00:00:17 <AnMaster> that should narrow the search field down
00:01:23 <AnMaster> "Lives in Las Vegas, 39 years or older, name of Mike Riley, lived in Switzerland some years ago (in or around Zurich)"
00:01:30 <AnMaster> pikhq, can't match too many people now can it?
00:02:06 <AnMaster> pikhq, alternatively you could contact the police. This sounds like a worse option to me though.
00:02:44 <pikhq> Kinda hard to call long-distance, anyways...
00:02:53 <pikhq> Not great phone service.
00:03:14 <AnMaster> pikhq, you are our only hope</bad taste>
00:16:05 -!- soupdragon has joined.
00:27:53 -!- jpc has joined.
00:52:29 <ehird> AnMaster: you must realise that in US is not very useful
00:52:37 <ehird> because the US is basically 51 countries :P
00:52:57 <pikhq> ehird: More than 50.
00:53:13 <pikhq> There's also the District of Columbia, Puerto Rico, ...
00:53:26 <ehird> isn't the District of Columbia small
00:53:45 <pikhq> But heavily urban.
00:53:45 <ehird> also, puerto rico doesn't really share the same services as the us afaik
00:53:53 <ehird> pikhq: alright then
00:54:19 <pikhq> Puerto Rico is set up similarly to the rest of the US.
00:54:43 <pikhq> Though, they pay no federal income taxes, and don't have representation in Congress.
00:54:51 <pikhq> Also, their economy is a tiny bit t3h suck.
00:55:11 <ehird> why isn't D.C. a state anyway
00:55:41 <pikhq> Because the founders wanted the capital to be independent from the states...
00:56:07 <ehird> that's not an answer :P
00:56:35 <pikhq> Why they wanted it that way? Something like "desiring neutrality on possible inter-state conflicts"...
00:57:28 <ehird> it's not like the govt can't vote themselves largesse anyway :P
00:58:39 <Sgeo> I'm trying to promote a group to be an antidote to all the annoying "Add this to get a dislike button/to see who's stalking you" groups: http://www.facebook.com/group.php?v=info&ref=mf&gid=218522451550
00:59:16 <ehird> "If I were malicious, I could have taken over your Facebook account. Do not trust arbitrary Javascript."
00:59:16 <ehird> ooh, a morality tale
00:59:26 <ehird> this just keeps getting more and more Yawnsville, population: this
00:59:29 <ehird> "* Seth (creator)"
00:59:36 <ehird> i thought you didn't like people knowing your name was Seth
01:01:21 <Sgeo> How likely is a stalker to decide to google Sgeo along with facebook.com?
01:02:44 <AnMaster> <ehird> AnMaster: you must realise that in US is not very useful <-- in US, Nevada
01:02:55 <AnMaster> ehird, also they share country code
01:03:04 <Sgeo> ... the Seagull Extinction Organization?
01:03:24 <AnMaster> ehird, and does anyone have separate short/long distance taxes nowdays?
01:03:43 <AnMaster> At least they removed the difference in Sweden around 12 years ago or so
01:04:09 <pikhq> AnMaster: The US is about the size of Europe... Nevada is a long freaking ways away.
01:04:11 <ehird> the us even has different laws for the same things in its 50 countries :-P
01:04:17 <AnMaster> well different abroad costs for different countries
01:04:22 -!- coppro has joined.
01:04:41 <pikhq> It's got more in common with the European Union than any other sort of government, honestly.
01:04:42 <ehird> the us states have some baseline laws, HOPE AND CHANGE, and not all that much else with practical implications in common :P
01:05:10 <AnMaster> ehird, what about "do not commit murder"?
01:05:18 <ehird> "some baseline laws"
01:05:25 <ehird> also, that's one of the ten commandments given to moses by god
01:05:44 <AnMaster> ehird, I believe it is *also* a low in most countries
01:06:04 <ehird> it's not "do not commit murder", it's "if you commit murder we will make your life horrible by force"
01:06:12 <ehird> "do not" doesn't have many implications
01:06:17 <pikhq> No, the law in most countries is "if you are charged with murder, we will do X to you"
01:06:20 <ehird> then again i guess the ten commandments come with the threat of hell anyway
01:06:25 <AnMaster> ehird, the "do not" is what the intention is
01:06:40 <ehird> christian anarchism is a wonderful contradiction :)
01:06:44 -!- FireFly has quit (Operation timed out).
01:06:58 <ehird> http://en.wikipedia.org/wiki/Christian_anarchism
01:07:25 <ehird> "The state is illegitimate! Authority is false! ...but that guy up there in the sky, he can enforce laws through coercion any time he wants. If you catch my meaning. ;)"
01:08:32 -!- coppro has quit (Read error: 104 (Connection reset by peer)).
01:08:36 <AnMaster> any suggestions for a meaningful meaning of that?
01:09:02 <ehird> washington or franklin did that
01:09:05 * Sgeo actually came up with one a while ago
01:09:08 <ehird> wrote a book that was basically a secular bible
01:09:16 <ehird> basically, using the teachings of jesus as a moral code, without supernatural implications
01:09:26 <Sgeo> ehird, that wasn't Jefferson?
01:09:31 -!- coppro has joined.
01:09:31 <pikhq> The Jefferson Bible.
01:09:35 <ehird> founders, all the same really :P
01:09:39 <ehird> even though franklin isn't one
01:09:46 <ehird> they unfortunately overlooked the fact that the bible isn't really the best moral code
01:10:35 * pikhq notes that Franklin is a founding father...
01:11:02 <ehird> erm, huh, maybe it's something else he wasn't
01:11:07 <ehird> brainfart there :/
01:11:26 <pikhq> He was definitely never President.
01:11:27 <ehird> embarrassing, this.
01:11:39 <ehird> "BEN FRANKLIN: Not a cucumber. At least that's what THEY want you to believe."
01:11:48 <pikhq> Kinda died before the Constitution was signed, so...
01:11:59 <AnMaster> ehird, and not a tomato either
01:12:21 <AnMaster> anyway you can make up lots of stuff he wasn't
01:12:21 <ehird> pikhq: huh, franklin died before the us begun?
01:13:23 <pikhq> ehird: No, no. He died before the second constitution was signed.
01:13:51 <pikhq> The Articles of Confederation, however, were around in his lifetime.
01:14:01 <ehird> §By 2009, game developers will face…
01:14:02 <ehird> – 80+ hardware threads
01:14:02 <ehird> – >1 TFLOP of computing power
01:14:02 <ehird> §GPU’s with general computing capabilities.
01:14:04 <ehird> §Game developers will be at the forefront.
01:14:06 <ehird> §If we are to program these devices
01:14:08 <ehird> productively, you are our only hope!
01:14:09 <ehird> — Tim Sweeney, The Next Mainstream Programming Language
01:14:12 <ehird> that CPU line is a bit of an epic misprediction
01:14:29 <ehird> http://www.st.cs.uni-saarland.de/edu/seminare/2005/advanced-fp/docs/sweeny.pdf rest of it is top-notch stuff though. And its mentioning of dependent types makes me really want to find a practical way to do them.
01:15:15 <pikhq> He served as the first ambassador to France and Sweden, as well as being the first Postmaster, for the united States of America.
01:16:14 <pikhq> Strictly speaking, the first constitution is still around -- it declared itself to be perpetual. :P
01:16:34 <ehird> Someone use that in court.
01:17:13 <pikhq> In fact, near as I can tell, the second one merely replaces most of the functional provisions of the constitution, "to form a more perfect Union".
01:17:38 <pikhq> ... Oh, that's not just my interpretation.
01:17:54 <pikhq> That's the opinion of the Supreme Court, in Texas vs. White (1869)
01:18:38 <ehird> a working dependent type system should be purely compile-time of course...
01:18:49 <ehird> in fact i think using them will give the compiler more static information and thus let it compile better
01:19:09 -!- FireFly has joined.
01:25:41 <ehird> pikhq: hmm... the array type in a dependently-typed language should have the size as part of its type, shouldn't it?
01:27:58 <ehird> Index (Array n _) = Nat `That` (< n)
01:27:59 <ehird> Index (Array n _) = Set.filter (< n) Nat
01:28:06 <ehird> for the latter, the type of types would be Set
01:28:19 <ehird> dunno which i prefer more, former seems more "familiar"
01:28:30 <ehird> latter seems more general
01:29:50 <ehird> foo :: ary@(Array n a) -> Array m (Index ary) -> Array m a
01:30:40 <ehird> i.e. foo (makeArray [10..1,-1]) (makeArray [2,4]) → makeArray [8,6]
01:31:07 <ehird> dunno whether that's actually any more "meaningful" than having
01:31:20 <ehird> NatBelow n = Set.filter (< n) Nat
01:31:36 <ehird> foo :: Array n a -> Array m (NatBelow n) -> Array m a
01:32:17 <ehird> pikhq: have you read the "total fp" paper?
01:33:14 <ehird> http://lambda-the-ultimate.org/node/2003, direct link: http://www.jucs.org/jucs_10_7/total_functional_programming/jucs_10_07_0751_0768_turner.pdf
01:33:25 <ehird> pikhq: basically, it's sub-TC but not totally impractical FP
01:33:27 <ehird> no function can diverge
01:33:30 <ehird> i.e. partiality is a side-effect
01:33:43 <ehird> i.e. if a function types, it returns a result of that type when you call it. No exceptions
01:33:46 <ehird> anyway, the idea is
01:33:56 <ehird> in a dependently-typed language, you are often called upon to prove that a value has a certain type
01:34:01 <ehird> because of the TC type system
01:34:18 <ehird> what if the language had a mode in which it was a Total FP language
01:34:26 <ehird> and that language is what you do proofs in?
01:34:34 <ehird> that way, your proofs must be sound
01:34:43 <ehird> no using "undefined" to get around type requirements in your proof or whatnot
01:34:47 <ehird> thus ensuring the safety of the language
01:35:03 <ehird> incidentally, spot the bug in that paper
01:35:04 <ehird> > fib (n+2) = fib (n+1) + fib (n+2)
01:35:08 <ehird> slight oops there :)
01:35:29 <ehird> amusingly enough that wouldn't be valid in total fp
01:35:32 <ehird> since n+2 is not reduced
01:35:43 <ehird> see, an accidental case study right in the paper
01:56:10 <ehird> why does wikipedia need 7.5 M$
01:56:20 <ehird> bandwidth doesn't cost _that_ much
01:56:24 <ehird> nor does server space
01:57:31 <soupdragon> is this just an intellectual exercise, ehird?
01:58:39 <soupdragon> you seem to be designing a dependently-typed language
01:59:02 <ehird> [01:14] ehird: http://www.st.cs.uni-saarland.de/edu/seminare/2005/advanced-fp/docs/sweeny.pdf rest of it is top-notch stuff though. And its mentioning of dependent types makes me really want to find a practical way to do them.
01:59:17 <ehird> but also as intellectual masturbation, yes... like everything we do in this channel
01:59:24 <ehird> soupdragon: to do it in a more practical way
01:59:37 <ehird> to not be a proof system like coq and agda and the like
01:59:51 <ehird> to have reasonable io working with dependent types
01:59:58 <ehird> to be more haskelly, haskell gets most of the other stuff right
02:00:03 <ehird> no reason to deviate when it's not required
02:00:12 <ehird> and to also have compiles be relatively short and the like
02:00:20 <ehird> a practical dependently-typed language, then
02:04:45 <soupdragon> You will need a large library of (beginner level) mathematics to justify termination and correctness for less basic programs, and some kind of plug-in system to hook new decision procedures into elaborating programs
02:05:02 <pikhq> ehird: The Wikimedia Foundation does more than host Wikipedia.
02:06:15 <ehird> yes, but 7 and a half megabucks?
02:06:35 <ehird> soupdragon: not concerned about termination
02:06:44 <ehird> be partial all you want unless it's in the proof subsystem
02:06:51 <ehird> (which is a total subset of the language)
02:07:29 -!- BeholdMyGlory has quit (Remote closed the connection).
02:07:44 -!- FireFly has quit ("Leaving").
02:07:50 * pikhq pulls up the Wikimedia finance report
02:08:36 <ehird> soupdragon: point is, though, regular programming should just be like haskell but a little more type-strict goodness and a few more type annotations to prove to the computer that you're not being stupid
02:08:41 <ehird> or rather, as close to this goal as is possible
02:10:07 <pikhq> $3 million in salaries, $1 million in hosting, $0.2 million for fundraising, $0.3 for travel expensions, $0.7 for facilities...
02:10:23 <soupdragon> but what exactly do you mean not being stupid, there is a spectrum of correctness and if you want to reach certain levels the impact on the programmer will have a stronger effect
02:10:31 -!- poiuy_qwert has joined.
02:10:40 <coppro> ehird: what do you consider an annotation?
02:10:46 <AnMaster> <ehird> why does wikipedia need 7.5 M$ <-- secret Mind/Gene Ray control project
02:11:18 <ehird> coppro: well, technically it'd be a proof
02:11:25 <ehird> in a total subset of the language
02:11:33 <ehird> soupdragon: agreed
02:11:48 <ehird> coppro: if you don't know what a total language is, best to just give up on the discussion now :P
02:11:56 <pikhq> Most of those salaries go to tech support.
02:12:46 <coppro> oh wait, I misread your message
02:13:06 <ehird> pikhq: Tech support. Really now.
02:13:21 <pikhq> ... Wrong fucking term.
02:13:24 <ehird> 1 M$ in hosting... seems about right
02:13:26 <pikhq> It's been a long day. XD
02:13:31 <ehird> 0.2 M$ for fundraising?!
02:13:47 <ehird> All they do is tell the programmers: "Put a fucking big banner up and link to a video by Jimmy Wales that nobody will watch."
02:13:54 <ehird> "MAKE IT BIGGER THAN THE SUN"
02:14:26 <ehird> anyway, that's 5.2 M$
02:14:33 <ehird> so where did the 2.3 M$ come from?
02:14:57 <pikhq> That's not the whole thing.
02:15:07 <pikhq> Just some of the larger items.
02:16:19 <soupdragon> you didn't answer my question though -_-
02:16:48 <soupdragon> what exactly do you mean not being stupid
02:17:11 <ehird> soupdragon: as in, you read a string from stdin and parse it into a Nat
02:17:22 <ehird> and pass it to a function expecting a (Set.filter (< somenumber) Nat)
02:17:39 <ehird> at this point, the compiler goes "WHOA BOY! I'm gonna have to see some ID for that natural."
02:20:05 <ehird> basically your responsibility would be providing a proof that the number you read conforms to (Set.filter (< somenumber) Nat)
02:20:14 <ehird> i.e. providing a proof that the number < somenumber
02:20:34 <ehird> so you'd do an if/else to make sure it was, and in the clause where it is ... yer done
02:28:56 <ehird> soupdragon: wasn't that question going to lead onto something else? :P
02:29:39 <soupdragon> I'm trying to gauge where you are targeting but you've just said that it's possible to depend on preconditions
02:34:08 <ehird> soupdragon: as opposed to?
02:34:19 <ehird> admittedly I'm not the most familiar with dependent types; I know the basic structure but not the variations
02:34:48 <soupdragon> have you studied the ones I mentioned earlier
02:35:51 <AnMaster> <ehird> [01:14] ehird: http://www.st.cs.uni-saarland.de/edu/seminare/2005/advanced-fp/docs/sweeny.pdf rest of it is top-notch stuff though. And its mentioning of dependent types makes me really want to find a practical way to do them.
02:35:57 <ehird> no; I will. how are they different from coq/agda style?
02:36:01 -!- adam_d has quit (Read error: 110 (Connection timed out)).
02:36:07 <AnMaster> "§ By 2009, game developers will face…
02:36:20 <soupdragon> well they're a lot closer to what you seem to be describing (which is why I mentioned the)
02:36:23 <ehird> it's not his fault progress let him down
02:36:27 <ehird> how many cores does the ps3 have anyway
02:36:49 <AnMaster> ehird, "80+ hardware threads" is also off
02:37:02 <ehird> one of them is PPE the others are SPEs
02:37:09 <ehird> AnMaster: yeah, well, can't fault a man for being hopeful
02:37:13 <ehird> soupdragon: googlin' em up
02:37:35 * Sgeo can fault Kurzweil for giving him false hope
02:37:40 <Sgeo> If he turns out to be wrong
02:37:46 <ehird> Kurzweil is wrong.
02:38:00 <ehird> His dates, certainly.
02:38:08 <ehird> The other stuff, who knows.
02:38:22 <ehird> But he very much chooses and advances his dates based on his expected lifespan.
02:38:53 <Sgeo> Night AnMaster
02:39:04 -!- anmaster_l has quit ("Leaving").
02:40:37 -!- Pthing has quit (Remote closed the connection).
02:42:31 <ehird> soupdragon: Data Manipulation Language - Wikipedia, the free encyclopedia
02:42:32 <ehird> Data Manipulation Language (DML) is a family of computer languages used by computer programs and/or database users to insert, delete and update data in a ...
02:42:33 <ehird> presumably not that
02:42:39 <ehird> and I can't get meaningful results for She
02:42:51 <ehird> could you link me to appropriate documents for DML and She?
02:43:25 <ehird> "While ATS is primarily a language based on eager (aka. call-by-value) evaluation" laaaaame :)
02:43:34 <soupdragon> it's basically ML with arithmetic in the type system
02:44:01 <soupdragon> She is the Strathclide Haskell Extention
02:44:21 <ehird> integer arithmetic?
02:44:32 <ehird> ok, so dml begat ATS
02:44:38 <ehird> soupdragon: ah yes, _that_ she
02:45:10 <ehird> soupdragon: how does she work, btw? it's just a preprocessor, isn't it?
02:45:14 <ehird> not sure how it can do dependent types like that
02:45:25 <ehird> [[The Strathclyde Haskell Enhancement is an experimental preprocessor for the Glasgow Haskell Compiler, concocted hastily by Conor McBride at the University of Strathclyde. Its current functionality includes]]
02:45:30 <ehird> Self-admittedly a preprocessor.
02:45:36 <ehird> soupdragon: but do you not distinguish cpp from gcc
02:45:43 <ehird> when I read She output
02:45:47 <ehird> it looks very much like haskell, tbh
02:45:54 <ehird> can it really do the full shebang of dependent fun?
02:45:58 <ehird> type-safe printf, for instance?
02:47:27 <ehird> ok, so dml is a restricted form of dependent types
02:47:42 <ehird> seems to be adequate for basic things
02:48:14 <ehird> ats is impure, it seems.
02:51:32 <soupdragon> she doesn't support full spectrum dependent types
02:51:58 <ehird> admittedly i don't even know if full dependent types are useful
02:52:28 <ehird> type-safe array indexing, absolutely, type-safe printf, almost certainly
02:52:43 <ehird> going more expressive than that, though, probably gets very annoying for the programmer thrust those types upon him fast
03:21:19 <ehird> fizzie: didn't you say debian is using grub 2 these days?
03:21:23 <ehird> installed debian testing, 1.97
03:30:38 <ehird> The Linux OOM killer: "it's like a big game of core wars on your computer".
03:36:13 <pikhq> Yes, 1.97 is Grub 2.
03:36:22 <ehird> Stupid versioning system.
03:36:32 <pikhq> Well, they never had a 1.0...
03:36:40 <pikhq> So they're using the 1.x for pre-release builds of 2.
03:36:45 <ehird> C-INTERCAL's is much more reasonable. It'd be -3.2
03:36:54 <ehird> Or 2.-3, in traditional major.minor form.
03:40:50 <ehird> technically i find that too restricting in the integer form
03:40:52 <ehird> I would do it like this
03:44:17 <ehird> making root accessible only by sudo for dummies
03:44:31 <ehird> I used to just do -l, but it turns out that leaves the original password after !
03:44:37 <ehird> which makes me uncomfortable, as it will never be used again
03:44:43 <ehird> this one replaces the entire field with a nice clean !
03:46:11 <ehird> wtf, default debian includes "vi" as vim but not "vim"
03:54:35 <ehird> pikhq: incidentally, here's the most retarded thing ever: Someone actually made their shell script explicitly execute with dash, not because they required some POSIX-compliant thing that bash and the like lacked, but because they were writing it in POSIX shell, and so used the only POSIX-compliant shell they knew of.
03:54:40 <ehird> You know, not like /bin/sh is supposed to be that.
03:54:52 <ehird> And it's not like bash suffices for... well, just about any POSIX shell use.
03:54:53 <pikhq> ehird: That's freaking retarded.
03:55:43 <pikhq> Default Debian kinda has a barebones install, but I didn't realise they were so barebones as to not install vim...
03:56:04 <pikhq> And I thought that dash was only used as a small shell for the installer...
03:56:37 <ehird> Ookay, I don't think people in the sudo group are meant to receive email sent to root.
03:56:52 <ehird> Oh, probably because I did "sudo apt-get dist-upgrade" apt decided to be helpful and send it to me, too.
03:56:56 <ehird> pikhq: They have vim.
03:56:59 <ehird> It's just called vi.
03:57:05 <ehird> And no, dash is /bin/sh on Debian.
03:57:11 <ehird> Nothing wrong with that.
03:57:17 <ehird> But what this person did? Dummmmmmb.
03:57:20 <ehird> pikhq: yeah, it's weird
03:57:22 <pikhq> vi being vim, but not vim being vim, that is.
03:57:29 <pikhq> /bin/sh being dash?
03:57:35 <pikhq> I can accept that.
03:57:41 <ehird> Well, it's vim-tiny, which is mostly intended for things-that-call-vi.
03:58:16 <pikhq> /bin/sh should only be a POSIX shell -- beyond that, I don't care so long as I can get me a zsh.
03:58:22 <ehird> ls: unrecognized prefix: hl
03:58:22 <ehird> ls: unparsable value for LS_COLORS environment variable
03:58:31 <ehird> Upgrading to sid breaking your current session's ls.
03:58:33 <ehird> There's a new one.
03:58:44 <pikhq> Yeah, that's a new one.
03:58:47 <ehird> At least shutdown still works.
03:58:50 <pikhq> export LS_COLORS=""?
03:59:05 <ehird> I just rebooted. Probably some bootup stuff changed, anyway.
03:59:09 <ehird> Might as well have it all happy-like.
04:01:31 <ehird> ...wait, "apt-get autoclean" exists?
04:01:42 <pikhq> BTW, you should totally use aptitude.
04:01:48 <ehird> I wonder if it was a bad idea to use autoclean.
04:01:54 <ehird> pikhq: apt has progressed enough that aptitude is useless
04:02:18 * ehird tries to figure out if there's an apt-get no-i-dont-fucking-want-that-old-kernel
04:02:22 <pikhq> I thought that aptitude had better dependency resolution than apt, and that apt-get was considered outmoded?
04:02:51 <ehird> Nope, the thing aptitude gives you (remove packages that aren't depended on any more) is now available as "apt-get autoclean".
04:03:00 <ehird> Admittedly, it's an extra step, but it informs you they exist whenever you do anything else.
04:03:22 <pikhq> Okay, so aptitude doesn't give you anything more than an ncurses interface.
04:03:31 <pikhq> (and not a great one)
04:04:31 <ehird> http://i.imgur.com/Dnozc.gif
04:06:09 <ehird> http://www.arrangebypenis.com/
04:06:31 <pikhq> Yes, yes, http://i.imgur.com/Dnozc.gif the http://www.arrangebypenis.com/!
04:08:38 <ehird> alias sag='sudo apt-get'
04:08:38 <ehird> alias sagi='sudo apt-get install'
04:08:52 <ehird> "sag remove" is rather disturbing though
04:10:21 * ehird tries to figure out where debian wants me to put things
04:10:31 <ehird> I think export EDITOR=vim should go in .profile
04:10:38 <ehird> and those aliases in .bashrc
04:11:08 <ehird> yeah, it's just that debian comes with tons of stuff in the files by default
04:11:23 <ehird> e.g. .profile includes .bashrc if we're running bash
04:11:39 <coppro> pikhq: Aptitude has some other features, like when you perform an operation, it tells you how many packages have changed status
04:11:49 <coppro> also, it has better conflict resolution
04:13:28 <ehird> has anyone ever installed sid by changing the mirrors in the debian-installer testing livecd :)
04:13:39 <ehird> i don't see any reason it shouldn't be as reliable as upgrading from unstable
04:13:45 <ehird> (for value of reliability equal to not at all)
04:14:13 <pikhq> ehird: I thought that that was a supported means of using the livecd?
04:14:23 <pikhq> Well, as supported as anything else in Sid.
04:14:35 <ehird> The canonical answer is: You don't. You can only upgrade to it from stable or testing. You do that by editing /etc/apt/sources.list and changing your sources from stable to unstable.
04:14:35 <ehird> There are some unofficial "sid ISO images" out there. They are dangerous, unofficial and obsolete (by definition!). Stay away from them.
04:14:36 <ehird> It may also be possible to install sid packages instead of testing packages if you're using a net install from the testing branch. This is not supported, but if you want to try it, you're free to do so. It's your machine, after all. Just don't cry if it breaks.
04:14:42 <ehird> (answer to "How do I install sid?")
04:14:54 <ehird> the answer being basically "It might work, and it might work. You know, just like sid LOL"
04:15:32 <ehird> "Should I use sid on my server?
04:15:32 <ehird> Are you insane? No!"
04:15:32 <ehird> FACTUALLY INCORRECT, FAQ-WRITER
04:18:06 <pikhq> The only thing crazier is using Gentoo ** on a server.
04:18:42 <pikhq> (meaning KEYWORD_ACCEPT="**", meaning that Portage will feel free to install any package that is marked as being able to compile)
04:19:32 <pikhq> I should note that that's "compile on at least one of Gentoo's architectures", not necessarily "compile on your architecture".
04:20:00 <pikhq> It'd even accept building FreeBSD libc on Windows doing that.
04:20:03 <ehird> It's not too crazy; I'd say rumours of sid's dog-eating are greatly exaggerated
04:20:19 <ehird> I mean, come on; it's not like Arch will be any stabler
04:20:53 <ehird> Debian folk are just the genteel, careful sort.
04:21:32 * pikhq wonders if FreeBSD libc can build on any non-FreeBSD system...
04:22:20 <pikhq> Well. Yeah, probably there.
04:23:26 <coppro> I'd expect it'll build on most systems. No clue if it will run
04:23:31 <pikhq> Gentoo only appears to support it on sparc-fbsd and x86-fbsd.
04:23:42 <pikhq> Doesn't mean much, though.
04:25:50 <ehird> What would be nice: A sort of blend of awk and sed.
04:26:05 <ehird> Say a script produces foo, a number of spaces depending on the width of foo, and then a size in kilobytes, lots of times..
04:26:10 <ehird> But you want it in megabytes.
04:26:45 <ehird> sewk '/\d+/ { print &/1024 }'
04:26:45 <pikhq> That idea, plus 20 years, is Perl.
04:26:52 <ehird> yes, but perl is shit.
04:27:19 <uorygl> "Perl sucks." --my dad, a Java web programmer
04:27:28 <ehird> actually if you used the awk derivative proposed in the structural regular expressions paper, you could do
04:27:51 <coppro> perl -e 'while (<>) { s=(\d+)=$1/24=e; print; }'
04:27:56 <uorygl> pikhq: so does that mean Portage will actually try to install all those packages, or just that it will be relatively uninhibited?
04:28:00 <ehird> awk '/\d+/ { print $1/1024 } /.*/ { print $1 }'
04:28:16 <pikhq> uorygl: It will try to install them if you ask for them.
04:28:30 <ehird> coppro: using = as a dlimiter.
04:28:37 <ehird> that is just awful
04:28:41 <ehird> also, a manual while <> loop?
04:29:05 <coppro> ehird: feel free to use pipe or something
04:29:11 <pikhq> It's basically the "Fuck off, Gentoo, I know exactly what I want installed" mode.
04:29:29 <ehird> coppro: Fine, then at least:
04:29:38 <ehird> perl -pe 's=\d+=&/24=e'
04:29:43 <ehird> Or, less HORRIBLY CONFUSINGLY,
04:29:53 <ehird> perl -pe 's|\d+|&/24|e'
04:30:02 <uorygl> What does while (<>) mean?
04:30:19 <ehird> <> = read a line from input; if you don't assign it to anything (or maybe even if you do), put it in $_.
04:30:36 <ehird> input is either stdin, or if you put multiple files as command line arguments, them in succession (as if catted together)
04:30:45 <ehird> obviously it's false as a boolean if there's no more input
04:30:51 <uorygl> Is this "maybe" an ehird-uncertainty maybe or a the-way-Perl-actually-works maybe?
04:30:53 <ehird> so while (<>) continually slurps lines of input, for processing
04:31:11 <pikhq> uorygl: Perl is crazy, but not that crazy.
04:31:22 <coppro> $_ is only used by default if no other variable is specified
04:31:31 <ehird> Ugh, if you specify e as a regexp option, & isn't expanded.
04:31:36 <ehird> Why are you fuck-shit retarded, Perl. Why.
04:31:39 <pikhq> It's merely crazy enough to make parsing equivalent to solving the halting problem. :P
04:31:44 <ehird> coppro: no, that's not what & is
04:31:52 <ehird> & should work to avoid needless parenthesising of the whole expression
04:31:58 <ehird> & is what $0 would be\
04:32:04 <ehird> if $0 wasn't taken
04:32:16 <coppro> & might be used in the expression
04:32:32 <ehird> so escape it, the regex terminator mighht be too
04:32:44 <ehird> okay, this works, somehow it fucks up the alignm— wait a second, those results are wrong
04:33:06 <ehird> $_ is wrong, bitch :P
04:33:53 <ehird> \d was replacing the numbers in the package names
04:34:24 <ehird> since some of them are
04:34:36 <ehird> the replacement messes it up
04:35:01 <ehird> perl should have a thing you can enable so that it analyses how the data is aligned, and keeps that alignment.
04:36:35 <uorygl> It often seems like other languages use syntactic sugar where Haskell would use a user-definable function.
04:36:49 <coppro> Perl is all syntactic sugar
04:37:33 <uorygl> Yeah, Lisp has macro thingies. I don't know if I want to wrap my entire program inside one function that changes the program's semantics perhaps significantly.
04:37:41 <uorygl> And, of course, Haskell has lots of syntactic sugar.
04:38:19 <pikhq> I count only a few bits.
04:38:20 <ehird> you don't have to wrap your entire program, macros can be used in subexpressions you know :P
04:38:22 <uorygl> It has too many pieces of syntactic sugar to count on one hand.
04:38:32 <coppro> I mean, C++ is C with syntactic sugar. *ducks*
04:38:42 <ehird> http://www.reddit.com/r/pics/comments/ahl6z/i_dare_you_to_set_this_as_your_desktop_background/c0hm413?context=4
04:38:53 <ehird> oh god, now I want a rotatable monitor
04:39:43 <uorygl> Newtype declarations, guards, do notation, pattern guards, case statements, list notation, list builder notation...
04:39:54 <ehird> Newtype declarations aren't sugar.
04:40:01 <pikhq> Case statements aren't sugar.
04:40:16 <ehird> Well, either case or pattern matching is sugar.
04:40:21 <ehird> Kind of irrelevant which one.
04:40:25 <ehird> "If you put it as your desktop background and then start pink floyd's "the wall" at the same time as the lion roars on the Wizard of Oz...
04:40:25 <ehird> Flying monkeys come out of YOUR BUTT"
04:40:26 <pikhq> ehird: I thought that the other pattern matching got desugared to case?
04:40:30 <uorygl> If case statements aren't sugar, then ordinary pattern matching is sugar.
04:40:35 <ehird> It's equivalent, either way.
04:40:36 <pikhq> Moot point, though.
04:40:41 <uorygl> Semantically, it doesn't... what they said.
04:41:20 <pikhq> uorygl: You managed to name most of the syntactic sugar.
04:41:45 <pikhq> Infix functions and list comprehensions are the other two that I can think of.
04:42:01 <uorygl> Then you have some exotic things like mdo notation, do guards, arrow do notation.
04:42:17 <uorygl> List comprehensions are what I meant by list builder notation.
04:42:21 <pikhq> mdo notation, do guards, and arrow do notation are GHC extensions.
04:43:28 <pikhq> You also omitted the (soon-to-be-gone) n+k matches.
04:44:04 <uorygl> I thought they might have already been gone.
04:45:01 <ehird> They are in Haskell 2010. Thank god.
04:45:28 <pikhq> Oh, right. Haskell 2010 has been ratified.
04:49:29 <uorygl> ehird: I have set that image as my desktop background.
04:49:34 <pikhq> The Haskell Commitee.
04:49:44 <ehird> Pantomime moment there.
04:49:57 <ehird> "Who ratified it?" "The committee" "Which committee?" "The Haskell Committee."
04:50:48 <coppro> Which Haskell Committee?
04:50:58 <ehird> The only Haskell Committee!
04:51:21 <uorygl> Does this Haskell Committee have a website?
04:51:37 <ehird> Which www.haskell.org?
04:52:03 <uorygl> The www.haskell.org endorsed by haskell.org's nameserver!
04:52:56 <uorygl> Which haskell.org? The haskell.org endorsed by .org's nameserver! Which .org? The one operated by Afilias Limited! Which Afilias Limited? I dunno, is there more than one?
04:53:13 <pikhq> uorygl: Which .org? The .org endorsed by .'s nameserver!
04:53:33 <uorygl> But . has lots of nameservers, each operated by a different company.
04:54:34 <coppro> There are 11 root nameservers iirc?
04:54:52 <uorygl> That's too many letters.
04:54:59 <pikhq> Under root-servers.net
04:55:19 <pikhq> uorygl: I redefine arithmetic to make you wrong.
04:55:42 <ehird> SWEET BABIES OF LUXURY
04:55:42 <pikhq> succ(12) = 11, dammit!
04:55:50 <ehird> Suck 12 equals 11.
04:56:00 <ehird> YOU KILLED SOMEONE WHILE PERFORMING ORAL SEX UPON THEM?!
04:56:12 <uorygl> www.haskell.org is the same thing as bugs.haskell.org, according to serv1.net.yale.edu. serv1.net.yale.edu is an authoritative nameserver for haskell.org, according to A2.ORG.AFILIAS-NST.INFO. A2.ORG.AFILIAS-NST.INFO is an authoritative nameserver for .org, according to G.ROOT-SERVERS.NET. G.ROOT-SERVERS.NET is an authoritative nameserver for ., according to 209.20.72.4.
04:56:17 <uorygl> And 209.20.72.4 is not an authoritative nameserver.
04:56:29 <pikhq> ehird: No, I am declaring that out of ever 12 instances of oral sex, 11 will survive.
04:56:45 <ehird> ORAL SEX: The hidden killer... IN YOUR PANTS
04:56:56 <ehird> FOX NEWS SPECIAL REPORT
04:57:35 <pikhq> uorygl: Out of band knowledge confirms that g.root-servers.net is an authoritative nameserver for . (according to ICANN)
04:57:46 <uorygl> So in order to figure out what www.haskell.org is, one must first know what serv1.net.yale.edu and A2.ORG.AFILIAS-NST.INFO are.
04:57:49 <ehird> I wish with Debian-Installer you could say "regular install but prompt me for this extra step"
04:57:55 <ehird> as opposed to trundling through the boring expert install
04:58:19 <uorygl> In order to figure out what serv1.net.yale.edu is, one must first know what C.GTLD-SERVERS.NET is.
04:58:33 <ehird> In order to figure out what any domain is, one must first know what any of [A-M].ROOT-SERVERS.NET are.
04:58:54 <pikhq> The query goes from . down.
04:58:55 -!- zzo38 has joined.
04:58:56 <uorygl> In order to figure out what C.GTLD-SERVERS.NET is, one must first know what C.GTLD-SERV--wait, hang on.
04:59:11 <ehird> pikhq: Yes, but . is defined by the root serverrs.
04:59:14 <zzo38> Hay! Wait! Hang on!
04:59:26 <ehird> If you have an IPP for any of [A-M].ROOT-SERVERS.NET, you're sorted.
04:59:37 <zzo38> Now I can look at the logs
05:00:13 <uorygl> It appears that C.GTLD-SERVERS.NET is inaccessible without prior knowledge of C.GTLD-SERVERS.NET
05:00:19 <pikhq> ehird: Yes. The root servers are kinda stuck into BIND.
05:00:50 <uorygl> How do you figure out what it is, then? It's the nameserver for .net.
05:01:00 <ehird> It's cool that you only need to know one single IP to be able to browse all the web you want.
05:01:14 <uorygl> Assuming that /[A-M].ROOT-SERVERS.NET/ matches only one string.
05:01:37 <ehird> You could be locked in a room with only a Forth console plugged into an internet connection, and as long as you can remember one single IP, you can build yourself a full web browser.
05:01:48 <ehird> And browse the interrwebnets.
05:01:52 <pikhq> uorygl: Here's how it works: you query [a-m].root-servers.net what net is. You query net what gtld-servers.net is. You query gtld-servers.net what c.gtld-servers.net is.
05:02:17 <pikhq> ehird: It'll be difficult, but yes.
05:02:23 <uorygl> pikhq: net isn't a server; you can't query it.
05:02:31 <pikhq> uorygl: Yes it is.
05:02:39 <ehird> Question. When Debian "installs the base system", from CD instead of the network, is any of that left at the end of the installation?
05:02:45 <ehird> Or is it all upgraded from the repos.
05:02:55 <ehird> pikhq: DNS is a pretty easy protocol, isn't it?
05:03:08 <uorygl> pikhq: ping tells me that net is an unknown host. dig tells me that net has no A record.
05:03:13 <ehird> TCP/IP would fuck you up, but let's say TCP/IP was on a piece of paper next to you.
05:03:39 <uorygl> Doesn't DNS operate greatly over UDP?
05:03:42 <pikhq> Oh, fine. You query [a-m].root-servers.net what gtld-servers.net is. You query gtld-servers.net what c.gtld-servers.net is.
05:03:44 <zzo38> Of course, it would be simpler to make a client to telnet and make like a raw dumb terminal, or to make a simple gopher client, and so on.
05:03:59 <pikhq> UDP or TCP; both are valid.
05:04:00 <ehird> Then it wouldn't be hard from just an internet link, the TCP/UDP/IP specs, one single IP, and knowledge of how to do basic DNS and HTTP requests to load google.com.
05:04:16 <ehird> A little string manipulation later, voila, dumb-ass web browser.
05:04:23 <zzo38> Yes, you could load google.com easily like that
05:04:26 <pikhq> UDP is generally used for smaller queries, but supporting it isn't mandatory from a client.
05:04:42 <ehird> A better path may be to connect to IRC and ask for help because dammit they aren't giving me food and I'm not sure where I am and I don't know if they'll let me out and I don't know who they are help helph elp
05:04:46 <zzo38> But you would need HTML and various image file formats, JavaScript, etc, to make the full use.
05:04:58 <ehird> zzo38: HTML and JavaScript are quite self-describing.
05:05:03 <pikhq> I can browse the web with freaking nc.
05:05:14 <pikhq> It's kinda annoying, but you can do it just fine.
05:05:16 <ehird> If you can do a basic HTTP request, you're savvy enough to work out how HTML, CSS and JS work through observation and testing.
05:05:18 <zzo38> pikhq: Well, yes you certainly can, but it isn't very good
05:05:29 <pikhq> We're not asking for good.
05:05:33 <pikhq> We're asking for functional.
05:05:44 <uorygl> Success; the root nameservers tell you who the gtld-servers.net people are.
05:05:55 <zzo38> It is not too difficult to write a proper HTML, with most things, and a bit harder for JavaScript, although, you would still need it if you wanted it complete
05:06:07 <uorygl> Hmm, I should have realized that before. The root zone file is loaded with hints.
05:06:09 <zzo38> But, yes, just netcat is good enough to connect
05:07:03 <uorygl> I think that if it mentions a domain name, it gives you both an A record and an NS record for it.
05:08:09 <ehird> Actually, that's a good point. Why does Debian netinstall first install from the CD?
05:08:11 <ehird> That is poopy-stupid.
05:08:32 <ehird> ;; AUTHORITY SECTION:
05:08:32 <ehird> .10800INSOAa.root-servers.net. nstld.verisign-grs.com. 2009122201 1800 900 604800 86400
05:08:40 <ehird> Wonder if a is somehow more authoritative than the others, or whether someone was just lazy.
05:09:40 <ehird> "dig net." isn't very helpful.
05:09:47 <zzo38> Which system is best for text-adventure games, is it Glulx, or TADS, or Z-machine
05:10:20 <uorygl> ehird: yes, because there's no server there.
05:10:32 <uorygl> dig looks for A and maybe AAAA records for whatever you give it.
05:10:34 <ehird> There isn't any at ., either.
05:10:51 <ehird> Those are the special magic root server records.
05:11:08 <uorygl> It would be cute if . were a domain name of an actual server.
05:11:16 <ehird> Glulx is just a vm
05:11:22 <ehird> and so is z-machine
05:11:26 <ehird> it's a full system
05:11:27 <zzo38> Inform 6 or 7? And it compiles to Glulx and Z-machine, which of those is better
05:11:39 <ehird> Glulx is "cooler" but Z-machine is much more commonly implemented
05:11:43 <ehird> glulx never took off afaik
05:11:44 <uorygl> Anyway, I think that in theory, all the root nameservers are mirrors of a.root-servers.net.
05:11:55 <ehird> inform 7 if you can stomach the syntax, it's where most the work goes today
05:11:56 -!- jpc has quit (Read error: 60 (Operation timed out)).
05:12:01 <uorygl> Since there has to be only one primary authoritative nameserver.
05:12:40 -!- oerjan has joined.
05:12:40 <uorygl> DNS's email address notation is so cute, you know.
05:13:05 <uorygl> It almost makes you want email addresses to actually be domain names.
05:13:51 <zzo38> I make my own text-adventure game system, too, it is called TAVSYS, see the example http://zzo38computer.cjb.net/img_10/tavsys_example_1.png
05:14:13 <ehird> Why am I not surprised.
05:14:51 <pikhq> I'm surprised zzo38 doesn't make his own computers.
05:15:08 <pikhq> And core memory. On a loom. Of his own design.
05:15:18 <uorygl> I want to make my own computers. Can I rent a microchip fabrication plant?
05:15:32 <zzo38> I would make my own computers, one day. But not yet, because I need the equipment and stuff I would get from help from someone I know
05:15:36 <uorygl> As in send them a design and get back a microchip.
05:15:57 <ehird> Heck, I'm surprised zzo38 doesn't make his own *physics*.
05:15:57 <zzo38> And I have the similar question(s) like you, too
05:16:18 <zzo38> Physics??? Really?
05:16:27 <oerjan> zzo38: maybe he does, it's just he only communicates through the net so no one notices it's different
05:16:54 <zzo38> I don't only communicate through the net
05:17:13 <zzo38> Making my own computers is something I plan to do soon
05:17:59 * ehird attempts again to install sid via the testing cd
05:18:29 <zzo38> The "G" in the corner is short for "Glk"
05:18:30 <ehird> "It is not possible to install sid from a netinst or full CD. Use the netboot installation method, a businesscard CD image, or floppy images (with the net-driver floppies)."
05:18:35 <ehird> Thank you, Debian-Installer FAQ!
05:19:40 <ehird> http://www.debian.org/releases/stable/i386/ch04s05.html.en
05:19:44 <ehird> Ookay... noot netboot then...
05:19:47 <ehird> Businesscard image it is.
05:21:37 <zzo38> They say Inform 7 with English sentences make it easier, but I think it actually makes it more confusing, for various reasons, including that you might think it implies something, even though it doesn't imply that
05:24:46 -!- jpc has joined.
05:24:56 <uorygl> zzo38: you're not the Loper OS guy, are you?
05:25:32 <zzo38> I don't know what the Loper OS guy is.
05:25:46 <uorygl> He's the guy who writes here: http://www.loper-os.org/
05:25:50 -!- coppro has quit (Remote closed the connection).
05:26:01 <uorygl> He rants and says very interesting things.
05:26:23 <uorygl> Kind of like Eliezer Yudkowsky, who fears and writes very interesting things.
05:26:33 -!- coppro has joined.
05:30:01 -!- ehird_ has joined.
05:36:15 <pikhq> zzo38 is quite a bit saner than the Loper OS guy.
05:36:21 <ehird_> vm.overcommit_memory = 2
05:36:21 <ehird_> vm.overcommit_ratio = 100
05:36:21 <ehird_> I would like to call these two lines the "Not a Turing Machine" Maneuver.
05:36:25 <ehird_> pikhq: Hey, I *like* that guy.
05:36:46 <pikhq> ehird_: He's enjoyable, just somewhat off his rocker.
05:36:58 <pikhq> ... The same is true of most notable mathematicians. :P
05:37:03 <uorygl> I want to be like that.
05:37:12 <ehird_> I don'tt think I've read anything to suggest he's particularly crazy.
05:37:22 <ehird_> Very strong unorthodox opinions, yes...
05:37:22 <pikhq> I might be thinking of someone else.
05:37:33 <ehird_> but I don't think his ambitions are very crazy, just utopian.
05:37:35 <pikhq> For some reason I was thinking of Losethos.
05:37:36 <ehird_> pikhq: Did you mean the Losethos guy?
05:37:48 <ehird_> In that case, I absolutely challenge the "somewhat" part.
05:37:50 <uorygl> Yeah, the Losethos guy is a bit crazy.
05:37:56 <ehird_> This guy never had a rocker, and continually beats up the rockers of everyone else.
05:38:18 <uorygl> Heh heh. "Bill Gates may be richer than Captain Kirk, / but the Windows OS blows! / And sucks! / At the same time!"
05:38:18 <pikhq> Yeah, loper-os is just unorthodox.
05:38:28 <ehird_> "If people think everyone has premarital sex or everyone does drugs, they have no will power to resist. We're gonna have lots of people deciding they're gay.
05:38:28 <ehird_> I don't like gays. I don't want them openly acting gay. It's yucky.
05:38:28 <ehird_> God says... doubted fitter stipend containest instituted Hierius
05:38:28 <ehird_> We're gonna be forced to hire them."
05:38:28 <pikhq> And that's perfectly fine by me.
05:38:29 <ehird_> — Losethos, in a post to reddit. Not a comment, a post. Title: "Gay Marriage".
05:39:20 <ehird_> "God's a child molester.
05:39:20 <ehird_> http://www.usccb.org/nab/bible/ezekiel/ezekiel16.htm
05:39:20 <pikhq> The only reason why I don't think Losethos is *completely* freaking crazy is because he appears capable of programming.
05:39:20 <ehird_> Big thoughts forced into puny heads."
05:39:26 <ehird_> Wait... if he's God-obsessed...
05:39:30 <pikhq> Which I assume requires some hold on reality.
05:39:32 <ehird_> And he thinks God is a child molester...
05:39:39 <ehird_> Perhaps this guy is not as harmless as you might think P
05:39:41 <pikhq> (as tenous as it may be)
05:39:45 <soupdragon> since when does ability to program tell you anything about a persons sanity/intelligence?
05:39:52 <ehird_> soupdragon: intelligence it does
05:40:03 <ehird_> if you're completely dissociated from reality you couldn't program
05:40:05 <pikhq> soupdragon: I'm assuming it demands at least a *modicum* of sanity.
05:40:31 <pikhq> I think losethos is as insane as you can be and still program decently.
05:40:43 <ehird_> soupdragon: unless you're going to offer arguments, so be it.
05:40:46 <uorygl> I think it would be possible to teach a dog programming, if you had an eternally young dog.
05:40:59 <ehird_> do you think someone with IQ 1 could program?
05:41:06 <ehird_> yes, I know IQ isn't a measurement of intelligence really
05:41:16 <ehird_> but anyone who scores 1 is either doing it intentionally or is really fucking retarded
05:41:26 <uorygl> ehird_: hey, there's more to discussing than providing argument.
05:41:27 <coppro> or just speaks the wrong language
05:41:40 <ehird_> uorygl: yes, but it's a good step up from assertion
05:41:49 <ehird_> coppro: they're mostly symbol-based, you know
05:42:45 <pikhq> A reputable one is.
05:42:56 * pikhq invokes the True Scotsman fallacy for the win
05:43:02 <ehird_> No IQ test is reputable.
05:43:25 <pikhq> And ehird snatches victory out of pikhq's hands.
05:43:37 <soupdragon> I'm not trying to convince you of something
05:43:55 <pikhq> Random assertions, then? *shrug*
05:44:05 <ehird_> "since when does ability to program tell you anything about a persons sanity/intelligence?" is usually interpreted as the start of some kind of back and forth.
05:44:23 <uorygl> coppro: si hay dos latas, una que contiene cinco galones de agua y una que contiene tres, ¿cómo se mide cuatro galones de agua?
05:44:35 <ehird_> coppro: Si. Si. Si. Uh... si.
05:44:35 -!- ehird has quit (Read error: 110 (Connection timed out)).
05:44:35 -!- ehird_ has changed nick to ehird.
05:44:52 <uorygl> (No, my Spanish is not getting rusty; it's never been better.)
05:45:36 <pikhq> uorygl: Nani? Watasi ha anata no hen na getsugo wo wakaranai, yo.
05:45:46 <zzo38> Which verse of Ezekiel 16 do you mean?
05:45:46 <coppro> uorygl: Fill the five-gallon container, pour three gallons into the three-gallon container, dump the three out, pour the remaining two into the three-gallon container, fill the five-gallon again, pour the last gallon into the three and voila! you have four
05:45:52 <coppro> also, I don't know Spanish
05:45:53 <uorygl> pikhq: by "ha", do you mean "wa"?
05:46:13 <ehird> pikhq: Kanji or GTFO.
05:46:20 <pikhq> uorygl: Yes, I was doing a very very literal transcription of kana.
05:46:20 <pikhq> ehird: I don't have an IME.
05:46:30 <ehird> zzo38: I was just quoting Losethos.
05:46:35 <ehird> God knows what he was thinking in his little mind.
05:46:40 <ehird> pikhq: COPY AND PASTE, FUCKER
05:46:40 <uorygl> Use kanji for every word, including particles.
05:46:51 <ehird> I wish I knew other languages.
05:46:53 <uorygl> pikhq: I have an IME, and I don't even know Japanese.
05:46:54 <zzo38> ehird: Yes, but which verse number?
05:47:01 <pikhq> I'm not about to write kanbun for your sick and twisted pleasure. :P
05:47:04 <ehird> zzo38: I don't know. He did not specify, and I am not telepathic.
05:47:24 <pikhq> (kanbun being Classical Chinese with annotations on how to read it as Japanese)
05:47:29 <uorygl> I'm guessing "watasi" is a pronoun. "Anata" looks familiar. What do those two mean?
05:47:49 <zzo38> I can read/write kana, too, and some words, and some kanji, but I don't have any IME software on my computer
05:47:58 <pikhq> watasi = I, "anata" = you.
05:48:17 <pikhq> The translation is: I don't speak your strange moon-language.
05:48:41 * uorygl arranges that sentence into a Japanese-ish order.
05:49:10 <pikhq> I ha you no strange na moon-language wo understand-not
05:49:17 <uorygl> I your strange moon language not speak.
05:49:41 <uorygl> Neat, it's the order that I guessed, except that the not is a suffix.
05:50:03 <ehird> pikhq: Wait, Japanese is postfix?
05:50:09 <ehird> Mental stacks. Wonderful.
05:50:21 <pikhq> ehird: Somewhat, yes.
05:50:34 <uorygl> Isn't Japanese extremely postfix?
05:50:48 <pikhq> uorygl: Makes the "hen" into an adjective.
05:50:56 <zzo38> Yes, it is postfix in some ways, like, you put verb at the end, for one thing
05:51:21 <ehird> "It's postfix in some ways, like, it's postfix."
05:51:44 <soupdragon> you don't even need to be concious to write programs
05:52:08 <uorygl> Wikipedia says that Japanese is quite strictly left-branching: modifiers tend to precede heads.
05:52:10 <pikhq> "ha" = subject, "no" = possessive, "na" = adjective, "wo" = object...
05:52:43 <uorygl> What sorts of things are "hen" and "getsugo"?
05:53:00 <zzo38> You can search on WWWJDIC.
05:53:09 <pikhq> "hen" is an adjective of Chinese origin, and "getsugo" is a noun.
05:53:17 <zzo38> You need Japanese fonts on your computer to use WWWJDIC
05:53:26 <pikhq> Composed of "moon" (getsu) and "go" (language)
05:53:29 <zzo38> But IME is not required
05:54:07 <uorygl> I hope that Go, the board game, is not named that because that word means "language".
05:54:09 -!- Oranjer has joined.
05:54:22 <ehird> http://harmful.cat-v.org/software/xml/soap/simple
05:54:28 <uorygl> Can "hen" be used without that "na" after it?
05:54:50 <ehird> [05:51] soupdragon: you don't even need to be concious to write programs
05:54:51 <ehird> you need to be conscious to write programs as anything but a spontaneous action
05:55:04 <pikhq> uorygl: The word for go in Japanese is "igo".
05:55:11 <pikhq> And not as an adjective.
05:55:12 <ehird> in which case you could argue that any group of particles in the universe could suddenly spontaneously turn into a program
05:55:21 <ehird> and that would be writing a prograagm
05:55:32 <uorygl> How else can it be used?
05:55:33 <ehird> which is, wossname, ideotic
05:55:36 <pikhq> (it can be combined with other words, though. See the adjective "hentai")
05:56:06 <ehird> Just bring the word "hentai" into a discussion about Japanese. That's utterly surprising and unexpected.
05:56:22 <uorygl> It is in Japanese, I guess.
05:56:44 * ehird looks it up on Wikipedia. Yes.
05:57:00 <ehird> "Graphic hentai representation." —a caption
05:57:06 <ehird> Well, that's one way of wording it, Wikipedia.
05:57:20 <Oranjer> ehird: does not the act of writing require a writer? and does not spontaneous creation require the lack of a creator?
05:57:35 <ehird> Does not your mother require the lack of a creator?
05:57:45 <ehird> She is so hideous, after all.
05:57:57 <zzo38> You can find a lot of Japanese words in WWWJDIC. But, some are still missing. But you can search both kana and kanji, and it will tell you the kana for every word, and examples of Japanese writings, too. And also stroke-orders
05:58:07 <pikhq> Doest thou not know of ehird's particular desire for thy mother?
05:58:19 <Oranjer> I'm saying that the problem seems to reside completely in soupdragon's use of the word "write"
05:58:32 <Oranjer> although, not understanding the context, I have no fucking clue
05:58:45 <ehird> zzo38: What Would Wally Jones Dickinson Ian Conjure?
05:59:23 <ehird> Hey hey, sid installed from scratch.
05:59:27 <ehird> No filthy testing influence here, nosiree.
06:00:34 <uorygl> World-Wide War One Dictionary.
06:00:38 <ehird> And with sudo set up by default, too. Who says Debian don't do none of that thingymagic.
06:01:58 <ehird> Although it adds your username, instead of adding you to the sudo group.
06:04:51 -!- augur has quit (Read error: 54 (Connection reset by peer)).
06:04:56 -!- augur has joined.
06:05:24 -!- augur has changed nick to Guest90348.
06:10:13 <zzo38> ehird: I don't know the answer do your question, but that isn't what WWWJDIC is supposed to be short for.
06:10:44 <zzo38> World Wide Web Japan Dictionary
06:11:55 -!- zzo38 has quit ("QUIT :").
06:12:40 <pikhq> It's the IC in DICtionary.
06:16:01 * ehird believes that "xorg" is the package to install for x magic on debian
06:16:10 <ehird> as opposed to any more complicated, xorg-involving name.
06:21:18 -!- soupdragon has quit ("* I'm too lame to read BitchX.doc *").
06:22:32 <coppro> http://en.wikipedia.org/wiki/Once_More*_With_Footnotes <- must find
06:22:48 -!- poiuy_qwert has quit ("Leaving").
06:24:33 -!- Oranjer has left (?).
06:26:42 -!- Guest90348 has changed nick to augur.
06:27:46 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
06:41:28 * ehird makes xdm actually look acceptable
06:45:45 <ehird> http://imgur.com/CKKDi.png ;; ok, admittedly, the actual login window thingy could do with a slightly lighter background
06:45:48 <ehird> but it sure as hell beats http://upload.wikimedia.org/wikipedia/commons/e/ec/Xdm_Screenshot.png
06:54:10 <ehird> my mouse produces keyboard events in debian
06:54:33 <ehird> seemingly unpredictably
06:54:36 -!- Asztal has joined.
06:54:46 <ehird> also, holding down the middle button and dragging seems to select everything aand middle-click-paste it forever
06:55:26 * ehird decides to see if debian would like it better as a usb device
07:03:47 -!- mental_ has joined.
07:09:24 <ehird> ugh, and I am left again with the task of figuring out what file debian wants me to put x resources in
07:10:50 <coppro> to see what your mouse is doing
07:11:10 <coppro> also, I like playing with xev :D
07:17:28 <ehird> erm, what's the proper way to say yes in xresources files
07:19:04 -!- lament has quit (Read error: 110 (Connection timed out)).
07:21:24 -!- Azstal has joined.
07:25:40 <ehird> anyone know if there's xfontsel for xft?
07:33:59 -!- oerjan has quit ("leaving").
07:34:17 -!- Asztal has quit (Read error: 110 (Connection timed out)).
07:52:12 * ehird tries to find out where the hell the x11 cursor themes are
07:52:22 <ehird> when i enter my wm i get an ugly red cursor theme...
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:06:33 -!- mental_ has changed nick to lament.
08:09:38 * coppro wants to write a CSS renderer
08:11:09 <ehird> you need a whole layout engine for that
08:11:15 <ehird> a whole rendering engine
08:11:20 <ehird> can't really write just a css renderer...
08:11:25 <ehird> besides, the box model is hell
08:11:32 <coppro> you need a rendering engine for CSS
08:11:45 <ehird> yes, but it's integrally tied to the layout engine
08:11:48 <coppro> but you don't need to implement any markup or anything
08:11:52 <ehird> and really has to be part of it tbh
08:12:03 <ehird> and writing a layout engine is a super-massive-gigantic task
08:12:54 <ehird> #000 on #BBB terminal, dejavu sans mono 10pt
08:12:58 -!- MizardX has joined.
08:13:03 <ehird> with #888 desktop background
08:13:15 <ehird> (and like 5px #000 window borders from lwm...)
08:13:36 <coppro> It'll go data-from-some-source + CSS -> render
08:14:19 <ehird> 8pt is not soothing unless you have a low dpi screen with the dpi set to 96 or something that isn't really the dpi
08:14:41 <coppro> not 100% sure about the dpi
08:14:47 <ehird> #000 on #BBB is quite close to a book in non-bright lighting conditions
08:15:01 <ehird> coppro: well, css operates on html/xml
08:15:12 <ehird> which both parse to basically the same thing (xml's parse tree being a subset)
08:15:26 <coppro> CSS operates on data of any form
08:15:41 <coppro> it's often used on HTML/XML, but the spec need not be specific to them
08:15:52 <ehird> This document specifies level 1 of the Cascading Style Sheet mechanism (CSS1). CSS1 is a simple style sheet mechanism that allows authors and readers to attach style (e.g. fonts, colors and spacing) to HTML documents
08:16:08 <ehird> This specification defines Cascading Style Sheets, level 2 revision 1 (CSS 2.1). CSS 2.1 is a style sheet language that allows authors and users to attach style (e.g., fonts and spacing) to structured documents (e.g., HTML documents and XML applications). By separating the presentation style of documents from the content of documents, CSS 2.1 simplifies Web authoring and site maintenance.
08:16:15 <ehird> so css level 1 is html only
08:16:19 <ehird> css 2 may say structured documents
08:16:25 <ehird> but the syntax used, really, won't lend itself to anythhing else
08:16:32 <ehird> it totally is html/xml specific
08:16:48 <ehird> well as long as it has the same semantics i guess
08:17:09 <ehird> anyway, if you're calling it a "CSS renderer" that's a really bad name, as it undermines the immense difficulty of a layout engine :P
08:17:39 <coppro> I don't think it's necessarily html/xml-specific. It could be used for JSON, for all CSS cares
08:17:58 <coppro> (granted, there would be a limited subset of usable features, simply because JSON is less powerful)
08:18:20 <ehird> I'd highly recommend structuring it as XMLDoc → Rendered, though
08:18:24 <ehird> simply because everything else reduces to that
08:18:28 <ehird> due to the immense complexity of xml
08:18:43 <ehird> and because xml gives you xhtml, which is very common in practice and so probably should be supported
08:19:00 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
08:19:04 -!- MizardX has joined.
08:19:38 <coppro> The input will be through an API, I'm thinking
08:19:44 <coppro> so writing an XML plugin could be done
08:21:25 <ehird> needless complexity if xml is a subset of all the others
08:21:36 <ehird> if you gave the base stuff simple enough names and made the rest optional you wouldn't even have to call it XML
08:21:47 <coppro> Then I don't need to write or use an XML parser
08:21:54 <ehird> you don't here either
08:22:07 <ehird> i'm just saying to use one data structure, and have it be XML's structure
08:22:20 <ehird> then "plugins" simply become render(parseXML(...))
08:22:22 <coppro> The data structure will be binary, though, not text
08:22:26 <ehird> or render(parseJSON(...)) all returning a Doc
08:22:28 <ehird> coppro: it'll be a struct
08:22:36 <ehird> are you unable to comprehend that XML has an internal structure?
08:22:47 <coppro> and CSS does require a hierarchical structure
08:22:57 <coppro> so it's not like it will be all that dissimilar from XML
08:23:01 <ehird> tag = namespace+attributes[name→str]+children[tag]+...
08:23:12 <ehird> that way, for e.g. json, you'd just set tag.name and tag.children
08:23:20 <ehird> but the thing is, "tag" there is xml
08:23:31 <ehird> but i don't see where plugins come into it
08:23:41 <coppro> well, I simply meant it would not be standalone
08:23:45 <ehird> has C++ addled your mind so much that you can't comprehend the idea of a function returning a Doc or whatever?
08:25:10 <coppro> I think it will render to an OpenGL surface
08:25:32 <ehird> coppro: seriously? rendering engine is separate from the actual display
08:25:44 <ehird> for one thing, the layout constantly drastically changes in quite a lot of renderings
08:25:53 <ehird> especially if you're loading the content (not the css) incrementally
08:26:07 <coppro> ehird: What do you recommend I choose as the target data then?
08:26:23 <coppro> an OpenGL surface seems like the LCD
08:26:25 <ehird> well, that's up to you, innit :P I'm no expert in writing t hem, I just know a little about how they work
08:26:41 <ehird> coppro: well, let's put it this way
08:26:43 <ehird> resize a browser window
08:26:47 <ehird> do you think it totally re-renders the page?
08:26:57 <ehird> my functional brain tells me to make it based on fluid constraints
08:27:06 <ehird> elements pushing away from other elements, etc
08:27:12 <coppro> sure, but what's that got to do with OpenGL?
08:27:22 <ehird> because you don't "render to an opengl surface"
08:27:28 <ehird> you render to an abstract data structure, then draw that
08:27:44 <ehird> coppro: btw opengl has problems with the idea of a "pixel"
08:28:11 <coppro> I'll need to look into this I guess
08:28:23 <coppro> (I would anyways, but now I need to look into it more!)
08:28:33 <ehird> coppro: if you come out of this anything other than gibbering I will be astounded.
08:29:17 <ehird> anyone know the proper way to change x11 resolution in this hal day and age
08:29:54 <ehird> no, that's on the fly
08:30:02 <ehird> I mean changing the initial resolution permanently
08:32:16 <ehird> but that's so... obsolete...
08:33:49 * ehird wonders if he can get away with just
08:34:13 <ehird> SubSection "Display"
08:34:27 <coppro> you can get it to generate the current config for you
08:34:32 <ehird> yes, but that stops the hal stuff
08:36:56 <ehird> well that just made x give up
08:38:54 <coppro> oh, you can also put xrandr in your x startup script
08:39:32 <ehird> oh, since i rebooted virtualbox is now just letting me resize the vm to my preferred window size
08:39:37 <ehird> and adjusting the resolution appropriately
08:39:48 <coppro> oh, you installed the extensions
08:40:04 <ehird> yeah, the OSE ones from debian's repository. works with the proprietary version :P
08:40:30 * ehird wonders what browser to stick on this thing
08:41:13 <ehird> firefox is shitty, midori has weird interface quirks, arora had some annoying glitch last time I used it
08:41:57 <ehird> alias sag='sudo apt-get'
08:41:57 <ehird> alias sagi='sudo apt-get install'
08:41:57 <ehird> alias acs='apt-cache search'
08:42:39 * ehird wonders why x11 mouse acceleration sucks so much
08:42:52 <ehird> coppro: has konqueror even switched over to webkit yet
08:43:01 <ehird> or is it still KH"It is 2003"TML
08:43:04 <coppro> haven't used it in ages
08:43:10 <ehird> also, I kinda dislike the whole kitchen-sink thing :P
08:43:15 <ehird> and the mass of KDE dependencies thing
08:43:24 <coppro> I agree in principle but not in practice
08:43:41 <ehird> i've never ever thought "I wish I could just type in a file URL now and start browsing my files"
08:43:51 <coppro> except in Windows where the shell doesn't exist
08:43:55 <coppro> but that doesn't count
08:44:13 <coppro> (practice being the fact that I use KDE, so a mass of KDE dependencies is largely a non-event)
08:46:03 <ehird> wow, andrew cooke packed even more text into his site: http://www.acooke.org/
08:46:09 <ehird> ...and dropped the lowercase fun!
08:46:25 -!- Slereah_ has joined.
08:46:30 * coppro feels an urge to link the CSS Zen Garden
08:48:03 <ehird> css zen garden was fun.
08:48:10 <ehird> oh god, arora still has the glitch
08:48:13 <ehird> the maddening begins now
08:48:35 <ehird> i just realised i'm using klingonux
08:50:50 <ehird> http://i.imgur.com/NLThT.png bask in the soothing colours and fonts
08:51:02 <ehird> coppro: arora semi-randomly either underlines or non-underlines underlined links
08:51:12 <coppro> http://createyourproglang.com/ roffffffffl
08:51:14 <ehird> it makes the whole thing feel unstable
08:51:33 <coppro> ehird: is there a bug filed?
08:52:04 <ehird> coppro: rather lame marketinig site, marc-andré cournoyer's little language implementations are cool thouough
08:52:16 <ehird> i mean they're all llvm and stuff, so probably the book has that too
08:52:22 <ehird> coppro: i don't know whether there's a bug filed
08:52:28 <coppro> ehird: it was linked from that acooke guy
08:52:31 <ehird> probably most people can't reproduce it
08:52:38 <ehird> andrew cooke is the one who proved malbolge TC
08:52:45 <ehird> he's the one who did hello world in it
08:52:49 <ehird> mixed up my momentous tasks there
08:52:53 <ehird> http://i.imgur.com/NLThT.png
08:52:55 <coppro> ehird: if you scroll down, it says it does LLVM
08:53:36 <coppro> it's just funny because it looks like Plain English in terms of quality, but clearly the guy actually knows what he's talking about
08:53:59 <ehird> yeah, i think it's aimed at the ruby post-ironic hipstercore market
08:57:02 -!- Slereah has quit (Read error: 110 (Connection timed out)).
08:58:37 <ehird> it's cool how lwm's resize widgets show the size in pixels for everything but terminal windows, where it shows lines/cols
08:58:59 <coppro> Pretty sure KDE can do the same
08:59:09 <ehird> it's still awesome
08:59:12 <coppro> I have it turned off for normal window resizes
08:59:47 <ehird> lwm seems to be a pretty rad window manager
09:00:17 <ehird> minimalist, not-entirely-wacky mouse controls, window hiding (goes to a right-click menu on the root window), and easy program triggering (buttons 1 and 2 on the root window; just 2 by default)
09:00:38 <coppro> huh, there doesn't appear to be a plugin for that. I thought there was.
09:00:49 <coppro> There is a neat effect that highlights areas of the screen that get rerendered
09:01:09 <coppro> so you can see how frequently your application is painting individual areas
09:01:41 <ehird> ugh, i think vbox is telling vm my screen dpi
09:01:45 <ehird> thus weirding everything up
09:01:51 <coppro> it even manages to refrain from counting the mouse movements
09:01:56 <coppro> I should leave it like this; it's trippy
09:02:50 <ehird> did you hear that I'm using xdm and it's not breaking my eyes? pretty astonishing news imo
09:02:53 <ehird> didn't know it was possibble
09:03:15 <coppro> the dm doesn't do very much, really
09:04:04 <ehird> http://upload.wikimedia.org/wikipedia/commons/e/ec/Xdm_Screenshot.png
09:04:08 <ehird> this is what xdm normally looks like
09:05:15 <ehird> http://i.imgur.com/CKKDi.png this, but with a lighter background (#AAA) on the actual login window, is what i'm using
09:05:24 <coppro> the problem is obvious. Log in and use startx
09:06:22 <ehird> i want x to start auto after login, no real easy way to do that withotu a dm
09:06:28 <ehird> might as well change the display mode while we're at it
09:06:47 <coppro> put it in your .bashrc?
09:06:55 <ehird> and ruin all shells?
09:07:06 * ehird considers trying chromium
09:07:10 <ehird> but i like my window decorations...
09:07:51 <coppro> no, just make all shells start an X server when you log in!
09:08:15 <coppro> (actually, if you silence the command and run it in the background, it will work fine)
09:08:43 <coppro> startx will fail because :0.0 is already in use, and you're happy
09:09:28 <ehird> do you think an ubuntu repo for chromium will wowrk?
09:09:33 <ehird> eh, i'll install a deb first
09:09:35 <ehird> just to see if i want it
09:10:02 <coppro> ba-ba-ba-bum bum ba-ba-bum
09:10:19 <ehird> i need to sleep soon
09:11:10 <coppro> roughly this time last night
09:11:16 <coppro> sleeping != disconnecting
09:11:29 <coppro> in fact, me being disconnected is usually a good indication I'm not sleeping
09:12:07 <ehird> it's 9am here for me, so that's connfusing
09:12:15 <ehird> all i know is i didn't sleep the whole night
09:12:23 <ehird> and i woke up at like 4pm the day before
09:12:26 <ehird> problematic for brain.
09:22:21 <ehird> Ehh. Chrome would be perfect for this if I could make it use the native GTK theme, except for the main background.
09:26:10 * ehird tries to remember the name of that simple program that did alt+f2 launching
09:27:32 -!- MizardX- has joined.
09:27:37 * coppro goes to sleep, expecting Mars to wake him up... ba-ba-ba-bum bum ba-ba-bum
09:27:58 -!- MizardX has quit (Read error: 104 (Connection reset by peer)).
09:28:05 -!- MigoMipo has joined.
09:28:09 -!- MizardX- has changed nick to MizardX.
09:34:13 -!- ehird has quit.
09:40:54 -!- jpc has quit (Read error: 60 (Operation timed out)).
10:00:55 -!- uorygl has quit (Remote closed the connection).
10:00:59 -!- uorygl has joined.
10:16:44 -!- adam_d has joined.
10:32:54 -!- ais523 has joined.
10:38:31 <ais523> gah, some spammer got around my mental spam filters by writing in first person plural
10:38:40 <ais523> and I read a whole half a sentence before I realised it was spam
11:09:34 -!- pikhq has quit (Read error: 104 (Connection reset by peer)).
11:15:59 -!- pikhq has joined.
11:29:14 -!- kwertii has quit ("bye").
11:41:28 <AnMaster> ais523, merry xmas btw I guess (since I will be away tomorrow)
11:42:04 <AnMaster> (and you will probably be away the day after that?)
11:42:32 <ais523> actually, I'll more likely be here to be with the family than I would be otherwise, I think
11:42:42 <ais523> they'll all be here I suspect, due to not being able to fit everyone anywhere else
11:42:55 <ais523> btw, what day is Christmas in your country?
11:43:00 <ais523> it's 25th here in the UK
11:43:04 <ais523> but apparently 24th in Germany
11:43:05 <AnMaster> ais523, presents are on the 24th
11:43:14 <AnMaster> due to it being the Christmas eve
11:43:24 <ais523> here, they're technically the 26th but everyone but very religious people ignores that
11:43:27 <AnMaster> giving the presents on the Christmas day? What a strange idea
11:43:34 <ais523> and gives on the 25th instead
11:43:50 <AnMaster> ais523, the 25th being the Christmas day or Christmas eve?
11:44:06 <ais523> christmas eve is when half of people traditionally /buy/ the presents
11:44:10 <ais523> due to having left it until the last minute
11:44:16 <ais523> and the shops stay open late and triple their prices
11:44:22 <AnMaster> also there is a simple explanation for it. Work hours rules for Father Christmas :P
11:44:44 <AnMaster> of course in Russia he uses a subcontractor iirc
11:44:54 <AnMaster> St. Nicolaus or something iirc?
11:45:10 <ais523> he's secretly helped by all the dads in the country
11:45:20 <ais523> it's how he manages to get into houses, they unlock the door for him
11:45:31 <ais523> either that, or deliver the presents themselves if he can't route everywhere in time
11:45:39 <ais523> the travelling salesman problem hasn't been solved efficently yet...
11:45:43 <AnMaster> ais523, well yes, this is related to population growing exponentially, while his capacity growing geometrically
11:46:05 <AnMaster> (that that rule is about food supply is a common misconception)
11:46:54 <AnMaster> ais523, isn't an A* search (or whatever the name was) reasonably efficient for the traveling salesman problem
11:47:02 <AnMaster> oh wait no, that was just route finding
11:47:12 <ais523> A* is decent for route finding if you know both endpoints in advance
11:47:23 <ais523> Dijkstra if you only know one endpoint in advance
11:47:56 <AnMaster> ais523, can't you do some sort of dynamic programming or such to make it reasonably manageable?
11:48:05 <ais523> for NetHack routing you really need an algorithm that works even if you know zero endpoints in advance, and the map is changing meanwhile
11:48:07 <AnMaster> iirc there was some xkcd about it
11:48:19 <ais523> AnMaster: oh, there are algorithms which do good enough for practical uses
11:48:42 <ais523> but if you want /the best/ answer, you can't do it quickly with current maths
11:49:01 <AnMaster> ais523, hm btw can you explain what that thing about NP complete problems being reducible to each other is about?
11:49:19 <AnMaster> is it just the same as you can express it as a variant of the other problem?
11:49:52 <ais523> yes, well with NP-complete problems, the idea is that you can set up one problem in such a way that solving it would be a solution to another as well
11:49:58 <ais523> pretty much like compiling esolangs into each other
11:50:19 <AnMaster> so then if you solved one of them efficiently you could just use that to solve all the other ones?
11:50:37 <AnMaster> or is that only true for some disjunct subsets of the NP complete problems?
11:51:27 <AnMaster> I mean, could you for example use the quantum integer factorization algorithm to solve the travelling salesman problem?
11:52:46 <ais523> NP-complete are all in the same computational complexity class
11:52:58 <ais523> sort-of the same way Turing-complete works
11:53:03 <AnMaster> but does that mean that they can be reduced to each other
11:53:12 <ais523> NP is "below" in the sense that NP-complete can be reduced to NP-complete, or anything else in NP
11:53:27 <AnMaster> and if it did, would that mean that all problems in P were also reducible to each other?
11:53:39 <ais523> no, I don't see why that would be the case
11:53:46 <ais523> it's not the case that all NP is reducible to each other
11:54:01 <ais523> just that all NP-complete is reducible to all NP (including all NP-complete)
11:54:46 <ais523> all P is trivially in NP, by the way
11:54:54 <ais523> the whole P = NP problem is to prove that it's also the other way round
11:55:08 <ais523> (or alternatively show that it isn't)
11:55:44 <AnMaster> ais523, I'm having trouble parsing "<ais523> just that all NP-complete is reducible to all NP (including all NP-complete)"
11:56:03 <AnMaster> as in, do you mean there is a common "root problem" that all NP complete problems can be reduced to?
11:56:23 <ais523> yes, any NP-complete problem
11:56:58 <ais523> hmm... just like any TC language can implement any program that a turing machine can run
11:57:04 <ais523> that includes other TC languages
11:57:10 <ais523> but also, things like BF-PDA
11:57:26 <ais523> which isn't Turing/complete/, even though you can run it on a Turing machine
11:57:31 <ais523> NP-completeness is much the same
11:57:55 <AnMaster> ais523, so then you could in theory express traveling salesman in terms of integer factorization?
11:57:56 <ais523> if something is in NP, then you can 'emulate' it with any NP-complete problem (as in, a solution to the second is a solution to the first)
11:58:09 <ais523> given that those are both well-known, it's probably already been done
11:58:29 <ais523> well, ordinary integer factorization probably isn't NP-complete
11:58:39 <AnMaster> ais523, and then solve it quickly with Shor's algorithm?
11:58:58 <ais523> AnMaster: integer factorization is in NP, but not known to be NP-complete
11:59:05 <ais523> it's sort-of, dupdog range in our analogy
11:59:22 <ais523> http://esolangs.org/wiki/Dupdog
11:59:55 <ais523> we (as in this channel) think it's probably sub-TC, but are unable to prove it
12:00:01 <ais523> or at least, were last time the subject came up
12:00:36 <AnMaster> ais523, it says unknown computational class for dupdog
12:01:11 <ais523> dupdog's much nastier than slashes
12:01:22 <AnMaster> which I did spend quite a bit of thought about before he decided to try it
12:01:36 <AnMaster> and I didn't come up with any sensible way to do non-trivial loops for example
12:01:39 <ais523> I tried /// once; I failed, but I could see sort-of how to do it and think the issue was just bugs in my compiler, rather than a fundamental failure of the method
12:02:34 <ais523> AnMaster: basically, quining
12:02:35 <AnMaster> hm sensible is not the opposite of insensible is it? And aspell suggests unsensible doesn't exists
12:02:49 <ais523> oppositie of sensible is senseless, or sily
12:03:16 <ais523> neither's an exact opposite; English's weird like that
12:03:22 <AnMaster> and inflammable means something doesn't burn easily. Who said English had to make sense...
12:03:33 <ais523> AnMaster: no, inflammable means it does burn easily
12:03:42 <ais523> flammable also means it does burn easily
12:03:55 <ais523> AnMaster: hmm... stating a blatantly wrong fact then whooshing when people correct you?
12:04:08 <AnMaster> ais523, I thought it was obvious it was sarcasm
12:04:17 <ais523> no, it wasn't obvious, it's a common mistake
12:04:20 <AnMaster> there was an iwc joke about that some time ago, forgot you didn't read it
12:04:35 <ais523> probably only among actual English people, though, foreigners are more likely to look up what a word means
12:04:39 <ais523> whereas the natives just guess
12:05:29 <AnMaster> ais523, btw have you heard about Mike Riley
12:06:09 <AnMaster> ais523, planning to commit suicide, depression. Ehird and me has been working on trying to get him not to do it and trying to contact people who might help
12:06:48 <ais523> you could try contacting the police where he lives
12:07:32 <AnMaster> ais523, yeah, major depressions, getting worse very time,
12:07:46 <AnMaster> ais523, also that is Las Vegas, a bit hard to find him there I imagine
12:08:14 <AnMaster> ais523, underlying cause he said was a "major birth defect" and didn't want to get into more details
12:10:32 <ais523> the issue is, I'm really not sure what to do beyond that
12:10:33 <AnMaster> ais523, but yes, we probably will contact them today. Tried various other ways first. (Why do good samaritans not have email except in a few places, none of them being Nevada...)
12:12:07 <AnMaster> wait that sed expression won't work
12:12:41 <AnMaster> anyway yes police probably is the only way left
12:26:21 -!- pikhq has quit (Read error: 104 (Connection reset by peer)).
12:26:57 -!- pikhq has joined.
12:29:45 <AnMaster> ais523, I will need your help with formal English in a bit
12:44:19 -!- FireFly has joined.
13:23:15 -!- Deewiant has quit (farmer.freenode.net irc.freenode.net).
13:23:15 -!- fungot has quit (farmer.freenode.net irc.freenode.net).
13:23:15 -!- dbc has quit (farmer.freenode.net irc.freenode.net).
13:23:15 -!- yiyus has quit (farmer.freenode.net irc.freenode.net).
13:23:15 -!- Cerise has quit (farmer.freenode.net irc.freenode.net).
13:24:29 -!- Cerise has joined.
13:24:29 -!- yiyus has joined.
13:24:29 -!- dbc has joined.
13:24:29 -!- fungot has joined.
13:24:29 -!- oerjan has joined.
13:24:29 -!- Deewiant has joined.
13:25:36 -!- oerjan has left (?).
13:25:36 -!- oerjan has joined.
13:29:45 <oerjan> <AnMaster> ais523, something for oerjan?
13:29:56 <oerjan> i'm pretty sure i've pondered dupdog
13:31:06 <oerjan> <ais523> AnMaster: hmm... stating a blatantly wrong fact then whooshing when people correct you?
13:31:20 <oerjan> clearly the whoosh here consists of runaway flames...
14:03:23 <AnMaster> oerjan, btw d&d has been timing out for me today. is it down for you too?
14:06:06 <oerjan> i haven't checked, since it's wednesday
14:06:31 <AnMaster> oerjan, yeah I wanted to check the annotation, think I forgot to read it yesterday (the annotation, not the strip)
14:06:41 <oerjan> except i read yesterday's a bit late, this morning, and it was fine
14:18:02 -!- Azstal has quit (Read error: 110 (Connection timed out)).
15:05:05 -!- benuphoenix has joined.
15:07:09 -!- benuphoenix has left (?).
15:13:33 -!- BeholdMyGlory has joined.
15:24:38 <oerjan> AnMaster: d&d is loading again
15:27:41 -!- wawl has joined.
15:27:44 -!- wawl has left (?).
15:27:49 -!- wawl has joined.
15:29:09 -!- wawl has left (?).
16:31:09 -!- soupdragon has joined.
16:40:03 -!- Pthing has joined.
17:03:15 -!- sebbu3 has changed nick to sebbu.
17:04:16 -!- MizardX has quit (Connection timed out).
17:05:27 -!- jpc has joined.
17:18:51 -!- MizardX has joined.
17:20:44 -!- soupdragon has quit (Read error: 104 (Connection reset by peer)).
17:20:48 -!- soupdragon has joined.
17:50:17 -!- soupdragon has quit ("* I'm too lame to read BitchX.doc *").
17:50:34 -!- soupdragon has joined.
18:06:19 -!- Pthing has quit ("Leaving").
18:06:34 -!- Pthing has joined.
18:40:20 -!- adam_d_ has joined.
18:51:08 -!- soupdragon has quit ("* I'm too lame to read BitchX.doc *").
18:58:30 -!- adam_d has quit (Read error: 110 (Connection timed out)).
19:01:05 -!- MizardX- has joined.
19:01:22 -!- MizardX has quit (Read error: 104 (Connection reset by peer)).
19:01:41 -!- MizardX- has changed nick to MizardX.
19:02:50 -!- MizardX- has joined.
19:19:41 -!- MizardX has quit (Success).
19:19:45 -!- MizardX- has changed nick to MizardX.
19:20:56 -!- soupdragon has joined.
19:52:07 -!- oerjan has quit ("Good night").
20:04:10 -!- ehird has joined.
20:04:19 <ehird> "Two young men caught cycling with no clothes on have escaped charges of offensive behaviour, but received a warning to wear protective headgear."
20:05:02 <ais523> and technically, going around naked is legal just so long as nobody complains
20:05:20 <ehird> where it's illegal
20:06:01 <ehird> 03:44:06 <ais523> christmas eve is when half of people traditionally /buy/ the presents
20:06:01 <ehird> that's worryingly close to where i'm heading...
20:06:22 <ais523> I buy and give the presents at any time at random, more or less
20:06:27 <ais523> when I can think of something worth giving
20:06:37 <ehird> christmas is more trouble than it's worth
20:06:37 <ais523> I have a rather non-traditional approach to that sort of thing....
20:06:58 <ais523> secular christmas confuses me, I think it was invented by shops to sell useless stuff
20:07:06 <ais523> and it lasts far too long nowadays, months in some cases
20:07:53 <lament> why is that confusing?
20:08:10 <ais523> oh, I believe too much in economics
20:08:15 <ais523> as in, I don't get why people buy overpriced stuff
20:08:23 <ehird> re santa: http://www.main.com/~anns/other/humor/physicsofsanta.html
20:08:33 <lament> ais523: you need to make people want to buy your stuff
20:08:48 <lament> ais523: one good way to do that is to change the whole culture to make this stuff more desirable
20:08:49 <ehird> ais523: you buy an overpriced item if there is no suitable alternative
20:08:59 <lament> ais523: which is what happens with christmas
20:09:09 <ais523> ehird: then it arguably isn't overpriced
20:09:44 <ehird> i'm so happy that lwm does clever window placement
20:10:10 <lament> ais523: also, it's not overpriced
20:10:24 <ais523> lament: that's a very vague statement
20:10:28 <ais523> some things are overpriced, some aren't
20:10:31 <ais523> yet people seem to like buying both
20:10:42 <lament> not sure what that has to do with christmas
20:10:51 <lament> you said christmas confused you
20:11:04 -!- Asztal has joined.
20:11:10 <lament> buying overpriced stuff is completely orthogonal to christmas
20:11:14 <ais523> oh, people seem a lot more inclined to buy useless things at christmas
20:11:16 <lament> christmas is about buying *useless* stuff
20:11:22 <lament> useless, not overpriced
20:11:30 <ais523> useless is overpriced by definition
20:12:03 <ehird> ais523: do you know of any decent non-epiphany webkit browser for linux?
20:12:09 <lament> ais523: i thought "overpriced" meant "above market value"
20:12:11 <ehird> not arora, not midori, not google chrome
20:12:27 <ais523> ehird: no, my browser knowledge is rather small
20:12:37 <ais523> I like the way you describe chrome as non-decent, though
20:12:50 <ehird> I do wish lwm let me raies a window to the top by clickiing on its contents, though, not just the title bar...
20:12:59 <ais523> apparently it installs a cronjob that add's google's deb repo to the repo list (on Ubuntu at least)
20:13:01 <ehird> ais523: no, chrome is alright
20:13:11 <ehird> also, yes, it's rather weird, but the browser itself is fine
20:13:15 <ehird> except for a few things
20:13:34 <ehird> best I've found so far, though
20:13:42 <ais523> what issues do you have with it?
20:14:19 <ehird> well, the update thing doesn't sit well with me of course; i can't use the ubuntu chromium ppa, which I'd prefer to, because I'm on debian (sid)
20:14:35 <ehird> because i have it set to use my WM's window decorations, because I like them,
20:14:44 <ehird> the background of the tab bar looks kinda weird
20:14:50 <ehird> and the tabs are too close to the title bar
20:14:55 <ehird> i could fix this partly
20:15:06 <ehird> if I made the background of the tab bar the colour of my WM's decorations
20:15:12 <ehird> (black focused, grey unfocused)
20:15:20 <ehird> but then I'd expect to be able to focus the window, drag the window, etc by it
20:15:27 <ehird> and besides I can't figure out how to do it
20:15:30 <ehird> ais523: oh, did you know?
20:15:31 <ais523> I'd love to stay and talk longer, we keep missing each other
20:15:34 <ehird> mike riley is going to commit suicide...
20:15:42 <ehird> just remembered to tell you
20:15:44 <ais523> AnMaster told me, and wrote an email to the police
20:16:21 <ehird> ais523: anyway, bye
20:16:33 <ais523> gah, just waiting for CPAN to finish
20:16:42 <ais523> stupid CPAN, I keep forgetting to check for prompts
20:16:50 <ehird> ais523: here, let me give you two screenshots first that you don't care about!
20:18:15 <ehird> ais523: http://i.imgur.com/ELeEq.png believe it or not, this is actually xdm
20:18:32 <ehird> yes, horrible-pseudo-3d-italic-blue-text-with-rubbish-logo-to-the-side-and-the-awful-X11-checkered-background xdm
20:18:37 <ehird> but I tamed the beast!
20:18:43 <ehird> http://i.imgur.com/NLThT.png
20:18:43 <ais523> is the grey pattern on the title bar correct?
20:18:53 <ehird> and this is lwm with a urxvt
20:18:57 <ehird> and sooothing colours and fonts
20:19:07 <ehird> ais523: you mean the gradient on the OS X window?
20:19:17 <ais523> I have fond memories of xdm, anyway
20:19:27 <ehird> what do you mean by "correct"?
20:19:34 <ais523> looking exactly as in Mac OS X
20:19:37 -!- ais523 has quit (Remote closed the connection).
20:19:40 <ehird> oh, that's my VM window
20:35:11 -!- ehird_ has joined.
20:35:16 <ehird_> Good morning, America!
20:35:34 <ehird_> Good morning, good morning, good morning.
20:35:38 <ehird_> Or something like that, anyway.
20:35:39 -!- ehird_ has left (?).
20:35:43 -!- ehird_ has joined.
20:35:49 -!- ehird_ has left (?).
20:35:56 -!- ehird_ has joined.
20:37:07 -!- adam_d has joined.
20:39:05 <ehird_> Mike Riley update: he's seeing a therapist
20:40:00 <ehird_> being in here shatters the psyches of even the strongest men
20:42:08 -!- atrapado has quit (Read error: 104 (Connection reset by peer)).
20:44:09 <Deewiant> ehird_: As in "is now seeing" or "has been seeing"
20:46:10 <ehird_> In other news, /set theme colorless makes irssi nice.
20:46:31 <ehird_> Still wish it somehow integrated with my terminal's scrollbar, but you know, that's just too much to ask for.
20:52:38 -!- adam_d_ has quit (Read error: 110 (Connection timed out)).
20:55:41 <pikhq> ehird_: How'd you set XDM to look nice, anyways?
20:56:10 <ehird_> Firstly, put "xsetroot -solid rgb:8/8/8" or whatever you want in /etc/X11/xdm/Xsetup.
20:56:34 <ehird_> Then, well, look at Xresources in the same directory.
20:56:53 <ehird_> I set the fonts, changed borderWidth, frameWidith and innerFramesWidth to 0,
20:57:02 <ehird_> shdColor and hiColor to black, background to #AAAAAA (both of them).
20:57:05 <ehird_> And some other tweaks.
20:57:18 <ehird_> Oh, and commented out the lines that add the Debian logo.
20:57:30 <ehird_> And changed greeting to just CLIENTHOST.
20:58:46 <ehird_> pikhq: You still using Conkeror as your browser? I'm <-> this close to surrendering to the Gecko forces... but yeck.
20:59:31 <pikhq> ehird_: Yeah, still using it.
21:01:06 * pikhq wonders if xft: fonts work for this
21:01:26 <ehird_> I have -adobe-helvetica thingies in the font, and yet it uses the vector version.
21:01:27 <pikhq> It looks like, since I've got xft in my USE flags, "yes".
21:01:35 <ehird_> Couldn't tell you why, but clearly it's using Xft for every font.
21:01:41 <ehird_> So I assume an explicit xft: will work.
21:02:20 <ehird_> You know, -*-fixed-bold-*-*-*-15-*-*-*-*-*-*-* isn't such a bad font.
21:02:25 <ehird_> I have it for my titlebars here.
21:02:46 <pikhq> man page agrees...
21:04:02 <ehird_> RESTRICTIONS Xedit is not a replacement to Emacs.
21:05:15 <ehird_> You know, "foo 2>| bar" should work.
21:06:04 <pikhq> Hmm. Just about got it nice.
21:06:24 <ehird_> I can give you my Xresources file if you're a fan of gray and Helvetica.
21:06:26 <pikhq> xdm is capable of *not* looking like shit. :)
21:06:48 <pikhq> Gray and Dejavu Sans for XDM.
21:07:34 <ehird_> http://i.imgur.com/ELeEq.png
21:07:43 <ehird_> I think that _is_ DejaVu Sans.
21:07:49 <ehird_> xdm must be substituting it for Helvetica.
21:07:53 <pikhq> Looks like it, actually.
21:07:57 <ehird_> Guess I should put Sans in directly.
21:08:07 <ehird_> I somehow like the serifs on the hostname.
21:08:11 <ehird_> Breaks the monotony. :P
21:08:51 <ehird_> I wish bash-completion wasn't so darned slow.
21:09:13 <ehird_> Incidentally, I bet sid is stabler than Gentoo. :P
21:10:16 <pikhq> Knowing Debian? Probably.
21:10:32 <ehird_> I installed this system yesterday and there haven't even been any updates!
21:10:32 <pikhq> "We've only tested it for a couple of months! Straight!"
21:11:09 <AnMaster> <ehird_> I wish bash-completion wasn't so darned slow. <-- it isn't?
21:11:19 <AnMaster> well it is first time after boot IME
21:11:33 <ehird_> Sure it is, like .3s delay completing just a lowly filename.
21:11:47 <AnMaster> ehird_, are you running native or in VM?
21:11:56 <ehird_> VM, but on properly virtualising hardware.
21:12:32 <ehird_> 351 megs of ram, 299 free
21:12:39 <ehird_> (that is the -/+ buffers/cache one)
21:12:44 <ehird_> (obviously only 31 megs free in the normal line)
21:14:08 <ehird_> Does anyone know how much stuff supports XDG's where-to-put-dotfiles stuff?
21:14:25 <ehird_> There's a horrible LD_PRELOAD hack that rewrites all writes to ~/.foo to it, too, I think.
21:14:32 <ehird_> But that's a bit too cowboy for my liking.
21:15:15 <ehird> Ugh, my control key has become stuck in the VM again.
21:15:40 <ehird> Did you know that urxvt has menus?
21:15:55 -!- ehird_ has quit (Remote closed the connection).
21:16:11 <ehird> And xterm has a ton of menus, but that's widely-known.
21:16:19 <ehird> But urxvt's include "evaluate Perl expression".
21:16:47 <ehird> Ah. It seems urxvt's menus do not agree with my WM.
21:18:29 <ehird> You know, stock Debian sid boots quite quickly.
21:19:45 -!- pikhq has quit (Read error: 54 (Connection reset by peer)).
21:20:24 <soupdragon> More recently, Gregory J. Chaitin of IBM has found arithmetic propositions whose truth can never be established by following any deductive rules.
21:20:53 -!- pikhq has joined.
21:22:08 <pikhq> And apparently the urxvt menus don't agree with *anything* X related.
21:22:25 <ehird> Should have done it in an Xnest
21:22:46 <ehird> Gah, Debian. Why will you not give me lovely upgrades?
21:24:42 * ehird tries out conkeror
21:27:02 <ehird> Conkeror review: "Meh."
21:27:36 * ehird decides to try trimming down iceweasel
21:37:15 <pikhq> It's the least annoying interface I've found, but the Gecko bit is t3h suck.
21:42:03 <ehird> AnMaster: Because all the other browsers suck more.
21:42:30 <ehird> pikhq: do you know a lightweight program that handles Alt-F2 program<enter> launching? I found one once but have forgotten it
21:43:02 <pikhq> That's a part of Ratpoison, you see...
21:45:10 <ehird> Deewiant: No, it was literally: input box, runs it in a shell
21:45:25 <ehird> dmenu has completion and stuff, which I don't need.
21:45:39 <ehird> And I want something in the middle of the screen; this was.
21:45:50 <ehird> It wasn't anything well-known, I don't think. Which is why this is probably hopeless.
21:47:03 <ehird> "Ratpoison? But that's SINFUL!"
21:47:48 <ehird> Ratpoison is designed for a certain workload.
21:48:05 <ehird> That workload consists of Emacs, one or two unrelated terminals, and a browser.
21:48:20 <ehird> And a screen that isn't too big (because otherwise everything will be in the corners.)
21:48:35 <ehird> Given those, since they won't interact much, ratpoison is probably close to the most optimal window manager.
21:48:52 <ehird> Heck, Emacs' buffer management is strikingly similar to Ratpoison.
21:51:02 <ehird> So, I figured out why my X11 cursor is red.
21:52:08 -!- poiuy_qwert has joined.
21:52:15 <pikhq> My workload consists of a terminal, a browser, and possibly a virtual machine or two.
21:52:24 <pikhq> For that, Ratpoison is just about optimal.
21:52:39 <ehird> Gtk's Mist style is pooping on my grey colour scheme party.
21:52:45 <ehird> I guess I should tweak it to behave.
21:52:51 <pikhq> I would *not* want to use it for heavier workloads.
21:53:09 <ehird> (lwm is pooping on my I-like-clicking-on-window-contents-to-raise party, too)
21:55:01 <ehird> Is it actually possible to disable xpdf's ugly menus?
21:58:26 -!- FireFly has quit (Remote closed the connection).
21:58:35 -!- FireFly has joined.
21:58:57 <pikhq> Better question, though: is it possible to make Motif not ugly?
21:59:08 <ehird> Yes. Want a page that shows you how?
21:59:30 <pikhq> ... I may start using Motif programs. XD
21:59:33 <ehird> http://www.nedit.org/technotes/looks-1.php
21:59:38 <ehird> http://www.nedit.org/technotes/looks-before.gif
21:59:41 <ehird> http://www.nedit.org/technotes/looks-after.gif
22:00:03 <ehird> The issue is that usually, Motif programs have a bad interface beyond the looks :P
22:01:11 <pikhq> Only so much that can be done with 20 year old programs. :P
22:01:58 <ehird> For xpdf, I'd remove the page navigation buttons (useless, I have a scrollbar for that).
22:02:07 <ehird> I don't need the print or the help buttons either, but I can live with those.
22:02:15 <ehird> The zoom and search items are probably good.
22:02:18 -!- ehird_ has joined.
22:02:40 <pikhq> Also, the icons are fugly.
22:02:51 <pikhq> Could probably be fixed just by making them vector icons, though.
22:02:59 <ehird> Yeah, no biggie though imo.
22:03:17 <ehird> Also, I'm not sure how to adapt that nedit page to non-edit programs; it uses "nedit" as the resource. What's it inherited from? Motif?
22:03:42 <ehird> Also, what's the difference between foo*bar and foo.bar? I forget.
22:06:18 <ehird> http://toastytech.com/guis/win7101apps.png
22:06:18 <ehird> Windows 7: the best platform to run your Windows 1.01 programs.
22:07:08 <ehird> Grr; I may just patch lwm to let me raise a window by clicking inside it.
22:07:14 <ehird> I think it's non-reparenting; that's likely to be the problem.
22:09:21 <ehird> Yeah, I'm totally tempted to switch to Debian sid.
22:10:38 <AnMaster> <ehird> http://toastytech.com/guis/win7101apps.png <-- that old calculator renders incorrectly it seems
22:10:56 <ehird> It was probably like that in Windows 1; not exactly the most polished OS.
22:11:15 <AnMaster> the text going out of the button
22:11:15 <ehird> Take a look at the menus in the Windows 1 programs.
22:11:19 <ehird> It's even retaining the non-antialiasedness.
22:11:25 <ehird> AnMaster: Yeah, that's what I meant.
22:11:41 <pikhq> That's pretty close to the Windows 1 rendering, yes...
22:11:47 <ehird> http://gallery.techarena.in/data/516/Windows_1_01_Calculator.png
22:11:55 <ehird> Just an aspect ratio vs font issue.
22:12:15 <ehird> Obviously the old Windows 1 font doesn't exist any more, and the old Windows 1 resolutions weren't the same aspect ratio.
22:12:22 <ehird> So it substitutes the font, which overflows.
22:12:30 <ehird> And it looks stretched in the Windows 1 rendering because of the res.
22:15:13 <ehird> pikhq: any luck with those motif adjustments?
22:15:34 <pikhq> ehird: Not been futzing with them.
22:15:51 <pikhq> Too busy trying not to scream at VMware Server.
22:16:28 <AnMaster> ehird_, wait, that looks like non-square pixels somewhere?
22:16:33 <pikhq> It appears to believe that the proper response to asking it to launch a VM is: chown -R root:root virtual_machine;chmod 660 virtual_machine
22:17:22 <AnMaster> ehird_, huh? What soft of monitor was that?
22:17:48 <ehird_> STOP FUCKING CREATING ~/DESKTOP!!!!
22:17:55 <ehird_> I don't HAVE a bloody desktop!
22:19:20 <pikhq> Don't you love how programs make stupid assumptions?
22:19:50 <pikhq> Like "Yes, I would *love* to have a ~/Desktop", or "Yes, I would *love* to have files in ~ be owned by root".
22:20:04 <ehird_> AnMaster: I'm pretty sure ~/Desktop is a default for some magic XDE "DESKTOPLOCATION" variable that I refuse to set.
22:20:17 <ehird_> "It's STANDARD. It's not like ~ is really *yours* or anything."
22:20:21 <ehird_> "Fuck you, user. Fuck you."
22:20:36 <ehird_> AnMaster: because I have no bloody desktop, and I don't bow to the authority of XDG to tell me that I do.
22:21:03 <AnMaster> ehird_, also that is trying to be noob-friendly. computer illiterates expecting files to download to desktop and such
22:21:28 <ehird_> if (File.exists("~/Desktop")) {
22:21:37 <ehird_> defaults.download_location = "~/Desktop";
22:21:45 <ehird_> defaults.download_location = "~/";
22:21:52 <AnMaster> ehird_, ubuntu renames it to Skrivbord on Swedish systems
22:22:02 <ehird_> So? If the variable is set, use it.
22:22:21 <AnMaster> ehird_, your solution fails at i18n just. You need to use said variable in place of "~/Desktop"
22:22:27 <ehird_> But you do _not_ create non-dot directories in ~ that aren't vital to your program's function and desired by the user.
22:22:32 <ehird_> AnMaster: IT'S PSEUDOCODE, FFS!
22:22:38 <ehird_> It doesn't HAVE to work.
22:30:51 -!- ehird_ has quit ("leaving").
22:31:36 -!- ehird_ has joined.
22:31:53 <ehird_> Anyone know how to configure Xpdf's defaults? Doesn't seem to be anything on the website about it. I'll check the man page.
22:34:24 <ehird_> Anyway, the reason my cursor is red is because lwm sets it to be when you're over window decorations or the root window -- presumably so that you know when you're in lwm land and (assuming you're not over a window decoration) can spawn programs with buttons 1 and 2.
22:40:18 -!- Sgeo has joined.
22:44:38 <ehird_> Gregor: You know things about Debian, right?
22:47:11 * Sgeo turned a popular Metaplace world into an Orgy world
22:47:15 -!- Halph has joined.
22:48:01 -!- coppro has quit (Read error: 104 (Connection reset by peer)).
22:48:10 -!- Halph has changed nick to coppro.
22:50:45 <ehird_> coppro: oi, you use debian don't you
22:51:53 <ehird_> "Postoffice accepts (and ignores) many of the same command line options that are passed to sendmail"
22:53:55 <fizzie> That is a popular thing to do; sSMTP accepts and ignores a whole lot of Sendmail options too.
22:54:06 <ehird_> Yes; I just found the wording funny.
22:54:19 -!- MigoMipo has quit ("Quitter!").
22:54:21 <ehird_> Accepting an option sort of tends to imply more than ignoring it to me.
22:54:28 <ehird_> fizzie: Hey, you use Debian! I know this.
22:54:47 <fizzie> It also sounds (to me, anyway) a bit like it implies sendmail ignores those options too.
22:54:55 <ehird_> fizzie: I installed xpdf which brought in some URW font thingies for X11 for its rendering pleasure; but this has caused defoma (you know, the Debian font manager doohickey) to decide that sans and related aliases should point to them.
22:55:00 <ehird_> Thusly my fonts are ugly and I am sad.
22:55:07 <ehird_> Why has it done such a horrible thing to my life?
22:55:28 <fizzie> It is possible that it hates you and hopes you die, but that's only a possibility.
22:55:53 <ehird_> Is there a button and/or buttons I can press that will make it stop hating me hoping I will die?
22:57:03 <fizzie> I don't really know; my approach to font-configuration is something you could classify as "agressively ignorant".
22:57:33 <ehird_> Well, everything *was* just working; I can't fathom why Debian thinks URW ported-to-X11 fonts are a better choice for sans and friends than the DejaVu fonts.
22:57:42 <ehird_> I guess I'll just uninstall xpdf and use some other reader, as a dumb fix.
22:58:26 <pikhq> Why in the world would one want a non-xft font for Sans?
22:58:27 <ehird_> This all-gray-and-black colour scheme I've got going on reminds me of greyscale NeXTStep machines.
22:58:32 <ehird_> pikhq: Oh, it's Xft, I believe.
22:58:35 <ehird_> At least, it antialiases.
22:58:40 <ehird_> See, they're conversions.
22:58:46 <ehird_> So that Xpdf can use fonts that look like the rest of the system.
22:58:49 <pikhq> Okay, so not as awful as is possible.
22:59:00 <ehird_> A noble goal, sure, but as a candidate for default fonthood?
22:59:27 <fizzie> Is it tht gsfonts-x11 thing?
23:00:02 <fizzie> It's a "recommends"-class dependency, so you might be able to fix it by uninstalling that, if you can live without it.
23:00:16 <ehird_> Right, but apparently it improves Xpdf's font display immensely.
23:00:21 <ehird_> So I'll just use another reader. No biggie.
23:01:25 <ehird_> It'd be nice if terminals had a quick-use command "<cmd> prog arg ..." that opened a new terminal running that.
23:01:33 <ehird_> Like tc, for terminal command.
23:01:42 <ehird_> $ xrdb - # oops, I forgot the option ^C
23:02:00 <ehird_> (It'd return immediately after spawning the terminal.)
23:04:16 <ehird_> Hooray; things are good once more. Now I need a pdf reader.
23:05:22 <ehird_> http://pdftohtml.sourceforge.net/ The example output of this thing is impressive; maybe I should use it as a PDF reader.
23:06:06 <lament> it sounds like xkcd so it must be bad
23:06:11 <fizzie> Well, uh... this is all just guesswork, but at least I have a /etc/fonts/conf.d/60-latin.conf (put there by the fontconfig-config package) that specifies a rather random-looking "prefer" list for serif/sans/monospace. Of course it's very much possible something overridizes it somewhere.
23:06:25 <ehird_> lament: Xresources is an okay system. :P
23:06:31 <ehird_> fizzie: That's generated by defoma, I think.
23:06:39 <ehird_> Or, at least, /etc/fonts/conf.d/??-defoma.conf is.
23:06:51 <fizzie> Yes, I do have that autogenerated 30-defoma.conf too.
23:06:59 <ehird_> So I'd *suspect*, though I'm not sure, that it generates the files in that directory from things elsewhere.
23:07:05 <ehird_> Otherwise, that'd just be weird.
23:07:27 <pikhq> ehird_: tc(){nohup urxvt -e "$@" &}
23:07:41 <fizzie> 60-latin.conf is one of the "static" files in the fontconfig-config package, as far as I can figure out.
23:07:45 <ehird_> nohup is crap because it makes nohup.out and stuff.
23:08:00 <ehird_> tc() { urxvt -e "$@" & disown }
23:08:19 <pikhq> Or just >/dev/null. :P
23:08:32 <ehird_> Yes, but disown is *meant* for that.
23:08:54 <ehird_> Anyway, unfortunately tc is forced to be suboptimal: "tc ls" should stay open even after ls returns, but "tc man ls" shouldn't.
23:09:06 <ehird_> Two separate commands would just be unneccessary mental overhead, though.
23:09:12 <fizzie> On the other hand, in 30-defoma.conf I end up with LMSans10-Regular and LMRoman10-Regular fonts (the Latin Modern set, which is a Computer Modern extension) as sans and serif, respectively. I'm not so sure that's very sensible; but on the other hand I don't think my Sans looks like that either.
23:09:29 <pikhq> That's pretty much a bug in Unix semantics.
23:09:42 <pikhq> Probably quite reasonable to do in Plan 9.
23:09:47 <ehird_> pikhq: Correction -- pretty much a bug in ncurses semantics.
23:09:56 <ehird_> In Plan 9, there's nothing like "man".
23:10:04 <ehird_> It'd just stay open after any command.
23:10:19 <ehird_> Well, you could pass it to the pager, but then it'd just stay after you go past the last line.
23:10:28 <ehird_> (The pager doesn't use ncurses-style stuff in Plan 9.)
23:10:31 <pikhq> The Unix semantic in question is "always close after the program exits".
23:10:46 <ehird_> But not closing after it exits would break modern man(1)s.
23:10:50 <ehird_> Because they're crap. :P
23:11:34 <ehird_> It's annoying that even GNU sleep doesn't have a "forever" option.
23:12:27 <pikhq> I'm surprised GNU sleep isn't at least as bloated as GNU hello.
23:12:38 <ehird_> "while true; sleep 1000d; done" does it.
23:13:42 <ehird_> urxvt -e "sh -c '$@; while true; do sleep 1000d; done'" &
23:14:09 <ehird_> pikhq: "Can you tell what it is yet?"
23:15:52 <ehird_> Whatever, I'll just not use this for things like ls.
23:16:15 <ehird_> You can easily run "ls" in your current terminal without disrupting things.
23:16:17 <ehird_> Ooh, if I do "tc irssi" it gets the title irssi.
23:16:28 <ehird_> I think I'll make tc a shell script, not a function.
23:20:52 <ehird_> "You are correct and I apologise. Your last project was actually both commercially viable and original. Unfortunately the part that was commercially viable was not original, and the part that was original was not commercially viable."
23:26:36 <ehird_> I think I need to improve my accuracy with a mouse.
23:27:46 <ehird_> By the way, if any Firefox users want typing a query in the address bar to search Google instead of I'm Feeling Lucky, and thus remove the need for the search box, set keyword.URL to:
23:27:53 <ehird_> http://www.google.com/search?ie=UTF-8&oe=UTF-8&sourceid=navclient&q=
23:34:00 <ehird_> Oh, fuck; and the system gets told about the DPI again and thusly fucks up.
23:37:02 <ehird_> coppro: Die, foul demon of non-rolling release and... and GNOME and... SHUTTLEWORTH
23:37:21 <ehird_> Time to reboot, anyway.
23:37:26 <ehird_> To fix the fucking DPI fucking fuckshit fucking fucker.
23:37:33 -!- ehird_ has quit ("Lost terminal").
23:39:22 <coppro> ehird_: Guess what fun I discovered today
23:39:40 <coppro> CSS Level 3 has a 3D transforms module - joy!
23:41:22 -!- ehird_ has joined.
23:41:31 <ehird_> Why don't terminals execute ~/.profile by default?
23:41:41 <ehird_> I mean, for a desktop machine, that means that ~/.profile is basically "console rc".
23:41:48 <ehird_> Which is dumbfuck retarded.
23:42:04 <ehird_> So what am I meant to do? Put things in bashrc? So only bash reads them?
23:43:35 <coppro> I think you're supposed to put ". .profile" in your .bash_profile
23:44:43 <ehird_> dash is your sh because dash is leaner and stuff and /bin/sh only has to be a POSIX sh, not bash
23:44:55 <ehird_> also, I have no .bash_profile.
23:45:07 <ehird_> And loading .bash_profile but not .profile? /etc/profile too? THE SYSTEM IS FUCKED UP
23:45:14 <ehird_> coppro: Or, I could just put . .profile in my bashrc.
23:45:30 <coppro> ehird_: True, you could! But that would be different behavior!!1!!11
23:45:41 <ehird_> But more pertinently: why don't terminals default to login shells?
23:45:59 <ehird_> In fact, why doesn't some distro completely abolish all the rc madness and just make there be one file?
23:46:07 <pikhq> Because "zomg it wasn't spawned by login(1)".
23:46:12 <ehird_> (You could make all of them be loaded in all cases, so people don't have to know which one to create.)
23:46:19 <coppro> because they won't all parse on every shell
23:46:24 <coppro> and knowing which one to create is easy
23:46:33 <ehird_> coppro: dude, you can test for shell
23:46:39 <pikhq> coppro: profile should parse on all Bourne shells.
23:46:47 <coppro> but what if you use csh?
23:46:57 <ehird_> uhh, csh doesn't load .profile afaik
23:46:58 <pikhq> csh never loads profile.
23:47:10 <coppro> but ehird_ wants one for every shell ever
23:47:11 <pikhq> I think it loads profile.csh
23:48:15 <ehird_> Anyway, it simply makes no sense:
23:48:21 <ehird_> # set PATH so it includes user's private bin if it exists
23:48:21 <ehird_> if [ -d "$HOME/bin" ] ; then
23:48:21 <ehird_> PATH="$HOME/bin:$PATH"
23:48:27 <ehird_> In Debian, this snippet is found in .profile.
23:48:32 <ehird_> Now: Why would you possibly want that?
23:48:38 <ehird_> WHO would expect consoles to be able to access ~/bin stuff but not terminals?
23:48:46 <ehird_> Why would this be desired default behaviour?
23:49:03 <coppro> As I said, finding the correct shell is easy: man $(egrep $(whoami) /etc/passwd | egrep -o "[^/:]+$")
23:49:06 <ehird_> It's idiotic to have such automagic behaviour if it NEVER RUNS for the usage you'd most want it in.
23:52:19 -!- MizardX has quit (Read error: 104 (Connection reset by peer)).
23:52:56 -!- MizardX has joined.