written that way for historical reasons.
23:23:29 <AnMaster> psygnisfive, doesn't that depend on dialect?
23:24:09 <psygnisfive> besides, "an" is not universally used just before vowels
23:24:16 <ehird> AnMaster: absolutely not
23:24:28 <ehird> an one hour sounds definitively wrong
23:24:41 <psygnisfive> for instance, its used because the "y" glide as in "a usually red hat"
23:24:52 <AnMaster> psygnisfive, as I mentioned my nick are from my initials
23:36:31 <ehird> 23:31 <oklopol> read that as "sleep eat haskell die"
23:36:32 <ehird> 23:31 <oklopol> i was like "wow, what a gambler!"
23:36:35 <ehird> those lines seemed familiar
23:36:39 <ehird> but then i realised oklopol said them
23:37:24 <ehird> just like I knew who would have said tem
23:37:34 <oklopol> well yeah, that's not a surprise
23:59:37 -!- KingOfKarlsruhe has joined.