←2010-03-25 2010-03-26 2010-03-27→ ↑2010 ↑all
00:00:09 <fizzie> I'm not qualified to guesstimate the lag, but I don't think arecord's been designed to minimize the time taken before the input sound is output into the "file". There's some --buffer-time option there, not sure how much that is related to writing.
00:00:29 <Gregor> DEAD DOVE
00:00:31 <Gregor> DO NOT EAT
00:00:51 <AnMaster> fizzie, hm
00:01:12 <fizzie> It's a low-effort thing to try out, though, given that it's just two commands basically, and a single mixer change.
00:01:14 <AnMaster> fizzie, would be a PITA to change system a to use jack or system b to use pulseaudio though
00:01:15 <AnMaster> :(
00:01:50 <AnMaster> Deewiant, well what is the bug then?
00:01:57 <AnMaster> Deewiant, I fixed this off by one error here
00:02:00 <fizzie> Well, it could work. And if it doesn't work out of the box, you might be able to use those arecord command line flags to make it better.
00:02:07 <AnMaster> Deewiant, I might be going to sleep soon
00:02:28 <AnMaster> fizzie, waiting for Deewiant to describe the bug first
00:02:47 <Deewiant> AnMaster: While I'm verifying this: are you going to do anything about the negative argument to q with efunge?
00:02:59 <AnMaster> Deewiant, what does efunge do with that?
00:03:02 <AnMaster> I'm interested
00:03:07 <AnMaster> but haven't tested yet
00:03:07 <Deewiant> (no error logger present) error: "Error in process <0.0.0> with exit value: {badarg,[{erlang,halt,[-34]},{init,do_stop,2}]}\n"
00:03:18 <AnMaster> Deewiant, I argue that it is probably undefined
00:03:25 <AnMaster> and that I might or might not do anything about it
00:03:32 <Deewiant> I argue that such behaviour is a pain in the butt ;-P
00:03:47 <AnMaster> Deewiant, I argue that no one sane would use a negative argument to q
00:03:59 <AnMaster> since, well, it is unlikely to be useful
00:04:00 <Deewiant> It can happen by accident
00:04:04 <AnMaster> the OS won't support it
00:04:10 <AnMaster> Deewiant, then the user finds it easily
00:04:11 <Deewiant> And of course it's done on purpose as well
00:04:16 <Deewiant> People do "return -1" and the like
00:04:16 <AnMaster> Deewiant, oh?
00:04:19 <Deewiant> Even in C
00:04:21 <AnMaster> not to an OS
00:04:33 <Deewiant> It's equivalent
00:04:38 <AnMaster> Deewiant, what am I supposed to do with it? take absolute value?
00:04:45 <AnMaster> that is about the only thing I'm prepared to do
00:05:00 <AnMaster> I'm not going to emulated unsigned 8 bit
00:05:01 <Deewiant> Presumably the docs will say what range of values it should take, and most likely the correct solution is mod 256 or something
00:05:13 <AnMaster> Deewiant, mod 256 will yeild -1
00:05:14 <AnMaster> :D
00:05:17 <AnMaster> well it is rem
00:05:44 <AnMaster> stop(Status) -> void()
00:05:44 <AnMaster> Types Status = int()>=0 | string()
00:05:45 <AnMaster> btw
00:05:47 <Ilari> Ah, I figured out why ALSA wasn't working: user 'Ilari' was not member of apporiate group. :-/
00:05:48 <AnMaster> from erlang docs
00:06:16 <AnMaster> Deewiant, what I can do is: leave it as it is, take absolute value, truncate it to 0
00:06:39 <AnMaster> Deewiant, unless you can point me to where in the 98 specs it says I should take modulo and mess around with emulating unsigned 8 bit
00:06:45 <Sgeo> o.O There's only one actual instruction in the Wireworld Computer?
00:07:01 <Deewiant> AnMaster: I'm just saying that getting an error which I can't even ^C out of easily is a pain in the ass
00:07:06 <Deewiant> My recommendation is absolute value
00:07:10 <AnMaster> Deewiant, doube ^C works
00:07:13 <AnMaster> most likely
00:07:34 <AnMaster> Deewiant, ^C in general doesn't work the way you prefer in erlang I suspect
00:10:15 <Deewiant> Verified that other issue: Mycology bug, not interpreter.
00:10:38 <AnMaster> Deewiant, arguably the negative argument to q *is* undefined though. And silently ignoring that is doing more bad than good
00:10:55 <AnMaster> Deewiant, causing possibly hard to debug issues
00:11:14 <Deewiant> I think that when you're doing q you're at the point where you just want to quit and don't care about errors :-P
00:11:53 <Deewiant> Even Java's System.exit doesn't throw InvalidParameterException ;-P
00:12:04 <AnMaster> Deewiant, well what if someone implements /usr/bin/test in befunge?
00:12:11 <AnMaster> then you *care* about exit code
00:12:29 <Deewiant> Yes, and you should be careful to get it right
00:12:37 <AnMaster> Deewiant, indeed
00:12:43 <Deewiant> But that's a bad example since it's really trivial: 0 or 1
00:12:49 <AnMaster> then you don't want your system to silently swallow the error code
00:12:54 <AnMaster> Deewiant, okay what about fsck then
00:13:00 <AnMaster> that has highly complicated error codes iirc
00:13:01 <Deewiant> I don't know ~anything about fsck.
00:13:15 <Deewiant> If you're writing fsck in Befunge you have other problems ;-P
00:13:30 <AnMaster> Deewiant, "The exit code returned by fsck is the sum of the following conditions: <list truncated from irc>"
00:13:31 <AnMaster> but
00:13:40 <AnMaster> basically a bitmask
00:13:55 <AnMaster> considering the values are 1 2 4 8 and so on
00:16:10 <AnMaster> Deewiant, anyway I think truncating to 0 is more DS9K and thus should be preferred
00:16:41 <AnMaster> (arguably efunge does have some DS9K traits, such as bignum in a language where that is not common)
00:16:48 <AnMaster> (it used to have i without o)
00:17:23 -!- FireFly has quit (Quit: Leaving).
00:18:37 <AnMaster> Deewiant, btw does ccbi run on vxworks?
00:18:42 <Deewiant> vxworks?
00:18:45 <AnMaster> yes
00:18:47 <Deewiant> What's that
00:19:00 <AnMaster> http://en.wikipedia.org/wiki/VxWorks
00:20:05 <Deewiant> The libraries it depends on don't
00:20:27 <AnMaster> Deewiant, iirc there is a (commerical only(?)) port of erlang to it
00:20:41 <AnMaster> I don't know if efunge would run of it though
00:23:09 <AnMaster> Deewiant, anyway it is likely some embedded controllers around you run vxworks
00:26:17 <AnMaster> Deewiant, I *might* make efunge print a warning for negative parameters
00:26:18 <AnMaster> or such
00:26:51 <AnMaster> Deewiant, perhaps clear screen, print drawing, enter line drawing mode, exit(0)?
00:26:52 <AnMaster> ;P
00:27:10 <Deewiant> >_<
00:28:15 -!- kar8nga has quit (Remote host closed the connection).
00:28:26 <AnMaster> Deewiant, or, maybe we should block everything but SIGKILL? and if root mount an nfs mount, launch a DDoS on the server, and then try to read something from it
00:28:51 <AnMaster> (that will put it in uninterpretable sleep)
00:28:55 <Deewiant> If efunge does that I'll disqualify it from Mycology results ;-P
00:29:16 <AnMaster> Deewiant, haha
00:30:54 <AnMaster> Deewiant, pushed behaviour change for efunge
00:31:19 <AnMaster> cache effects might apply
00:32:34 <AnMaster> fizzie, I can't do the http://alsa.opensrc.org/index.php/Network stuff
00:32:40 <AnMaster> there is no capture source control
00:32:44 <AnMaster> on the system I want to forward from
00:32:57 <AnMaster> there is only on the system I want to forward to
00:34:38 <fizzie> That's curious. Though the mixer controls are horribly divergent across systems.
00:37:22 <fizzie> You should be able to do some .asoundrc trickery, though.
00:38:24 <fizzie> There's this "loopback soundcard" alsa module, http://www.alsa-project.org/main/index.php/Matrix:Module-aloop
00:38:59 <fizzie> You should be able to use that, and have your game use one end of it for output, and arecord on the other end for feeding netcat.
00:39:25 <fizzie> http://dev.inzenet.org/~panard/hack_record_hda_sound also has some example bits, though it looks awfully complicated and uncommented.
00:41:22 <fizzie> It's also borderline possible that the very simple "dummy" soundcard actually would support the capture-mixer thing.
00:41:38 -!- augur has quit (Ping timeout: 246 seconds).
00:43:01 <fizzie> Yes, it might just work.
00:43:27 <fizzie> At least it's got the exact same capture and playback controls.
00:43:46 <fizzie> I could try playing something to the dummy soundcard and seeing if I can record it.
00:47:29 <fizzie> Strange, it records noisy bursts of sound every 0.1 seconds, and silence for the most part. Heh, I wonder what's up with that.
00:50:02 <fizzie> Apparently it just drops what's written and generates (mostly) zeroes out.
00:52:38 <AnMaster> hm
00:52:52 <AnMaster> <fizzie> http://dev.inzenet.org/~panard/hack_record_hda_sound also has some example bits, though it looks awfully complicated and uncommented.
00:52:59 <AnMaster> I think it is intel hda
00:53:27 <fizzie> Most motherboard-integrated things are, though even the actual HDA chipsets differ.
00:53:45 <AnMaster> yes indeed they do
00:54:04 <fizzie> I don't seem to have the mixer-recording thing on this box either, though I seem to recall there was.
00:54:37 <fizzie> Anyway, the "aloop" "soundcard" should work (not enabled in my kernel config, though), and alsa's "file" plugin might work too.
00:55:19 <fizzie> And also (assuming the card does full-duplex right) there's the ridiculously low-tech way of looping back the line out to line in. :p
00:56:04 <fizzie> You could check out the file plugin reference at http://alsa-project.org/alsa-doc/alsa-lib/pcm_plugins.html (pretty far down on the list).
00:56:32 <fizzie> It can even pipe directly into a shell command.
00:56:44 <fizzie> That should have less latency than trying to involve recording with arecord.
01:00:56 <fizzie> See for example Audacity's docs about it, http://wiki.audacityteam.org/wiki/Recording_audio_playing_on_the_computer#Using_the_ALSA_PCM_file
01:01:24 <AnMaster> mhm
01:02:19 <fizzie> Actually there even seems to be a "file" device.
01:02:26 <fizzie> "File device
01:02:27 <fizzie> The file device is file plugin with null plugin as slave. The arguments (in order: FILE,FORMAT) specify filename and file format.
01:02:27 <fizzie> Example:
01:02:27 <fizzie> file:'/tmp/out.raw',raw
01:02:27 <fizzie> "
01:04:05 <fizzie> mplayer -ao alsa:device="file='testfile.raw',raw" testfile.wav seems to do a sensible thing, at least based on file sizes.
01:04:32 <fizzie> (That = there after "file" is just mplayer's stupidity, you have to write the : as a = so that it won't interpret it as a parameter change.)
01:04:35 -!- augur has joined.
01:05:30 <fizzie> So in theory you should be able to just use file:'|/bin/nc -u server port',raw as the ALSA "device" specifier string for your game, without fiddling with any config files or anything.
01:05:59 <fizzie> Of course then you'll need to specify sample rates and types and so on on the listening side.
01:06:02 <AnMaster> fizzie, well I wanted the whole system sound
01:06:08 <AnMaster> not just a single game
01:06:20 <AnMaster> game was the primary example of programs in question
01:06:55 <AnMaster> if the built in sound wasn't so bad, and if I had a suitable cable, I could just feed one computer's line out into the line-in on the next
01:06:59 <fizzie> Well, then you can use the example given in http://wiki.audacityteam.org/wiki/Recording_audio_playing_on_the_computer#Using_the_ALSA_PCM_file as "set the teeraw device as your default device with something like".
01:07:16 <AnMaster> hah
01:07:32 <fizzie> Or maybe the final example there as a full setup.
01:07:41 <fizzie> It even has a command-piping thing.
01:08:08 <AnMaster> mhm
01:08:34 <AnMaster> actually setting up jack on that ubuntu system might be *less* of a PITA XD
01:09:58 <AnMaster> fizzie, what would I use on the listening side then? same as the original?
01:10:23 -!- augur has quit (Ping timeout: 265 seconds).
01:10:34 <fizzie> Well... I'm not really an ALSA expert, but if you use an .asoundrc as something like http://pastebin.com/NfALSAK8 -- with the format "wav" thing -- you should be able to use the original nc | aplay thing.
01:11:20 <fizzie> That should catch anything that's written using ALSA to the "default" PCM device. It might not catch programs that use the OSS-dsp emulation thing. And it might not work at all.
01:11:33 <AnMaster> fizzie, how does that interact with that pulseaudio is in use on that computer?
01:12:14 <fizzie> I know even less about pulse, but I guess it uses alsa libs to interact with the underlying hardware, and therefore might work.
01:12:29 <AnMaster> fizzie, it uses alsa to capture things then send it back to alsa
01:12:36 <AnMaster> basically ALSA-directly-apps
01:12:37 <AnMaster> end up
01:12:38 <AnMaster> like
01:12:49 <AnMaster> ->ALSA->pulse->ALSA
01:13:20 <fizzie> It could get confused, I don't know. Pulse has that own networking thing too, but then you'd need pulseaudio also on the receiving side.
01:13:46 <AnMaster> yes
01:13:53 <fizzie> Possibly it might be better to not set that "pipe-to-network" device as the default pcm one, but instead configure pulse to feed things into it.
01:13:58 <AnMaster> fizzie, and the receiving side uses jack/pure-ALSA
01:14:41 <fizzie> I'm also a bit concerned about what the "netcat | aplay" will do on the receiving side if the sample rates on the sending side change.
01:14:51 <AnMaster> fizzie, ouch indeed
01:15:21 <fizzie> Doing it fiddling like this is a bit awkwardly manual. If you seriously want all the sound, and keep it set up semi-permanently, I might go with the "just set up jack on Ubuntu too and use netjack" solution.
01:15:48 <AnMaster> fizzie, jack isn't really suitable for my laptop in *normal* use
01:16:21 <AnMaster> fizzie, anyway I could use jack alongside pure alsa without problems on my desktop, it has hardware mixer
01:16:25 <AnMaster> since it has an sb live
01:17:11 <AnMaster> fizzie, and yes this is semi-permanently as in, it will be used a lot but not all the time
01:17:39 <fizzie> If you configure jack to only have the into-network thing, no "actual" direct-sound use at all... well, I'm not sure how to make the rest of the system use that, then.
01:17:49 <fizzie> I think I'll need some sleep now.
01:18:43 <AnMaster> mhm
01:18:48 <AnMaster> thanks for your help
01:22:07 <fizzie> I wanted to have a setup like this for my laptop too, but (since it runs OS X most of the time) I finally gave up and just bought that USB digital-output sound-stick I mentioned, and connect it to the front panel inputs of the amplifier when necessary. (The alternative would've been to feed the digital audio over LAN to one of these Real Computers that are permanently hooked to the amplifier.)
01:22:22 <fizzie> Networked audio seems to be a bit nontrivial.
01:22:27 <fizzie> (Away now for real, night.)
01:23:54 <oklopol> night
01:56:47 -!- Tritonio_GR has quit (Read error: Connection reset by peer).
02:28:22 -!- coppro has quit (Ping timeout: 264 seconds).
02:45:48 -!- coppro has joined.
02:47:47 -!- augur has joined.
02:52:55 -!- augur has quit (Ping timeout: 246 seconds).
02:57:34 -!- augur has joined.
03:01:58 -!- coppro has quit (Quit: I am leaving. You are about to explode.).
03:02:28 -!- coppro has joined.
03:11:35 -!- Gracenotes has joined.
03:37:34 -!- Asztal has quit (Quit: .).
04:15:24 * coppro tries to think of a full version of his language that isn't overly influenced by Haskell
04:15:35 <pikhq> Antiskell.
04:15:47 <pikhq> Every decision made by Haskell, it chooses the opposite.
04:16:09 <pikhq> Unicode support? Nah, we're going for binary coded decimal EBDIC
04:16:38 <pikhq> (yes, the EBDIC is encoded into decimal, which is then binary coded.)
04:16:58 <pikhq> Strict typing? Nah. We're going for /dev/urandom typing.
04:17:22 <pikhq> And every statement is implicitly wrapped in unsafePerformIO.
04:17:52 <coppro> lol
04:18:00 <coppro> did you catch my idea earlier?
04:18:10 <pikhq> Except for the ones that are implicitly wrapped in unsafePerformIO . unsafeIntersperseIO . return
04:18:18 <pikhq> Nah.
04:18:50 <coppro> basically, rather than pure types, there's "datapacks"
04:19:29 * Sgeo wants Antiskell to become real
04:19:46 <Sgeo> Although obviously, it can't be a perfect opposite
04:20:16 <Sgeo> Otherwise, it would not be TC, and would not be able to do any IO, contrary to pikhq's earlier statements
04:20:22 <coppro> upi
04:20:25 <Sgeo> Hm, I think "opposite" is ill-defined
04:20:27 <Sgeo> upi?
04:20:33 <coppro> you'd need an anticomputer for that.
04:21:11 <pikhq> Sgeo: There is no requirement for unsafePerformIO to do much.
04:22:15 <pikhq> In fact, I suggest that all statements except for undefined be equivalent to bottom.
04:22:23 <coppro> sorry
04:22:25 <pikhq> Except in confusing circumstances.
04:22:39 <coppro> my father needed to ask me a question
04:23:08 <Sgeo> Can Antihaskell solve the halting problem?
04:23:12 <coppro> anyway, instead of types, there's "datapacks". All UDTs are actually datapacks.
04:24:07 <pikhq> Sgeo: GHC can, so Antiskell cannot.
04:24:09 <coppro> a given value can possess any number of datapacks; the union of the packs it has could be said to represent its actual type. Some datapacks could require other datapacks to exist on the type (*coughinheritancecough*)
04:24:09 <pikhq> :P
04:24:18 <Sgeo> pikhq, wait what?
04:24:43 <pikhq> Sgeo: Just a joke. Compile "main = main". ;)
04:25:13 <Sgeo> Too lazy to bother, I take it the compiler sees that somehow?
04:25:25 * coppro feels ignored
04:25:52 <pikhq> It results in an error. "infinite loop" or some such.
04:26:51 <coppro> ping
04:27:29 <pikhq> GNIP
04:27:53 <coppro> okay, so that's working
04:27:57 <Sgeo> Dangit
04:28:02 * coppro tries to investigate other causes of not being listened to
04:28:30 <Sgeo> The best thing to do when someone pings is to ignore that >.> [Ok, so I was being mean]
04:28:58 <Sgeo> So basically values have multiple types?
04:29:02 <coppro> yeah
04:43:25 <coppro> ugh, sorry, I got distracted again
04:44:17 <coppro> There would also be "type patterns" which could be used to match against types; e.g. {Int~Nil} means "Int
04:44:26 <coppro> *means "Int or Nil, but not both"
04:44:34 <coppro> and would be the implementation of a nullable value
04:45:06 <Sgeo> Hm
04:45:10 <Sgeo> Interesting
04:46:35 <coppro> {Int|Nil} or {Int&Nil} could probably also exist (I may choose to add a way to make datapacks mutually exclusive), but, by convention, would be pretty useless
04:46:42 <coppro> Ideally type inference would drive a lot of the system
04:49:51 <coppro> The real power, though, is in expressing reality better than most languages
04:50:36 * Sgeo wonders how you manipulate an Int+String
04:50:50 <coppro> Sgeo: it would have both an int value and a string value independently
04:51:09 <Sgeo> Ah
04:51:27 <coppro> geek example: A Magic card has a type; each type has a list of allowable subtypes for that type
04:51:59 <coppro> a Magic card may have multiple types
04:52:38 * Sgeo misunderstood "Magic card" for a second
04:53:02 * Sgeo points out "for a second" before coppro or anyone else elaborates about Magic
04:53:12 -!- oerjan has joined.
04:53:13 <coppro> huh?
04:53:17 <coppro> I assume you understand now
04:53:37 <pikhq> 私 見今 Oerjan !
04:53:47 <coppro> so you'd express this by creating a datapack for each list of subtypes, and then add these to the card value. Since they aren't mutually exclusive, you can have as many as you like
04:53:52 <pikhq> 正 今, too!
04:54:26 <oerjan> i am not a friend of people who greet me with unicode
04:54:34 <pikhq> Hah.
04:54:42 <pikhq> 今日, Oerjan. How are 前?
04:54:57 <oerjan> it looks like question marks.
04:56:27 <pikhq> Oh 本当?
04:57:09 <coppro> any thoughts about my idea?
04:57:22 <oerjan> pikhq, almost _no one_ here understands japanese. showing it off all the time is just annoying.
04:58:05 <oerjan> (even if i could see it properly.)
04:58:28 <pikhq> Hrm. When was the last time I used code-points out of ASCII in this room, anyways?
04:58:40 <coppro> 2 minutes 2 seconds ago
04:58:51 <pikhq> coppro: Before oerjan just entered.
04:58:59 <coppro> dunno
04:59:07 <coppro> last week I think?
04:59:35 <coppro> please ping me with any comments you may have; it's webcomic time
05:00:48 <pikhq> oerjan: Also. Sorry. I've been cramming kanji for the past couple months. And for some reason, I felt like using kanji in English. I'm going to blame my lack of sleep for that...
05:01:16 <oerjan> i'll blame my current grumpiness, then :)
05:01:27 <oerjan> *y?
05:03:00 <pikhq> Now, for random characters: 亜sdtf日小田鵜fhgpqeかsfqwさkじょ9tqぁsづてょdヴい
05:09:03 <oerjan> <coppro> in other words, a value can possess multiple types at once <-- they're called intersection types iirc. they were important for fundamental lambda calculus in a book i looked in (barendregt?)
05:10:03 * Sgeo wonders if he can use a cryonics article for his psychology assignmnet
05:10:22 <oerjan> if you have intersection types + a special "catch-all" type omega, then you have a simple type system in which all normalizable lambda expressions can be typed, iirc
05:10:52 <oerjan> without omega, you need strongly normalizable, iirc
05:11:54 <oerjan> oh and omega can occur only on the left side of -> or something, so things don't get to use just it for trivial typing
05:13:56 <coppro> oerjan: thanks for the name
05:14:35 <oerjan> hm i think the types may even have been preserved under reduction, both ways
05:14:54 <coppro> I don't see much that looks like a useful language with them though
05:15:39 <oerjan> coppro: higher order functions, if you use a polymorphic function on two different types
05:16:38 <coppro> oerjan: that's a special case
05:16:46 <coppro> as in, doesn't apply to all types
05:17:15 <oerjan> but they're probably not practical, type inference is probably impossible and forall types with type variables like in haskell/ML are more practical
05:18:17 <coppro> most of what I see intersection types used for is type inference
05:18:44 <coppro> and what I'd be doing is more like type unionization
05:20:44 -!- augur has quit (Ping timeout: 265 seconds).
05:20:52 <oerjan> oh well
05:33:26 <coppro> oerjan: ?
05:35:51 <oerjan> !
05:36:00 <coppro> [22:20:28]<oerjan>oh well
05:36:42 <oerjan> coppro: that was intended as a "conversation over" mark :P
05:36:49 <coppro> ah
05:37:26 <coppro> ugh, it's that time of night again
05:37:39 <oerjan> oh no, not the vampires
05:38:10 <coppro> where I really should go to bed, but I don't
05:39:02 <coppro> worse, I have an excuse
05:39:30 <oerjan> the bed is haunted!
05:39:53 <coppro> Today's Hansard isn't out yet
05:40:55 <oerjan> a principled follower of the british parliament, i take
05:41:32 <oerjan> or would that be canada
05:42:53 <oerjan> "In one instance, during a Liberal filibuster in the Canadian Senate, Senator Philippe Gigantès was accused of reading one of his books only so that he could get the translation for free through the Hansard." :D
05:44:16 <coppro> Canadian in this case
05:44:41 <pikhq> That's... Beautiful.
05:48:22 -!- Oranjer has left (?).
06:06:27 <Gregor> http://xkcd.com/719/ I've never had any of these dreams :(
06:11:26 <pikhq> I've had few dreams that I can actually recall.
06:11:31 <oerjan> i don't have car dreams either. might have _something_ to do with the fact i don't drive :D
06:12:00 <oerjan> ok maybe an occasional dream as passenger
06:13:08 <oerjan> wait i _have_ dreamed about my teeth falling out D:
06:26:08 -!- pikhq has quit (Read error: Connection reset by peer).
06:38:14 -!- oerjan has quit (Quit: leaving).
06:47:47 -!- jcp has quit (Quit: I will do anything (almost) for a new router.).
07:14:15 -!- FireFly has joined.
07:34:22 <oklopol> "oerjan: pikhq, almost _no one_ here understands japanese. showing it off all the time is just annoying." <<< no it's not!
07:36:15 -!- coppro has quit (Remote host closed the connection).
07:36:50 -!- coppro has joined.
07:39:39 <oklopol> nowadays it's less common for me to roll all the time, what's more common is i jump or something and realize i'm not falling down, but just rising and rising hundreds of meters first
07:39:55 <oklopol> in dreams
07:40:18 <coppro> I generally don't remember my dreams
07:40:22 <oklopol> also i become lucid quite often, but i still haven't mastered not waking up almost instantly after
07:40:44 <coppro> all I can really remember is that they tend to be things I don't want to remember for one reason or another
07:40:54 <oklopol> i usually manage to have sex with someone for a few seconds, but then it fades away
07:41:08 <oklopol> why wouldn't you want to remember something
07:41:35 <coppro> I think it's that I'm really good at getting into really awkward situations in dreams
07:41:39 <coppro> can't quite remember
07:41:52 <oklopol> oh lol right
07:42:40 <coppro> Pretty sure it's not scariness though
07:43:44 <coppro> the few good dreams I ever remember are only even vaguely remembered because they're the ones I get woken up in the middle of
07:46:44 <coppro> ah, finally, Hansard's out
07:46:50 <coppro> then I can sleep
07:47:45 <oklopol> yay hansard
07:48:10 -!- myndzi\ has joined.
07:51:18 -!- myndzi has quit (Ping timeout: 260 seconds).
07:52:37 <coppro> ooh, an MP quoted MacBeth
07:54:15 -!- augur has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:02:20 <coppro> man, this trade agreement with Colombia is causing quite a stir... 30 hours of debate already
08:02:37 <coppro> AFAICT it boils down to we should/shouldn't consort with the evils
08:02:47 -!- tombom has joined.
08:04:41 <oklopol> evil things are wrong
08:07:08 <coppro> the government says it hopes to encourage the Colombian government to clean up
08:07:13 <coppro> the whole debate is tl;dr
08:09:56 <coppro> ah, I love this country
08:09:59 <coppro> http://www.cbc.ca/canada/story/2010/03/25/prisoners-pensions025.html
08:10:21 <coppro> 25 life sentences and we still give him 13k a year
08:32:04 -!- BeholdMyGlory has joined.
08:51:49 -!- tombom has quit (Quit: Leaving).
09:00:18 -!- BeholdMyGlory has quit (Remote host closed the connection).
09:24:26 -!- FireFly has quit (Quit: Leaving).
09:40:12 -!- MigoMipo has joined.
09:53:02 -!- ais523 has joined.
09:53:11 -!- augur has quit (Read error: Connection reset by peer).
09:53:27 -!- augur has joined.
09:56:24 -!- MigoMipo has quit (Quit: Page closed).
10:47:16 -!- lereah_ has joined.
11:06:03 -!- augur has quit (Ping timeout: 240 seconds).
11:07:05 -!- augur has joined.
11:15:23 -!- Asztal has joined.
11:20:28 -!- augur has quit (Ping timeout: 268 seconds).
11:20:51 -!- augur has joined.
12:52:03 -!- Tritonio_GR has joined.
12:57:31 -!- Tritonio_GR has quit (Quit: Leaving.).
13:13:37 -!- augur has quit (Ping timeout: 264 seconds).
13:14:16 -!- augur has joined.
13:22:42 -!- FireFly has joined.
13:47:04 -!- Gracenotes has quit (Ping timeout: 246 seconds).
14:23:27 -!- fax has joined.
15:02:37 -!- coppro has quit (Quit: I am leaving. You are about to explode.).
15:48:06 -!- MigoMipo has joined.
15:53:38 -!- oerjan has joined.
16:06:02 -!- oklopol has quit (Ping timeout: 248 seconds).
16:09:18 -!- BeholdMyGlory has joined.
16:09:21 -!- BeholdMyGlory has quit (Changing host).
16:09:21 -!- BeholdMyGlory has joined.
16:35:49 -!- augur has quit (Ping timeout: 264 seconds).
16:46:03 -!- Oranjer has joined.
16:50:29 -!- lereah_ has quit (Quit: Leaving).
17:02:10 -!- MigoMipo has quit (Remote host closed the connection).
17:04:33 -!- MigoMipo has joined.
17:10:31 -!- kar8nga has joined.
17:24:10 -!- pikhq has joined.
17:25:10 <pikhq> Kikcking router in shins = WORKING!
17:25:19 <ais523> routers have shins?
17:25:29 <pikhq> Do now.
17:45:37 <ais523> `define mephitic
17:46:31 <HackEgo> * miasmic: of noxious stench from atmospheric pollution \ [15]wordnetweb.princeton.edu/perl/webwn \ * Avernus was an ancient name for a crater near Cumae (Cuma), Italy in the Region of Campania west of Naples. It is approximately 2 miles in circumference. Within the crater is Lake Avernus (Lago d'Averno).
18:00:52 -!- tombom has joined.
18:04:23 <AnMaster> <pikhq> Kikcking router in shins = WORKING! <-- wow
18:04:36 <AnMaster> pikhq, would that be in like the ethernet sockets?
18:05:16 <pikhq> Nah.
18:05:22 <pikhq> You let it grow legs.
18:05:29 <AnMaster> okay
18:05:32 <AnMaster> then ?
18:06:24 <AnMaster> pikhq, just kicking it?
18:06:30 <AnMaster> no more steps in between?!
18:06:47 <pikhq> Then you kick it.
18:06:49 <AnMaster> huh, I newer knew it was _that_ simple
18:11:04 -!- oerjan has quit (Quit: Reboot).
18:13:54 -!- oerjan has joined.
18:31:03 -!- ais523 has quit (Remote host closed the connection).
18:36:46 -!- Slereah has quit (Ping timeout: 240 seconds).
18:43:29 -!- Slereah has joined.
18:56:30 -!- Phantom_Hoover has joined.
18:58:22 <Phantom_Hoover> CURSE YOU, TIME ZONE DIFFERENCES!
18:59:32 <oerjan> oh no, you cursed the time zone differences!
18:59:47 <oerjan> i predict they will take a horrible revenge!
19:00:12 <oerjan> before this weekend is over, even!
19:00:15 -!- Sgeo_ has joined.
19:01:58 -!- Sgeo has quit (Ping timeout: 264 seconds).
19:06:42 -!- Gracenotes has joined.
19:08:41 * Sgeo_ loves when he randomly disconnects!
19:08:45 <Sgeo_> Love it love it love it!
19:09:06 -!- Sgeo_ has changed nick to Sgeo.
19:14:58 -!- sebbu2 has joined.
19:16:04 -!- sebbu has quit (Ping timeout: 246 seconds).
19:16:04 -!- sebbu2 has changed nick to sebbu.
19:29:06 -!- dixon has joined.
19:30:18 <Quadrescence> Hey Sgeo
19:30:23 <Sgeo> Hi
19:30:31 <Quadrescence> What's up?
19:30:54 <Sgeo> Not much
19:30:59 <Sgeo> Thinking about cryonics a bit
19:32:11 <Phantom_Hoover> What about it?
19:32:20 <dixon> That's... cool.
19:32:24 <Phantom_Hoover> It strikes me as far too faith-based.
19:33:09 <Quadrescence> The notion of freezing is faith-based?
19:33:47 <Phantom_Hoover> No, the notion that medical technology will be able to revive you.
19:34:16 <dixon> Hey, that's cold.
19:34:21 <dixon> Doctors are smart.
19:34:33 <pikhq> Yes, well. Faith IN TECHNOLOGY!
19:34:35 <pikhq> :P
19:34:43 <Phantom_Hoover> But dammit, they're doctors, not miracle workers!
19:35:01 <pikhq> Dammit Jim!
19:35:51 <oerjan> anyway, don't do it this way: http://www.darwinawards.com/personal/personal2000-25.html
19:36:26 <pikhq> Delicious, delicious cryonic nitrogen.
19:36:30 <Phantom_Hoover> Ah, the Darwin awards.
19:36:36 <Phantom_Hoover> A tale for every situation.
19:37:01 <Phantom_Hoover> Although it seems the character encoding doesn't work properly...
19:37:07 <oerjan> this one was on today's reddit frontpage
19:37:10 -!- wareya has quit (Ping timeout: 258 seconds).
19:52:10 * Sgeo wishes he never had to use Firefox
19:52:46 <AnMaster> <oerjan> before this weekend is over, even! <-- oh no
19:52:50 <AnMaster> the end of the world is near
19:52:58 * AnMaster stockpiles stuff
19:53:09 * Sgeo stockpiles AnMasters
19:53:14 <AnMaster> how?
19:53:18 <AnMaster> I mean, there is just one of me
19:53:33 <AnMaster> you can't stockpile anything that there is just one of
19:57:49 <oerjan> well, he could try folding you
20:02:34 -!- sebbu has quit (Ping timeout: 260 seconds).
20:03:06 -!- sebbu has joined.
20:04:51 <AnMaster> oerjan, eh?
20:05:05 <AnMaster> oerjan, how would that result in more than one of me?
20:07:21 <oerjan> always doubting are you
20:07:33 <fax> im just here to buy some soy sauce
20:08:15 * Sgeo hopes he can program despite using a trackpad
20:08:21 <Sgeo> Too lazy to sit up at the table next to my bed
20:09:15 <AnMaster> oerjan, yes!
20:09:46 <AnMaster> Sgeo, why do you need anything but a printer (alternatively a monitor) and a keyboard?
20:10:01 <oerjan> fax: old meme which i've never heard before is old
20:10:28 <Sgeo> Easy to move cursor for cut/paste
20:11:16 <oerjan> AnMaster: wait, you use a printer rather than a monitor?
20:11:43 <AnMaster> oerjan, no
20:11:48 <AnMaster> oerjan, did I say I did?
20:12:09 <oerjan> AnMaster: it was your first alternative, above
20:12:13 <AnMaster> Sgeo, mhm, I never found programming without a mouse to be a problem
20:12:36 -!- jcp has joined.
20:12:40 <AnMaster> oerjan, well yes. I wish I had one of those oldstyle hard copy terminals
20:12:42 * oerjan doesn't use a mouse. admittedly he doesn't program much either.
20:12:59 <AnMaster> oerjan, I use a mouse when image editing for example
20:13:10 <AnMaster> a trackpoint really isn't very good for that
20:13:16 <AnMaster> a touchpad even less so
20:13:29 <oerjan> mhm
20:13:41 <AnMaster> well, in general touchpad are useless I find
20:13:48 <AnMaster> trackpoint is fairly usable but a mouse is better
20:13:57 <AnMaster> but for programming I find I only use the keyboard
20:14:22 * oerjan recalls nearly 30 years ago, when my dad borrowed home such a terminal from the job
20:14:43 <AnMaster> wow, I would love to see one in "reality"
20:15:18 <oerjan> we connected to the mainframe via the phone, and used BASIC on it
20:15:27 <AnMaster> heh
20:15:44 <AnMaster> oerjan, I was rather thinking about connecting to something at MIT that ran lisp
20:16:42 <oerjan> i'm not sure if the internet had reached scandinavia yet...
20:17:29 <AnMaster> true
20:18:09 <AnMaster> oerjan, actually Norway was the first place in Europe (and iirc outside US too, but not completely sure about that) to get ARPANET
20:18:23 <oerjan> that does ring a bell
20:21:21 * pikhq rings a bell for Oerjan's skae
20:21:23 <pikhq> Sake, even.
20:22:07 * AnMaster rings a bell for pikhq's skate
20:22:53 * oerjan rings a bell for AnMaster's steak
20:23:23 * AnMaster stakes oerjan with a very pointed bell
20:23:31 <oerjan> ouch
20:23:58 <AnMaster> made of silver.
20:24:05 <oerjan> fancy
20:24:37 <AnMaster> oerjan, well, it offers both werewolf and vampire protection that way.
20:24:48 <oerjan> brains..
20:25:04 <AnMaster> hm what is effective against zombies?
20:25:24 * oerjan lurks toward AnMaster
20:25:28 * AnMaster gives oerjan some plush brains
20:25:38 <AnMaster> (says made in china on them)
20:26:10 * AnMaster beams away to a badly rendered space ship (TOS)
20:26:21 * oerjan thinks they taste like soy
20:26:51 <AnMaster> strange
20:26:56 * oerjan reverses the polarity
20:27:10 <AnMaster> ooh that made it go faster
20:27:11 <AnMaster> thanks
20:27:21 <AnMaster> which is trange
20:27:23 <AnMaster> strange*
20:27:27 <AnMaster> since it is alternating current
20:28:25 <oerjan> it's just a phase(r)
20:28:38 <AnMaster> what is?
20:28:44 <AnMaster> the warp drive?
20:28:51 <AnMaster> that was what you reversed polarity of
20:28:56 <oerjan> the current reversal
20:28:57 <AnMaster> the mains to it
20:29:50 <oerjan> reversing the polarity should be a pi radians face shift, no?
20:29:53 <oerjan> *phase
20:30:46 <AnMaster> oerjan, that only makes sense assuming sinusoidal current
20:31:07 * oerjan thought alternating current was sinusoidal
20:31:12 <AnMaster> it often is
20:31:15 <AnMaster> but it may not be
20:31:36 <AnMaster> what matters is that it is alternating.
20:32:04 <oerjan> oh well
20:32:36 <AnMaster> oerjan, but sinusoidal is the far most common one I think
20:34:26 <AnMaster> oerjan, I'm surprised you didn't ask me what I meant with alternating
20:34:31 <AnMaster> but I shall tell you anyway
20:34:50 <oerjan> nooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
20:34:51 <AnMaster> it means it oscillates across zero.
20:35:09 <AnMaster> you can have square wave shape for example
20:35:15 <AnMaster> or triangular
20:35:23 -!- kar8nga has quit (Remote host closed the connection).
20:35:43 <Phantom_Hoover> Or pentagonal.
20:35:49 <Phantom_Hoover> Maybe.
20:35:55 <AnMaster> Phantom_Hoover, how would that look like
20:36:01 <AnMaster> I have a hard time imagining it
20:36:06 <Phantom_Hoover> Yeah, hexagonal makes more sense.
20:36:13 <AnMaster> and with triangular I mean \/\/\/ style
20:36:19 <Phantom_Hoover> Yes.
20:36:19 <AnMaster> not /_\
20:36:29 <AnMaster> (because that wouldn't make any sense)
20:36:41 <Phantom_Hoover> Perhaps pentagonal would be like triangular, but with less sloping.
20:36:57 <Phantom_Hoover> Or perhaps the wave would go back in time at some points.
20:37:48 <AnMaster> oerjan, btw in electrical engineering it seems common to mix radians and degrees. I have seen stuff like: u(t) = 4sin(50*2*pi*t + 45°) stuff, where (if your unicode is still broken) the thing after 45 is a degree sign
20:38:21 <AnMaster> Phantom_Hoover, hm
20:38:37 <Phantom_Hoover> ^
20:38:39 <Phantom_Hoover> / \
20:38:40 <Phantom_Hoover> \ /
20:38:42 <Phantom_Hoover> /
20:38:43 <Phantom_Hoover> \
20:38:45 <Phantom_Hoover> Or somesuch.
20:38:46 <AnMaster> argh
20:38:57 <Phantom_Hoover> Too much space? Sorry.
20:39:02 <AnMaster> Phantom_Hoover, that is not a function of either x or y
20:39:03 <AnMaster> :/
20:39:20 <Phantom_Hoover> The drawing is awful...
20:39:23 <AnMaster> true
20:39:23 <oerjan> degree signs are fine
20:39:55 <AnMaster> oerjan, but mixing them in one expression? Since the "50*2*pi" stuff is in radians
20:40:12 <fizzie> There was a signal generator at our electronics lab that reset the wave shape (from triangle to sine or the other way around) when you changed some other settings. Ruined our otherwise accurate magnetic field -related measurements, though we got the report accepted by doing some integration; the change in shape translated into a fixed fudge factor in whatever it was we were supposed to measure.
20:40:14 <Phantom_Hoover> How is a triangle wave a function of x or y, unless you use stuff like absolute values.
20:40:45 <fizzie> (Actually it might even have been square.)
20:40:50 <AnMaster> oerjan, oh also that would be: 4*e^(j45°) (you can usually ignore the factor that t is multiplied with if you are working with a single frequency iirc)
20:40:53 * oerjan swats Phantom_Hoover -----###
20:41:04 <oerjan> absolute values are perfectly good functions!
20:41:06 <AnMaster> oerjan, there are two things to annoy you there. j for imaginary unit
20:41:08 * Phantom_Hoover whimpers.
20:41:11 <AnMaster> and degrees in exponential form
20:41:38 <Phantom_Hoover> Can you even differentiate functions with absolutes?
20:41:50 <oerjan> well ° = pi/180
20:41:58 <AnMaster> Phantom_Hoover, everywhere except in the point 0
20:42:08 <AnMaster> or wherever the tip is
20:42:10 <oerjan> it's not exactly logically inconsistent
20:42:25 <AnMaster> (I mean, it would obviously not be in 0 if you did |x+1|)
20:42:43 <Phantom_Hoover> But if we have y=|x|, what is dy/dx?
20:42:46 <oerjan> Phantom_Hoover: not into functions, no. who says all functions have to be differentiable?
20:43:04 <Phantom_Hoover> No-one, but it irks me.
20:43:07 <AnMaster> Phantom_Hoover, would you treat division as a function then?
20:43:09 <AnMaster> as in
20:43:11 <AnMaster> 1/x perhaps
20:43:15 <AnMaster> yes lets take that one
20:43:19 <oerjan> however there is something called distributions, corresponding to such differentials and stuff
20:43:29 <AnMaster> and you more or less indicated above that sin() is a function from your point of view
20:43:38 <AnMaster> Phantom_Hoover, now differentiate sin(1/x) :P
20:43:57 <AnMaster> in fact, I want to know the value of it in x=0
20:44:09 <AnMaster> of (dx/dy that is)
20:44:17 <Phantom_Hoover> Of course, it's undefined at 0.
20:44:27 <AnMaster> Phantom_Hoover, and so is the limit
20:44:37 <oerjan> Phantom_Hoover: more importantly, absolute values are continuous and so triangle waves have perfectly good fourier series
20:44:51 <fizzie> The square wave shape is, I think, even worse than the triangle. At least the / and \ slopes of a triangle have a different derivatives; for the square it's just 0 everywhere where it plays nice, and then there's that horreeble jump.
20:45:40 <oerjan> indeed the differential of a triangle wave is intuitively a square one...
20:45:52 <AnMaster> hah nice
20:47:17 <Phantom_Hoover> Re the pentagonal wave, what's specifically wrong with it?
20:48:25 <Phantom_Hoover> I don't really know about this stuff.
20:48:27 <fizzie> The fact that it has multiple values at a single time?
20:48:27 <oerjan> well nothing really
20:48:33 <AnMaster> yeah what fizzie said
20:48:41 <Phantom_Hoover> fizzie: so does y=sqrt(x).
20:48:46 -!- coppro has joined.
20:48:47 <AnMaster> Phantom_Hoover, no it doesn't
20:48:52 <AnMaster> because sqrt is always the positive one
20:49:00 <AnMaster> that is why you put the "plus or minus" in front of it
20:49:11 <AnMaster> if it wasn't you wouldn't need that "plus or minus" symbol
20:49:15 <Phantom_Hoover> sin-1x?
20:49:23 <AnMaster> Phantom_Hoover, ?
20:49:30 <Phantom_Hoover> Inverse sin?
20:49:30 <AnMaster> sin of what?
20:49:32 <oerjan> Phantom_Hoover: not a function
20:49:38 <Sgeo> There's a pentagonal wave?
20:49:46 <AnMaster> Phantom_Hoover, inverse of sinus is only defined for a limited range
20:49:47 <oerjan> Sgeo: we could invent one
20:49:57 <AnMaster> Phantom_Hoover, and it only maps to a limited range
20:49:59 <Phantom_Hoover> Sgeo: No, I just thought of it in a fit of insanity.
20:50:18 <Phantom_Hoover> AnMaster: I have been *lied* to, then.
20:50:37 <AnMaster> Phantom_Hoover, well it is done in order to make it a function
20:51:05 <AnMaster> Phantom_Hoover, however in calculations like sin(x) = 4 you need to remember the extra posibilities
20:51:10 <AnMaster> and thus would write something like
20:51:21 <AnMaster> actually wait
20:51:23 <AnMaster> that won't work
20:51:24 <fizzie> You could call the pretty boring _/''\_ -style wave, with the slope angle of 72 degrees, a pentagonal wave; it's not like the triangle wave has more than two edges of the triangle either.
20:51:26 <AnMaster> change that equation to:
20:51:33 <AnMaster> sin(x) = 0.5
20:52:38 <AnMaster> x = arcsin(0.5) + pi * n (n in Z)
20:52:39 <AnMaster> or whatever
20:52:50 <AnMaster> I forgot what the repetition rate was
20:53:13 <Phantom_Hoover> Yeah, it's pi.
20:53:17 <Phantom_Hoover> I think.
20:53:28 <AnMaster> I get negative value there, mhm
20:53:33 <Phantom_Hoover> Can the "I think" after everything I say just be implicit?
20:53:34 <AnMaster> so something is off
20:54:03 <oerjan> it's 2pi, but you need two values in that interval
20:54:36 <oerjan> x and pi-x
20:54:40 <AnMaster> ah right
20:54:51 <AnMaster> oerjan, was it cos that had the simpler one?
20:55:10 <oerjan> x and -x
20:55:51 <oerjan> sin (-x) = -sin x, cos (-x) = cos x
20:55:59 <AnMaster> right
20:56:07 <AnMaster> oerjan, It isn't something I use very often
20:56:25 <oerjan> me neither
20:56:26 -!- Gracenotes has quit (Read error: Connection reset by peer).
20:59:31 <AnMaster> oerjan, I find the difference between mathematics and applied mathematics amusing sometimes
20:59:46 <AnMaster> like math teachers telling us to always remember the "plus or minus" when taking square root
21:00:30 <AnMaster> and then when you actually use it you quite often don't need to remember that, because one is absurd (like you can't have a negative ratio between the number of turns in a transformer)
21:01:13 <oerjan> well as long as you're positive about it
21:01:18 <AnMaster> hah
21:02:04 <fizzie> AnMaster: sin x = 4 "works" if you let x be complex; sin(pi/2 + i*arccosh(4)) = 4.
21:02:14 <Phantom_Hoover> oerjan: You are charged with crimes against the English language.
21:02:18 <AnMaster> fizzie, ah, haven't dealt with that stuff at all
21:02:25 <AnMaster> oerjan, or that you don't actually need to solve most things exactly, numerically is good enough as long as you have enough decimals (of course deciding what is enough is another question)
21:03:52 <oerjan> Phantom_Hoover: as long as it is a positive charge
21:04:12 <oerjan> (MWAHAHA)
21:04:20 <fizzie> Ah, there was a lovely joke in the Concrete Mathematics book.
21:04:33 <AnMaster> I think almost all (I don't know if this is in the mathematical sense or not) math teachers I ever had have hated approximate answers. They seem to prefer 4*sqrt(2)/44 rather than something like ~0,115
21:05:06 <oerjan> Phantom_Hoover: ais523 once complained after i had made so many puns he became physically ill
21:05:07 <AnMaster> IMO the latter actually tells you a lot more when reading it
21:05:25 <fizzie> It has student-provided comments scribbled all over the margins; now, in one place the actual book text says something about "looking at sums that involve actual numbers: [a sum expression with j, k and N in it]" and the comment in the margin there was "Caution! The authors of this book seem to think that j, k and N are 'actual numbers'."
21:05:31 <AnMaster> I mean, I can't tell from just looking at "4*sqrt(2)/44" that it is "0.1 and a bit"
21:06:00 <AnMaster> oerjan, when was that?
21:06:10 <oerjan> AnMaster: that's sqrt(2)/11, you infidel!
21:06:16 <oerjan> a few weeks ago?
21:06:22 <AnMaster> fizzie, wonderful
21:06:32 <AnMaster> oerjan, whatever, to me it is ~ 0.115
21:06:55 <AnMaster> oerjan, in practise your measurement tools probably didn't give you that many decimals even
21:07:05 <fizzie> Then you should positively (no pun intended) love floating-point numbers, given that they're inexact all the time.
21:07:33 <AnMaster> most multimeters I have seen tend to give you 3 or 4 significant digits
21:07:55 <AnMaster> (well, some of them seem to vary and be inconsistent even within the same range, but that is another, weird, story)
21:10:12 <AnMaster> I have seen the *same* multimeter display, without me touching anything on it displaying: 1.010 V, 1.01 V, 1010 mV
21:10:28 <coppro> Oo
21:10:34 <AnMaster> and all I did was vary the amplitude on the signal generator used
21:10:49 <AnMaster> to approach the values from different directions
21:11:04 <AnMaster> like, going from just above downwards or just below upwards
21:11:49 <AnMaster> some sort of the delay in switching between V and mV could explain part of it
21:11:53 <Phantom_Hoover> AnMaster: What's wrong with surds?
21:12:00 <AnMaster> but not 1.010 vs. 1.01
21:12:04 <AnMaster> Phantom_Hoover, what
21:12:05 <AnMaster> ?
21:12:33 <Phantom_Hoover> The giving of numbers with square roots and fractions unevaluated.
21:13:04 <AnMaster> Phantom_Hoover, well, it doesn't really give you any sort of feeling for what size the value actually is
21:13:28 <Phantom_Hoover> True, but it's *exact*, and any calculations you make with it are also exact.
21:13:43 <AnMaster> could you tell what 476531489094403/6171473383353049 actually is
21:13:56 <AnMaster> Phantom_Hoover, like just from looking at it
21:14:04 <Phantom_Hoover> No.
21:14:07 <AnMaster> see
21:14:14 <oerjan> !haskell 476531489094403/6171473383353049
21:14:27 <fizzie> Hey, where's fungot?
21:14:28 <EgoBot> 7.721518987342642e-2
21:14:32 <AnMaster> Phantom_Hoover, you can't really tell which fraction is largest if given two ridiculous fractions like that
21:14:46 <Phantom_Hoover> It really depends what you're using it for.
21:14:54 <AnMaster> fizzie, didn't you set up a supervisor for it?
21:15:04 <fizzie> AnMaster: No, I just thought about it. :p
21:15:05 <AnMaster> err that might be erlang speak
21:15:06 <Phantom_Hoover> Halfway through a calculation it's probably better to leave it as a surd.
21:15:35 <AnMaster> oerjan, btw I doubt you can simplify that fraction
21:15:39 <AnMaster> oerjan, since both are prime
21:15:43 <AnMaster> but I might be wrong
21:15:58 <fizzie> I've seen "supervisor" used also in a daemony context.
21:16:03 <AnMaster> right
21:16:07 <Phantom_Hoover> How can you possibly simplify a fraction of two primes?
21:16:15 <AnMaster> Phantom_Hoover, well yes probably, unless that means you end typoing stuff because the expression takes up two lines
21:16:16 <fizzie> Gah, my hostmask is again wrongly.
21:16:22 -!- fizzie has quit (Quit: jumpin' jumpin').
21:16:24 -!- fizzie has joined.
21:16:45 <AnMaster> Phantom_Hoover, I don't know, I'm pretty sure you can't but I might have forgotten something
21:16:51 -!- fungot has joined.
21:17:16 <AnMaster> fizzie, not identified to services?
21:17:24 <AnMaster> why not just edit the host mask in fungot then
21:17:24 <fungot> AnMaster: it just wasn't my brand of humor.
21:17:40 <Phantom_Hoover> AnMaster: Two primes by definition share no common factors, and as such no fraction with them as numerator and denominator can be simplified.
21:17:43 <AnMaster> wait you are now
21:17:45 <AnMaster> hm
21:17:46 <AnMaster> misread
21:18:02 <AnMaster> Phantom_Hoover, perhaps you could break out a root or something silly like that?
21:18:48 <Phantom_Hoover> Well, you can obviously divide top and bottom by anything, but you wouoldn't exactly be simplifying.
21:18:50 <AnMaster> you probably can't what with roots not being rational though
21:19:02 <AnMaster> Phantom_Hoover, indeed
21:19:22 <AnMaster> oerjan, anyway one thing I do agree with mathematicians about is units
21:19:27 <oerjan> !haskell import Ratio; main = print (476531489094403/6171473383353049::Rational)
21:19:31 <EgoBot> 476531489094403%6171473383353049
21:19:36 <AnMaster> oerjan, base units is best
21:19:51 <AnMaster> even if that means 5*10^-9 F rather than 5 nF
21:20:06 <AnMaster> because that is so easy to forget in your calculations
21:20:11 <Phantom_Hoover> F?
21:20:17 <AnMaster> Phantom_Hoover, farad isn't it?
21:20:36 <Phantom_Hoover> Oh, yes.
21:20:47 <AnMaster> Phantom_Hoover, capacitors. And yes the usual range tends to require you to use *10^-something for them
21:20:52 <Phantom_Hoover> For an awful second I though you meant Fahrenheit.
21:20:59 <fizzie> Wasn't it capacitance where the old-fashioned name for a picofarad was μμF, the micro-microfarad?
21:21:07 <AnMaster> fizzie, I never seen that
21:21:33 <fizzie> "A micro-microfarad (μμF) that can be found in older texts is the equivalent of a picofarad." -- An unsourced statement, though.
21:21:34 <AnMaster> fizzie, I have seen nF, µF mostly. pF (?) I don't remember
21:21:43 <fizzie> Yes, but we're not so old.
21:21:44 <AnMaster> ooh wait
21:21:49 <AnMaster> tell me what pH is
21:21:59 <AnMaster> pico-henry right?
21:22:02 <AnMaster> XD
21:22:09 <coppro> lol
21:22:11 <Phantom_Hoover> Or potential of hydrogen.
21:22:18 <Phantom_Hoover> Oh, wait.
21:22:25 <AnMaster> Phantom_Hoover, I prefer picohenry
21:22:48 <AnMaster> Phantom_Hoover, but yes pH is usually the "is it acid or basic" thingy
21:22:48 <oerjan> this is henry, who is very small
21:23:03 <AnMaster> oerjan, henry is a unit for inductors
21:23:10 <AnMaster> it measures, um, something
21:23:16 <AnMaster> I know how to calculate with it thought!
21:23:19 <AnMaster> though*
21:23:20 <Phantom_Hoover> Inductance, presumably.
21:23:23 <fizzie> AnMaster: Yes, and 1 kB is one kilobel.
21:23:24 <AnMaster> Phantom_Hoover, ah yes
21:23:31 <AnMaster> fizzie, bel?
21:23:40 <AnMaster> oh right
21:23:40 <fizzie> AnMaster: The B in dB (decibel).
21:23:42 <AnMaster> as in dB
21:23:49 <AnMaster> fizzie, indeed it must be
21:23:52 <oerjan> is bel an SI unit?
21:24:00 <fizzie> oerjan: It's not exactly, but a bit.
21:24:03 <AnMaster> fizzie, awfully loud noise I guess
21:24:18 <fizzie> oerjan: At least it's on this "Units outside the SI that are accepted for use with the SI" table, but not on the actual "Units of SI" one.
21:24:18 <Phantom_Hoover> Just 10 dB.
21:24:23 <AnMaster> I think units is bad:
21:24:24 <AnMaster> mad*
21:24:26 <AnMaster> You have: dB
21:24:27 <AnMaster> You want:
21:24:27 <AnMaster> Definition: deci B = 0.8 bit
21:24:29 <AnMaster> XD
21:24:35 <oerjan> a kilobel should be enough to shatter the universe
21:24:42 <AnMaster> (just entering a unit is supposed to show it's definition)
21:24:48 <AnMaster> but decibit is just silly
21:25:16 <AnMaster> wait, since dB is logarithmic.... kB would be horribly extremely large
21:25:19 <fizzie> oerjan: Depends on your reference level, of course.
21:25:28 -!- Oranjer has left (?).
21:25:30 <AnMaster> I mean, even outside sound you don't get tha
21:25:31 <AnMaster> that*
21:25:36 <oerjan> AnMaster: ^
21:25:49 <AnMaster> oerjan, [citation needed]
21:25:53 <AnMaster> btw dB it is used in electricity too
21:26:02 <AnMaster> for stuff like amplifiers of signals
21:26:10 <AnMaster> which may or may not be sound
21:26:15 <fizzie> They use it for everything where there's a ratio of two things.
21:27:23 <Phantom_Hoover> How is a bel even defined?
21:27:25 <oerjan> AnMaster: a kilobel is 10^1000 times as loud as 0 bel, that's many orders of magnitude larger than the number of atoms in the visible universe
21:27:45 <AnMaster> 20*log_10(amplification in electronic amplifier) wasn't it?
21:27:48 <AnMaster> something silly like that
21:28:06 <AnMaster> to get the thing in dB
21:29:36 <fizzie> For sound, it's 20*log_10(p_rms/p_ref), where p_rms is the sound pressure level you see, and p_ref is something like 20 µPa (according to Wikipedia); that's around the threshold for hearing in the usual circumstances.
21:30:24 -!- MizardX has quit (Ping timeout: 276 seconds).
21:31:00 <AnMaster> fizzie, I'm talking about http://en.wikipedia.org/wiki/Operational_amplifier kind of stuff. Had a course that included calculation on such recently at university
21:31:43 <AnMaster> the fun thing was calculating on "ideal" such
21:31:53 <AnMaster> like you deal with ideal resistors or inductors or whatever
21:32:13 <AnMaster> I'm not sure the stuff quite would pass the scrutiny of a mathematician
21:32:51 <Phantom_Hoover> What is it with Visual Basic?
21:33:07 <Phantom_Hoover> I see it referred to all over the place, but what is it acually used for?
21:33:07 <AnMaster> eh?
21:33:19 <AnMaster> what?
21:33:30 <AnMaster> visual basic? isn't that a shitty language that microsoft made
21:33:30 <fizzie> Anyway, 10^500 (because the ratio is taken of the squares of the pressure levels) times p_ref would be "just" 2*10^495 Pascals.
21:33:47 <AnMaster> fizzie, I have no idea of what the usual scale is for pascals
21:33:50 <oerjan> it's basically visual
21:33:52 <AnMaster> I don't really deal with sound stuff
21:34:05 <AnMaster> fizzie, didn't you do some sound stuff at university at some point?
21:34:18 <AnMaster> or was that your main area even?
21:34:22 <Phantom_Hoover> AnMaster: Whence "Visual"?
21:34:30 <oerjan> no, everything fizzie did was unsound
21:34:32 <AnMaster> Phantom_Hoover, no idea
21:34:38 <AnMaster> Phantom_Hoover, maybe because it had GUI stuff?
21:34:42 <Phantom_Hoover> Same with Visual C++.
21:34:54 <AnMaster> Phantom_Hoover, iirc that is just a product line of microsoft
21:34:56 <AnMaster> *shrug*
21:34:57 <fizzie> AnMaster: Pascal is a derived unit, 1 Pa = 1 N/m^2. Just imagine some 10^494 kilos of stuff on top of you to get a "gut feeling" of that magnitude.
21:35:15 <AnMaster> Phantom_Hoover, why "Vista", why "XP", why "excel"
21:35:25 <AnMaster> I don't think there is a reasonable answer to any of those
21:35:30 <fizzie> And Visual Basic, compared to Basic, is indeed pretty Visual; at least the development environment is (or used to be) pretty GUI-form-editor-oriented.
21:35:42 <Phantom_Hoover> And what's shitty about it?
21:35:44 <fizzie> I don't know which one they named first, though; Visual Basic or the Visual Studio line of stuff.
21:35:52 <AnMaster> fizzie, ouch
21:36:05 <fizzie> "BASIC rots your brain; Visual Basic is bad for the eyes", like they say.
21:36:18 <Phantom_Hoover> ...Why?
21:36:29 <Phantom_Hoover> I have no experience in any BASIC stuff.
21:36:30 <AnMaster> Phantom_Hoover, you might as well ask why COBOL is bad
21:36:43 <fizzie> AnMaster: If you want to use a timer somewhere, you put an invisible timer element in a GUI window somewhere.
21:36:48 <AnMaster> (not the same reason, but equally hard to explain)
21:36:50 <Phantom_Hoover> Nerds obsessing over minutiae?
21:36:50 <oerjan> my gut feeling seems to be in a black hole right now
21:37:00 <AnMaster> oerjan, heh
21:37:09 <AnMaster> oerjan, that depends on density
21:37:20 <AnMaster> fizzie, sounds like delphi?
21:39:49 -!- coppro has quit (Quit: I am leaving. You are about to explode.).
21:40:56 <fizzie> AnMaster: My exposure to Delphi is minimal, but yes, I think there are similarities.
21:41:14 <AnMaster> *shudder*
21:41:42 <Phantom_Hoover> I was raised on Pascal!
21:41:57 <AnMaster> apple script first for me. Yes I know, horrible
21:41:57 <fizzie> I think they've sensiblized the language (and the environment) at least somewhat for the current, "VB.NET" generation.
21:42:00 <Phantom_Hoover> Though I didn't like it very much.
21:42:03 <AnMaster> then dephi, then sane stuff
21:42:07 <AnMaster> delphi*
21:42:10 <fizzie> You can write web stuff with it nowadays, for example.
21:42:21 <AnMaster> fizzie, you could in theory before I gues
21:42:24 <AnMaster> guess*
21:42:27 <AnMaster> if you had stdio
21:42:39 <Phantom_Hoover> Then I changed schools, and had to use JavaScript.
21:42:54 <fizzie> AnMaster: Well.. you could declare win32 api components, though the FFI was a bit limited.
21:43:07 <AnMaster> heh
21:43:15 <AnMaster> fizzie, what I meant was a cgi
21:44:29 <fizzie> Yes, I don't know how that stuff works in Windows. I don't recall ever seeing a Visual Basic "console application", but it is possible it could make one.
21:46:25 <fizzie> "Even if you use the Win32 API for writing into the console screen, Your application won't work, because the Visual Basic compiler always creates GUI application and it doesn't provide any compiler options for changing it to console application."
21:46:42 <fizzie> There is, however, a third-party binary-mangling tool that translates a GUI PE executable to a console one.
21:47:16 * Phantom_Hoover shivers.
21:47:27 <AnMaster> heh sounds fun
21:47:45 <fizzie> It seems to be a one-byte change.
21:47:58 <Phantom_Hoover> Betwixt what and what?
21:48:12 <fizzie> A "GUI mode" PE executable and a "console mode" one.
21:48:22 <AnMaster> where is that quote from btw?
21:48:33 <fizzie> AnMaster: http://www.nirsoft.net/vb/console_application_visual_basic.html
21:49:24 <fizzie> That might be a bit outdated too. Probably it's about as easy to write a console app with VB.NET as with any other .NET language.
21:49:53 * Phantom_Hoover shivers even more.
21:51:19 <fizzie> "On Error Resume Next" is a nice one-line summary of a Visual Basic mindset.
21:51:34 <AnMaster> heh
21:51:42 <Phantom_Hoover> Wait, so it just skips errors?
21:51:42 <AnMaster> what does it mean?
21:51:55 <Phantom_Hoover> Like /INTERCAL/?
21:52:15 <AnMaster> Phantom_Hoover, why a regex?
21:52:22 <fizzie> It means pretty much what it says. If there ever is a runtime error detected by the system, it is "handled" by resuming execution on the next line.
21:52:23 <AnMaster> (or what the heck where those / / ?)
21:52:30 <fizzie> They could be italics.
21:52:37 <AnMaster> mhm
21:53:13 <fizzie> The other alternative, mind you, is "on error goto x", where x is a line number. (Or a line label, which sounds more likely nowadays.)
21:53:19 <AnMaster> Phantom_Hoover, also iirc intercal doesn't do that sort of thing
21:53:32 <AnMaster> rather it gives you a cryptic error message
21:53:42 <Phantom_Hoover> I thought it just skipped them.
21:53:54 <Phantom_Hoover> And that's how you write comments.
21:53:58 <AnMaster> oh that yes
21:54:03 <AnMaster> but that is syntax errors iirc
21:54:53 <oerjan> no it doesn't skip all syntax errors, but ignores syntax errors in lines that are otherwise skipped (due to starting, e.g. with DO NOT)
21:55:07 <AnMaster> ah right
21:55:47 <Phantom_Hoover> Oooh.
21:58:15 <oerjan> specific lines can have skipping turned on/off with the ABSTAIN/REINSTATE commands
21:58:21 <fizzie> "The statement DO ABSTAIN FROM ABSTAINING is perfectly valid, as is DO ABSTAIN FROM REINSTATING (although this latter is not usually recommended). However, the statement DO ABSTAIN FROM GIVING UP is not accepted, even though DON'T GIVE UP is."
21:58:25 <fizzie> One of my favourite sentences.
21:58:30 <fizzie> Or a pair of 'em.
21:59:03 <AnMaster> heh
21:59:41 <fizzie> I do like the non-sequiturs too, though.
21:59:52 <AnMaster> intercal needs to be... more self-modifying
22:00:08 <fizzie> "--, where exp represents any expression (except colloquial and facial expressions), --"
22:00:15 <AnMaster> heh
22:00:48 <Phantom_Hoover> I had an idea a while back for an "exclude" system.
22:00:50 <fizzie> "No other form of argument is permitted. For example, the following is an invalid argument: [an invalid proof of x != 0, y != 0 => x + y = 0]."
22:01:22 <Phantom_Hoover> Wherein one would export a function to another module of code, rather than importing it.
22:02:48 <AnMaster> fizzie, argument to what?
22:03:10 <fizzie> AnMaster: For REINSTATE, in this case.
22:03:22 <AnMaster> fizzie, also were those [ ] there in the original or was that your summary of it?
22:03:29 <fizzie> That was my summary.
22:03:32 <AnMaster> ah
22:03:47 <fizzie> The original has a nice proof that relies on the ambiguities of English.
22:03:52 <AnMaster> fizzie, link to this?
22:04:22 <fizzie> http://www.muppetlabs.com/~breadbox/intercal-man/s04.html has one HTMLization, section 4.4.10.
22:05:27 <AnMaster> hahaha
22:05:52 <fizzie> You hardly see any line numbers around nowadays either, which is a shame. There's something so pure about an integer label, instead of some silly thing like a name.
22:06:10 <Sgeo> When ais523 comes in, link him to http://www.reddit.com/r/programming/comments/bioxv/while_thinking_about_turing_machines_i_found_that/c0mypya
22:06:17 <AnMaster> even TI-BASIC uses 2 letter labels
22:06:21 <AnMaster> 1 or 2 letters
22:06:26 <AnMaster> I don't know if numbers work
22:06:34 <AnMaster> and I don't have my TI-83+ handy atm
22:09:30 <Phantom_Hoover> Sgeo: Aren't TC languages effectively equivalent to a Turing machine?
22:10:08 <lament> equivalently effective
22:10:45 <fizzie> TI-86's TI-BASIC allows pretty long labels, you just use the "Lbl ABCD" command, and then "Goto ABCD" later. I'm not sure what the length limit there is; four seems to work just fine.
22:10:56 <Phantom_Hoover> So what's the guy complaining about?
22:12:08 <pikhq> Complaints.
22:12:58 <fizzie> Empirical tests seems to suggest that "Lbl" accepts labels up to 8 characters long.
22:14:55 <pikhq> PLEASE ABSTAIN FROM COMING FROM
22:15:07 <pikhq> Man. INTERCAL.
22:16:18 <AnMaster> fizzie, on TI-83+ it was 2 letters
22:16:20 <Phantom_Hoover> fizzie: Presumably that afects TCness.
22:18:34 <fizzie> AnMaster: I don't have the 83+ ROM in the TI emulator in my phone, so can't verify that; I'll take your word for it.
22:20:32 <oerjan> fizzie: unless you can create labels on the fly (and somehow use that for memory), more than 8 characters seems unlikely to help for TCness
22:20:43 <fax> burghgrph'unrbafpun
22:21:09 <oerjan> even 2 letters are probably enough
22:21:30 <fizzie> oerjan: Yes, I do agree. It's got arrays (well, vectors and matrices), but I think those have specified fixed size limits.
22:22:11 <oerjan> oh, i meant that remark for Phantom_Hoover
22:22:22 <oerjan> but thanks for clarification
22:23:33 <Phantom_Hoover> oerjan: Surely a finite number of levels would restrict something important?
22:23:57 <Phantom_Hoover> For instance, you would only be able to nest loops up to a certain number.
22:24:35 <oerjan> Phantom_Hoover: if _everything_ is finite, yes. but you only need one mechanism for getting unbounded memory, labels don't need to be it
22:25:25 <Phantom_Hoover> Hmm...
22:29:48 <oerjan> and even unlimited label size won't help with that as long as you have fixed labels in each single program
22:30:13 <Phantom_Hoover> Yes...
22:30:24 <fizzie> It has strings, and the manual doesn't especially speak of length limits, but of course in practice it does have limits there.
22:30:34 <AnMaster> <fizzie> AnMaster: I don't have the 83+ ROM in the TI emulator in my phone, so can't verify that; I'll take your word for it. <-- on... your... phone?!
22:30:36 <AnMaster> wth
22:30:51 <AnMaster> fizzie, well I do have a ti-83+ emulator somewhere around here anyway
22:31:35 <Quadrescence> Windows Mobile has a TI 83-89 emulator
22:31:37 <fizzie> AnMaster: Yes, I have AlmostTI (which does 85, 86, 82, 83, 83+, 73, 83+se, and maybe 84+ and 84+se) on the phone.
22:31:49 <fizzie> Doesn't do 89, though.
22:32:12 <fizzie> (http://fms.komkon.org/ATI85/)
22:32:25 <AnMaster> yep syntax error with 3 letters
22:32:37 <AnMaster> fizzie, 89 is 68k isn't it?
22:32:40 <AnMaster> while the other are Z80
22:32:42 <fizzie> AnMaster: Yes, I think it is.
22:33:44 <fizzie> oerjan: It's got strings, and those don't have explicit length limits, but you can again derive some implicit limits from the fact that "lngth" needs to be able to return the length of string, and it returns a "number", and the numbers are something like 14-byte BCD-decimal-floats or some-such.
22:34:04 <AnMaster> is there no plain 84 btw?
22:34:07 <AnMaster> without the +
22:34:29 <fizzie> They might have skipped the number.
22:34:32 <AnMaster> Phantom_Hoover, you have loops in TI-83+ btw
22:34:33 <fizzie> I haven't heard of one.
22:34:38 <AnMaster> that are not done with label and goto
22:34:44 <AnMaster> you have for() and while iirc
22:34:51 <AnMaster> anyway the ram is very limited
22:34:58 <AnMaster> so you would run out of that early on
22:35:00 <fizzie> I think there was a nesting limit for the loops, too, but I might misremember.
22:35:05 <AnMaster> maybe
22:35:06 <AnMaster> I don't know
22:35:21 <AnMaster> "The TI-84 Plus is a graphing calculator made by Texas Instruments which was released in early 2004. There is no original TI-84, only the TI-84 Plus and TI-84 Plus Silver Edition models"
22:35:23 <AnMaster> from wikipedia
22:40:01 <AnMaster> what just happened to wikipedia
22:40:05 <AnMaster> everything changed
22:40:14 <AnMaster> horribly unreadable now
22:40:25 <AnMaster> and then it went back to normal
22:40:29 <AnMaster> with ctrl-shift-r
22:40:33 <AnMaster> ?????
22:41:07 <Sgeo> Bug with Beta?
22:41:53 <AnMaster> does beta have a smaller font?
22:41:57 <AnMaster> if so probably
22:42:04 <AnMaster> and kind of bluish theme
22:42:53 <Sgeo> http://mashable.com/2010/03/26/wikipedias-redesign-is-coming-soon/
22:43:30 <AnMaster> looked like a mix of them
22:47:29 <Phantom_Hoover> They're changing it! Therefore it will suck!
22:48:17 <oerjan> http://tvtropes.org/pmwiki/pmwiki.php/Main/TheyChangedItNowItSucks
22:48:25 <oerjan> (*MWAHAHAHA*)
22:53:00 <pikhq> eh, bttr thn th chngs thy md t nglsh. lawlfg!
22:53:56 -!- Oranjer has joined.
23:05:40 <Phantom_Hoover> Oerjan: Linking to TV Tropes? Have you no decency?
23:05:42 -!- adam_d has joined.
23:06:29 <oerjan> i thought my evil laughter would be warning enough
23:10:11 <Phantom_Hoover> Fortunately I was already overexposed.
23:19:43 -!- Azstal has joined.
23:20:46 -!- Aszstal has joined.
23:22:28 -!- Asztal has quit (Ping timeout: 246 seconds).
23:22:39 -!- Aszstal has changed nick to Asztal.
23:24:15 -!- Azstal has quit (Ping timeout: 245 seconds).
23:25:59 <AnMaster> "cpuemu.c:34537: warning: comparison of promoted ~unsigned with unsigned" <-- that is a warning I never seen before
23:26:05 <AnMaster> also what a huge line number
23:26:19 <pikhq> I suggest clang.
23:26:27 <AnMaster> pikhq, oh?
23:26:37 <pikhq> It gives useful errors and warnings.
23:26:42 <AnMaster> pikhq, I doubt it would work. This uses inline asm which llvm doesn't support
23:26:46 <pikhq> And it makes C++ type errors readable.
23:26:48 <AnMaster> specifically x87 asm
23:26:53 <pikhq> Aaaaw.
23:27:00 <pikhq> Wait, x87? Burn it with fire!
23:27:09 <AnMaster> pikhq, *shrug*
23:27:30 <AnMaster> I don't see what is wrong with using that.
23:27:46 <pikhq> It's quite slower than SSE.
23:27:59 <AnMaster> pikhq, sure, but SSE doesn't do long double
23:28:19 <AnMaster> what the heck dcop.cpp:(.text+0x52c): undefined reference to `KApplication::dcopClient()'
23:28:42 <AnMaster> shouldn't ./configure have, you know, detected that
23:29:15 <AnMaster> google says: No results found for "comparison of promoted ~unsigned with unsigned".
23:30:32 <pikhq> Yes, but I fail to see the advantage of an 80-bit floating point type that's hella-slow and unportable.
23:31:17 <AnMaster> pikhq, well because PPC has 128 bit
23:31:24 <AnMaster> while x86 only has 80 bits
23:31:38 <pikhq> x86 doesn't have floating point.
23:31:40 <pikhq> :P
23:31:51 <AnMaster> har
23:33:58 <Phantom_Hoover> Really?
23:34:07 <fizzie> Hm, PPC does real quadruple-precision floats? That's impressive.
23:34:50 <Phantom_Hoover> When would you even need floats that precise?
23:34:54 <Phantom_Hoover> Rocket science?
23:35:01 <AnMaster> fizzie, pretty sure it does
23:35:17 <fizzie> http://en.wikipedia.org/wiki/Long_double speaks only of software implementations, but who knows.
23:36:01 <fizzie> Also, GCC's http://gcc.gnu.org/onlinedocs/gcc/Floating-Types.html says __float80 and __float128 are only supported on x86ish things (the latter with software emulation), but I guess they might still do a "long double" like that.
23:37:59 <fizzie> Another place lists a 128-bit PowerPC "long double" format that's listed to be implemented as "fast software".
23:38:09 <fizzie> That double-double non-IEEE thing sounds bizarre.
23:38:26 <fizzie> "PowerPC supports long double using its fused multiply-add instruction, so its operations are relatively fast, though not fully IEEE-compliant."
23:38:39 <AnMaster> hm
23:38:45 <AnMaster> aha
23:38:55 <Phantom_Hoover> What?
23:38:59 <fizzie> I guess if all you have is a fused multiply-add, all your floats look like... nails? I don't know how that saying goes.
23:41:17 -!- oerjan has quit (Quit: Just don't blow a fuse).
23:41:28 -!- Oranjer has left (?).
23:50:57 <fax> mathematica can't even do diophantine equations over natural numbers....
23:51:20 <Ilari> Isn't there standard IEEE quad-prec float type also (but not widely supported)?
23:52:08 <fax> have to do && x >= 0
23:53:23 <dixon> Because Mathematica sucks.
23:54:15 -!- tombom has quit (Quit: Leaving).
23:59:09 <Slereah> You can't solve general diophantine equations, though
23:59:59 -!- jcp has quit (Read error: Operation timed out).
←2010-03-25 2010-03-26 2010-03-27→ ↑2010 ↑all