| --- Log opened Sat Dec 08 00:00:40 2012 |
| 00:01 | <&ToxicFrog> | That just means that whether you consider it a modifier or a classification in its own right, OO is a property of how the code is written, not the language, which I thought went without saying. |
| 00:01 | <&ToxicFrog> | I mean, you can write non-OO Java, too. |
| 00:24 | | shawn-p [Shawn@Nightstar-4db8c1df.mo.charter.com] has joined #code |
| 00:39 | | You're now known as TheWatcher[T-2] |
| 00:41 | < RichyB> | So, Lampson. Currently working at Microsoft. |
| 00:41 | | You're now known as TheWatcher[zZzZ] |
| 00:41 | < RichyB> | Responsible for a bunch of neat concurrent and distributed protocols like the Baker's Algorithm and Paxos, for instance. |
| 00:42 | < RichyB> | http://research.microsoft.com/en-us/um/people/blampson/33-hints/webpage.html <- is a pretty good writer, too. |
| 00:51 | | Vornicus [Vorn@Nightstar-35eb62f8.sd.cox.net] has joined #code |
| 00:51 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 02:02 | | mac [mac@Nightstar-fe8a1f12.il.comcast.net] has joined #code |
| 02:03 | < mac> | ok can some one explain the logic behind not making things in 64 bit, (clients use 64 bit computers and Os's) |
| 02:11 | <~Vornicus> | 32 bit is still more compatible. |
| 02:11 | <~Vornicus> | Also you don't have to recompile older libraries et al to use them |
| 02:13 | <~Vornicus> | --and you probably don't have the source for those in any case |
| 02:15 | < mac> | my face hurts? its the year 2012, and i 64 bit is still offered on windows. I think I'm going to go break some windows??. |
| 02:15 | <&ToxicFrog> | mac: if all of the software you depend on is 64-bit compatible, and all of your targets are 64-bit? There is no reason. |
| 02:15 | < mac> | 32 bit * |
| 02:16 | < mac> | toxic i guess they just want to be able to throw it on any computer. not really a big deal but i mean come on. |
| 02:17 | <~Vornicus> | But not all software is 64-bit compatible at the library level |
| 02:17 | < mac> | any guess as to when 32 bit will be completely deprecated? |
| 02:17 | <~Vornicus> | mac: they just this week announced that windows 3.11 would no longer be available on the OEM channel |
| 02:17 | <~Vornicus> | So, you know |
| 02:17 | <~Vornicus> | It might be a while. |
| 02:18 | <&ToxicFrog> | For people targeting windows? Not until XP is dead at the earliest. |
| 02:18 | <&ToxicFrog> | So like 2076 or something. |
| 02:18 | < mac> | lololol |
| 02:18 | < RichyB> | There are a few advantages. |
| 02:18 | <~Vornicus> | The computer I'm currently typing at you on is running win7 32bit |
| 02:19 | < RichyB> | Pointers are half the width, some programs whose memory use is mostly in trees of pointers end up using a lot less memory with the ia32 ABI than with the amd64 one. |
| 02:19 | < RichyB> | That translates into fewer L1/L2 cache misses. |
| 02:19 | < mac> | this is true. |
| 02:19 | <~Vornicus> | And pointers get you a shockingly long way |
| 02:19 | <&ToxicFrog> | (more seriously, in 2038 time_t exceeds the bounds of a int32_t, and in 2106 it exceeds uint32_t and everyone needs to be on 64-bit at that point) |
| 02:20 | <&ToxicFrog> | (everyone keeping time, anyways) |
| 02:20 | < RichyB> | amd64 has perf advantages over ia32 as well though, because it has a larger register file and a few other improvements. |
| 02:20 | <~Vornicus> | my Settlers of Catan db schema was something like 95% ID numbers |
| 02:20 | < RichyB> | For most programs, amd64 vs ia32 is a wash, performance-wise. |
| 02:20 | <&ToxicFrog> | Vornicus: my point was more that "XP is dead" is the first point at which we can say every version of windows in the wild has a working 64-bit build. |
| 02:20 | < mac> | yay. 2038 is my hero |
| 02:21 | < RichyB> | However, the Penguin Cult have a thing that's in mainline Linux recently called the x32 ABI, where they use the amd64 instruction set but restrict userland pointers to 32 bits. |
| 02:21 | < mac> | wait a min here. can you still use a 32bit pointer in a 64 bit program? |
| 02:21 | < mac> | can't you* |
| 02:22 | < RichyB> | Not unless you switch on all of the ia32-backwards-compatibility hacks that force the OS to always hand out addresses under the 4GB barrier. Which is how the x32 ABI works. |
| 02:23 | <~Vornicus> | Which is to say that you don't get the real advantages of 64bit processors. |
| 02:23 | < RichyB> | You deliberately sacrifice the wide address space. You don't lose the wider registers or the extra registers. |
| 02:23 | < RichyB> | The 64-bit speedup to algorithms like ElGamal and RSA speeds, as does the register pressure improvement. :) |
| 02:24 | < mac> | 128 bit coming to a world near us? |
| 02:24 | <~Vornicus> | Heard tell of a few 128bit procs but they don't get much play because, well |
| 02:24 | <~Vornicus> | 16 exabytes /really should/ be enough for everybody. |
| 02:24 | < RichyB> | IBM has a line of machines that have had 128-bit pointers for a long while. |
| 02:25 | < mac> | how is gpu computation done are they a bunch of 16 bit full adders? or what |
| 02:26 | < RichyB> | mac: read that, ask about any of the terms that you don't understand: http://www.yosefk.com/blog/simd-simt-smt-parallelism-in-nvidia-gpus.html |
| 02:26 | < RichyB> | Uh, that's a fairly comprehensive and realistic answer, but you might find it heavy-going. |
| 02:28 | <~Vornicus> | I also don't know of anythign that uses the ieee754 quad float, but something's got to be out there. most people who use over-8-byte basic types end up with /arbitrarily/ big basic types, as far as I've seen |
| 02:28 | < mac> | this parallel ability seems to look mighty impressive. RichyB have you messed around with cuda? |
| 02:29 | < RichyB> | The shorter answer is... you start with a 32-bit floating-point ALU. Now put 16 of them, and share an instruction-decoding unit between all of them (so they all execute the same instruction in parallel). Now make them barrel-processors, which means that they context-switch on every clock cycle in order to hide latency. Now repeat that whole unit a bunch of times and stick a really wide RAM interface on it. |
| 02:31 | < RichyB> | So for example, instead of issuing a load from memory and then waiting 8 cycles for the result to show up, you issue a load from memory and go and spend (at least) 7 cycles running instructions for other threads while the load is being resolved. |
| 02:31 | < mac> | humm ok, i thought for some reason they could do 128 bit calculations natively |
| 02:31 | < RichyB> | ? is what the "barrel-processor" bit is about. |
| 02:31 | < RichyB> | Uh, you might come across 128-bit, 256-bit or even 512-bit wide registers in SIMD units |
| 02:31 | < RichyB> | um, terminology fail |
| 02:31 | < RichyB> | Uh, you might come across 128-bit, 256-bit or even 512-bit wide registers in vector units |
| 02:32 | < RichyB> | What those are used for in practice is that one 128-bit wide register stores four 32-bit wide floating point values, though, not a single 128-bit value. |
| 02:33 | < RichyB> | You end up writing code like, "vec4f a, b, c; a = { 0, 0, 0, 1 }; b = { 1, 2, 3, 4 }; c = a + b;" and c ends up with the value { 1 2 3 5 }. |
| 02:34 | < RichyB> | Really wide individual integers are problematic because even fairly simple operations like addition end up having a lot of latency to compute all the carry bits. |
| 02:34 | < mac> | yeah. |
| 02:34 | < mac> | i just thought perhaps the gpu would have a slight advantage. |
| 02:34 | < RichyB> | But really wide vectors of individually narrow types don't have any latency problems, you just ram 'em through arrays of parallel ALUs! |
| 02:35 | < RichyB> | I have never used CUDA in anger (i.e. in shipping code). |
| 02:35 | < RichyB> | Nor any OpenCL stuff. |
| 02:37 | < mac> | do you know if addition is easier than multiplication in terms of instruction load? |
| 02:37 | < RichyB> | Honestly, I'd rather stick with OpenCL, even if NVidia's CUDA tools are nicer and more performant. OpenCL is a fairly sane vendor-neutral standard (I've read the OpenCL 1.0 spec, it's very well designed), whereas using CUDA just seems to me like bending over, dropping trou and screaming "TAKE ME, VENDOR LOCK-IN, PRICE-GOUGE AS MUCH AS YOU WANT!" in some technological/financial equivalent of a prison shower. |
| 02:38 | < gnolam> | Heh. |
| 02:38 | < mac> | ah lol |
| 02:38 | < gnolam> | I've never actually used CUDA. Only OpenCL. |
| 02:38 | < gnolam> | And I didn't really have anything against it at the time. |
| 02:39 | < RichyB> | Outside work, I spend too much time playing mad videogames like Stalker and They Bleed Pixels to leave myself time to play with cool tech like openCL. |
| 02:40 | < RichyB> | Inside work, I write imperfect web applications in some damn Python framework written by druggies and drunks. |
| 02:40 | < mac> | lol |
| 02:40 | < RichyB> | (They're not stupid at all. They're mostly very smart people.) |
| 02:40 | < RichyB> | (Smart people who very much enjoy their herbs and hops, that is.) |
| 02:40 | < mac> | lol, i seem to know exactly what type of people you're talking about. |
| 02:41 | < mac> | It's perplexing to me why would do such a thing. |
| 02:42 | < RichyB> | Because drinking is fun if you have the temperament for it, and burning herbs makes most people giggle a lot? |
| 02:58 | | Attilla [Obsolete@Nightstar-5c0fce37.range109-151.btcentralplus.com] has quit [Ping timeout: 121 seconds] |
| 03:02 | <@Alek> | [13:55] <Syk> http://reddrgn.net/xbmc.jpg < i love xbmc and its skins |
| 03:02 | <@Alek> | Boom de yada? |
| 03:05 | | * Vornicus plays WaDF, gets the rubber ball. omg so damn bouncy |
| 03:06 | <@Alek> | what's WaDF? |
| 03:06 | <@Alek> | Wat Da F? |
| 03:06 | <~Vornicus> | Within a Deep Forest |
| 03:06 | <@Alek> | mmkay. *shrugs* |
| 03:06 | < RichyB> | Vornicus: I love that game! |
| 03:07 | < RichyB> | I remember I used to try to speedrun it while using the glass ball as much as possible. |
| 03:08 | < RichyB> | Freaked my flatmate the Hell out doing that, because he could barely get through the game because it had so many innovative ways to die :) |
| 03:09 | <~Vornicus> | It is a great game though very frustrating |
| 03:09 | <~Vornicus> | Especially because apparently I'm really good at missing platforms. |
| 03:14 | < RichyB> | Got the glass ball yet? |
| 03:14 | < RichyB> | You're going to enter a whole new universe of frustration with that. ;) |
| 03:15 | <&McMartin> | Glass ball is only required at two points, only one of which is mandatory |
| 03:15 | <&McMartin> | It's only a serious problem if you try to cheat like Hell with it |
| 03:15 | <&McMartin> | Which actually lets you skip something like 70% of the game |
| 03:15 | <~Vornicus> | I'm working my way through a sequential run |
| 03:15 | | * RichyB wasn't aware of a sequence-break involving the glass ball. |
| 03:16 | <~Vornicus> | The glass ball has no maximum velocity. |
| 03:16 | <&McMartin> | RichyB: You can enter the Shadowlands with the glass ball by abusing its failure to die on corners. |
| 03:16 | < RichyB> | I just used it because it has amazing lateral speed, so if you're careful about how you bounce, it can be really useful to speed-running. |
| 03:16 | <~Vornicus> | And you can bounce it off corners without it shattering. |
| 03:16 | < RichyB> | Corners? oO |
| 03:16 | <&McMartin> | It's possible to win the game with nothing but the Practice, Iron, Glass, Cold, and Flight ball. |
| 03:17 | <&McMartin> | ... and Pathetic Ball. |
| 03:17 | <&McMartin> | Which you get after Flight. |
| 03:17 | <&McMartin> | Practice, Cold, Flight, and Pathetic ball have an unavoidable hard gear check on them. |
| 03:17 | < RichyB> | Huh. |
| 03:17 | <&McMartin> | There's a fairly amazing speedrun for this. |
| 03:17 | <&McMartin> | (It is *intended* that the endgame not be unlockable until you have gotten every ball in the game.) |
| 03:18 | <~Vornicus> | I've played the game and beaten it several times on hard but I always forget how hard it really is |
| 03:19 | <&McMartin> | (There are hard gear checks on Pathetic and Flight Ball, and there is *supposed* to be a backwards chain of requirements that hit all the other ones.) |
| 03:19 | <~Vornicus> | but the glass ball can replace jump, speed, and rubber ball. |
| 03:20 | < simon`> | http://hpaste.org/78941 -- I'm abusing the State monad for several things at once. I think I should probably have some separation, but I want to KISS. one consequence is that I explicitly need to wipe a part of the state. is this bad Haskell? |
| 03:20 | <~Vornicus> | and if you don't need the rubber ball you don't need the yoga ball. |
| 03:21 | < simon`> | (I forgot to say that checkProgram = mapM_ checkModule) |
| 03:21 | < RichyB> | simon`: don't think so. There's #haskell on freenode too. |
| 03:22 | <&McMartin> | And it turns out the Normal Ball isn't *strictly* required to get Iron. |
| 03:22 | < simon`> | RichyB, ah, right. |
| 03:22 | < RichyB> | State monad with a big wide record is perfectly sensible. |
| 03:23 | < simon`> | RichyB, someone suggested a stack of monad transformers, one for each concern... but I've never used them, and the entire reason for my wide State is because I want this to be *easier* than writing a recursive function with a lot of parameter passing. |
| 03:23 | < simon`> | cool, cool :) |
| 03:23 | < simon`> | s!used them!used Reader/Writer/ |
| 03:24 | < RichyB> | Are some of those inputs and some of those outputs? |
| 03:25 | <~Vornicus> | I forgot also that the shadowlands are meeeaaan |
| 03:25 | < simon`> | to the monad? |
| 03:25 | < RichyB> | To the checkModule action, yes. Are you using some of the things inside the state as outputs that you update as you go along and will want to read out at the end? |
| 03:26 | < simon`> | yes |
| 03:26 | < simon`> | I've divided my wide state into "things I will wipe after having checked each module" and "things I will return as output of the type check so I don't have to traverse the AST again. |
| 03:27 | < RichyB> | Yeah, okay, I'd use Control.Monad.RWS. |
| 03:27 | < RichyB> | RWS is Reader-Writer-State. |
| 03:27 | < simon`> | ah. |
| 03:28 | < simon`> | heh. and then there's an RWST. |
| 03:29 | < RichyB> | The things that you're putting out as output, you spew with (MonadWriter m)'s "tell" function, the things that are constant, you pick up with (MonadReader m)'s "ask" action, and the stuff that's actually genuine state internal to the computation, you put in the State. |
| 03:29 | < RichyB> | Yeah, wiping part of your state between computations is indeed.. error-prone. |
| 03:30 | < RichyB> | RWS is just a convenient shorthand for one of the most common monad-transformer stacks, though. mtl is actually rather easy to use. :) |
| 03:31 | < simon`> | right... I did use ErrorT once. that was amazingly easy. |
| 03:31 | < simon`> | basically, it just grabbed all my "error" calls and wrapped them. |
| 03:31 | | * Vornicus finds himself wondering what kind of action game puzzles can be randomly generated and still come up with a sensible challenge |
| 03:31 | < RichyB> | simon`: oh and if you get stuck on the fact that MonadWriter's tell function requires that the thing you're spitting out be a Monoid, remember that there's a Monoid instance for (a -> a). :) |
| 03:36 | < RichyB> | e.g. |
| 03:36 | < RichyB> | ($[]) . snd . runWriter $ do { tell ('b':); tell ('a':); tell ('c':); tell ('o':); tell ('n':); } |
| 03:38 | < RichyB> | From right to left: runWriter $ do { ... } is a (WriterT Identity ([Char] -> [Char]) ()) action that emits individual functions which prepend one character each of the word "bacon" to a string |
| 03:38 | < RichyB> | er |
| 03:38 | < RichyB> | runWriter turns that do { ... } into a pair, ((), (String -> String)) |
| 03:39 | < RichyB> | snd picks out the (String -> String) function |
| 03:39 | < RichyB> | ($ []) applies the (String -> String) function to an empty String |
| 03:39 | < RichyB> | leaving you with a String, with the contents "bacon". |
| 03:39 | <~Vornicus> | or rather what top-down-zelda-ish-game puzzles can be etc etc |
| 03:41 | < RichyB> | simon`: did that make any sense? The idea of emitting functions from a Writer is that, that way, you can be sure that you never accidentally clobber the state that you're writing. |
| 03:46 | < simon`> | RichyB, hmm, interesting. |
| 03:53 | | * Vornicus spends a while at the start of the shadowlands and gets nowhere. |
| 03:54 | | * simon` trying to fix Haskell errors at 5AM for his commit-hooks to like him seems like a wrong idea. |
| 04:02 | | mac [mac@Nightstar-fe8a1f12.il.comcast.net] has left #code ["Leaving"] |
| 04:05 | < rms> | Vornicus: read the 300 project at all (squidi.net)? |
| 04:05 | < rms> | If not head to his PCG stuff |
| 04:05 | <~Vornicus> | rms: never heard of it until just now, thank you |
| 04:05 | < rms> | He describes how to make colored-key situations possible with PCG'd rooms. |
| 04:05 | < rms> | Among other things |
| 04:06 | <~Vornicus> | oh that thing, I remember seeing that |
| 04:07 | <~Vornicus> | Oooh, I should read like all of these. |
| 04:07 | < rms> | Heh |
| 04:08 | < rms> | Yeah, he's pretty interesting |
| 04:08 | < rms> | Has some real-world game-dev experience. |
| 04:31 | | * Vornicus finds himself considering items in vornda |
| 04:41 | | Kindamoody[zZz] is now known as Kindamoody |
| 04:41 | <~Vornicus> | Oh god he has arts too |
| 04:42 | <~Vornicus> | This might be the start of a beautiful friendship |
| 04:43 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [[NS] Quit: Program Shutting down] |
| 04:56 | | * McMartin checks to see if he was in fact a witchalock last night or not. |
| 04:57 | <&McMartin> | woot, success |
| 04:58 | <&McMartin> | This implies that I should write the rest of these as higher-order functions instead of as part of some abstraction, as well. -_- |
| 05:00 | <&McMartin> | (key insight: the map transforms are all functions of one argument. That means you can chain them with comp, and fixpoint is a generic function as well, jus tnot one built in.) |
| 05:17 | <~Vornicus> | a witchalock? |
| 05:19 | <&McMartin> | I've been doing this transformation kernel stuff for the map editing, as you may recall |
| 05:19 | <&McMartin> | It turns out some of the things I thought I would need to write are in fact part of the Clojure standard library. |
| 05:20 | <&McMartin> | Because when you squint at a transformation kernel, it's Just A Closure and there are functions for those |
| 05:21 | <~Vornicus> | Yep. |
| 05:36 | | * Vornicus gets to work again on Galactic Vorntiers. Is still poking vaguely at load/save, which thus far eludes him. |
| 05:37 | <~Vornicus> | Okay. |
| 05:37 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has quit [[NS] Quit: Leaving] |
| 05:40 | <&ToxicFrog> | How hard can it be, just serialize everything and then write it to disk~ |
| 05:40 | <~Vornicus> | Everything that saves (which is, well, everything) has a constructor that takes (as named) every single attribute it owns. |
| 05:40 | | * Vornicus replaces ToxicFrog with a penguin sitting atop a television. Let's see if anyone notices. |
| 05:40 | | * ToxicFrog explodes |
| 05:41 | <~Vornicus> | ...true to life, even |
| 05:50 | <~Vornicus> | Everything that saves also has to have a save() command that spews a thing acceptable to json. |
| 05:51 | | * McMartin suggests calling it asJson() |
| 05:51 | <~Vornicus> | Sensible idea |
| 06:01 | <~Vornicus> | Well, not quite asJson() so much as, uh |
| 06:01 | <~Vornicus> | Because it doesn't spew actual json, merely a json-serializable object |
| 06:04 | | Derakon[AFK] is now known as Derakon |
| 06:12 | <~Vornicus> | mrenh. Sometimes I wonder if all the objectifying is worth it; I could technically do all my logic on the serializable object tree |
| 06:31 | | Derakon is now known as Derakon[AFK] |
| 06:34 | <~Vornicus> | Might even be a good idea. |
| 06:44 | <&McMartin> | Mmm |
| 06:44 | <&McMartin> | This code is wildly inefficient |
| 06:44 | <~Vornicus> | the kernel code? |
| 06:44 | <&McMartin> | Yeah |
| 06:44 | <&McMartin> | I need a better way of determining that there has been no change |
| 06:45 | <&McMartin> | It took it 67 seconds to build consistent brickery for my level. |
| 06:45 | <~Vornicus> | yeah, that is a little long even for baking. |
| 06:45 | <&McMartin> | It is not however obvious that the programmer time cost is worth it -_-) |
| 06:48 | <&McMartin> | Oops, obvious derp at one point |
| 06:48 | | * McMartin modifies that, sees if it helps |
| 06:50 | <&McMartin> | Hm. I think it made it worse. |
| 06:50 | <&McMartin> | Oh, and that's because now it has to be consistent. |
| 06:51 | <&McMartin> | Well, brick-fill can be readily made more efficient. |
| 06:52 | <&McMartin> | It's not really a proper flood-fill algo right now. |
| 07:15 | <&McMartin> | Yeeeah, 4:39 |
| 07:15 | <&McMartin> | That is unacceptable |
| 07:18 | <~Vornicus> | Quite |
| 07:18 | <&McMartin> | However: correct first, then fast |
| 07:19 | | * McMartin draws a bath |
| 07:19 | | * Vornicus plays WaDF, gets the Ascent Ball |
| 07:19 | <~Vornicus> | I love how you get the most powerful ball in the game and then you must use the least powerful ball in the game. |
| 07:21 | | * syksleep gives Alek a much delayed thwap :| |
| 07:21 | | syksleep is now known as Syk |
| 07:49 | <~Vornicus> | the practice ball zone is a lot easier than I remember it being |
| 07:51 | <~Vornicus> | rms you bastard now I absolutely must build Vornda and to those specs but first I need to actually learn2game |
| 07:53 | | mode/#code [+o Azash] by ChanServ |
| 07:53 | | mode/#code [+o celticminstrel] by ChanServ |
| 07:53 | | mode/#code [+o Courage] by ChanServ |
| 07:53 | | mode/#code [+o cpux] by ChanServ |
| 07:53 | | mode/#code [+o EvilDarkLord] by ChanServ |
| 07:53 | | mode/#code [+o froztbyte] by ChanServ |
| 07:54 | | mode/#code [+o gnolam] by ChanServ |
| 07:54 | | mode/#code [+o iospace] by ChanServ |
| 07:54 | | mode/#code [+o Namegduf] by ChanServ |
| 07:54 | | mode/#code [+o Orthia] by ChanServ |
| 07:54 | | mode/#code [+o Pandemic] by ChanServ |
| 07:54 | | mode/#code [+o Reiv] by ChanServ |
| 07:54 | | mode/#code [+o Rhamphoryncus] by ChanServ |
| 07:54 | | mode/#code [+o rms] by ChanServ |
| 07:54 | | mode/#code [+o simon`] by ChanServ |
| 07:55 | | mode/#code [+o Tarinaky] by ChanServ |
| 07:56 | < Syk> | owO |
| 07:58 | | * Vornicus handed out hats to everyone he remembers as being around for quie a while. |
| 07:58 | < Syk> | also |
| 07:58 | < Syk> | bahahahah the USPTO preliminary invalidated the 'steve jobs patent' |
| 07:58 | <~Vornicus> | haa-haa! |
| 07:59 | <~Vornicus> | </div class="nelson-muntz"> |
| 07:59 | < Syk> | ""touch screen device, method, and graphical user interface for determining commands by applying heuristics" - 20/20 claims invalidated |
| 08:01 | < Syk> | Vornicus: class in a closing div tag? :P |
| 08:02 | <~Vornicus> | shut up |
| 08:02 | <~Vornicus> | I had to put it somewhere |
| 08:03 | < Syk> | man why do i even browse CNet |
| 08:03 | < Syk> | "That said, Google Android devices may be more vulnerable to attacks than the iPhone. The reason is that Apple's App store is considered to be more secure due to the fact that each app must be approved to be listed in the store for download." |
| 08:04 | < Syk> | I don't think anyone remembers when the iPhone was jailbreakable via *a PDF file* |
| 08:05 | <~Vornicus> | silliness. |
| 08:07 | | mode/#code [+o Thrae] by ChanServ |
| 08:07 | | mode/#code [+o Zemyla] by ChanServ |
| 08:07 | <~Vornicus> | K. Rest of you need to either identify or stick around. :P |
| 08:07 | | celticminstrel is now known as celmin|sleep |
| 08:08 | <@Tarinaky> | Does anyone know if KSP has some components for err... reconfigurable hulls? Like, parts that can extend/retract after lift off? |
| 08:13 | <@froztbyte> | iospace: aaaaaaamen |
| 08:13 | <@froztbyte> | (re tcl) |
| 08:16 | <@froztbyte> | and I'm too asleep right now to read through the rest of the arch discussion, but it looks like it has some nice detail and I'll definitely revisit it :3 |
| 08:40 | | Vornicus [Vorn@Nightstar-35eb62f8.sd.cox.net] has quit [[NS] Quit: Leaving] |
| 08:51 | | Kindamoody is now known as Kindamoody|out |
| 09:58 | < auREAX> | tfw no op |
| 10:02 | | You're now known as TheWatcher |
| 10:02 | | mode/#code [+oooo shawn-p AnnoDomini auREAX Syk] by Tamber |
| 10:03 | <@Tamber> | :p |
| 10:05 | <&McMartin> | Hum |
| 10:05 | <&McMartin> | OK, Debian and/or Ubuntu dudes |
| 10:05 | | * Syk aughs |
| 10:05 | <&McMartin> | Is there a way to parallel-install the 32-bit version of a package? |
| 10:05 | <@Syk> | McMartin: um |
| 10:05 | <@Syk> | parallel? |
| 10:05 | <&McMartin> | Dudettes may also answer, but really, "Dude" has been gender-neutral for awhile |
| 10:05 | <&McMartin> | Specifically, Mono |
| 10:05 | <@Syk> | McMartin: <package>:i386 |
| 10:05 | <&McMartin> | I have 64-bit Mono installed |
| 10:06 | <&McMartin> | oK |
| 10:06 | <@Syk> | i think |
| 10:06 | <&McMartin> | That won't wreck the pre-existing 64-bit install, then? |
| 10:06 | <@Syk> | you might need to enable multiarch |
| 10:06 | <@Syk> | no idea how to do it |
| 10:06 | <&McMartin> | Hrm |
| 10:06 | <&McMartin> | ISTR Ubuntu's bad at that |
| 10:06 | <&McMartin> | Well, let's see |
| 10:06 | <@Syk> | McMartin: if the packager wasn't an insane fuck, yearh |
| 10:06 | <&McMartin> | What Could Possibly Go Wrong (tm) |
| 10:06 | <@Syk> | since there's /lib and /lib64 or something |
| 10:07 | <&McMartin> | Nope, screwed |
| 10:07 | <&McMartin> | Oh well |
| 10:07 | <@Syk> | wut |
| 10:07 | <&McMartin> | The following packages have unmet dependencies: mono-runtime:i386 : Depends: mono-gac:i386 (= 2.10.8.1-1ubuntu2.2) but it is not installable |
| 10:07 | <&McMartin> | E: Unable to correct problems, you have held broken packages. |
| 10:07 | <&McMartin> | "mono-runtime" itself is in fact installed and sitting pretty |
| 10:07 | <@Syk> | is it not installable because multiarch isn't on? |
| 10:08 | <&McMartin> | I don't think 12.04 supports multiarch at all |
| 10:08 | <@Syk> | yes it does |
| 10:08 | | * Syk points at 32bit Steam on her 64bit Linux |
| 10:09 | <&McMartin> | Yeah |
| 10:09 | <&McMartin> | That's exactly what's breaking |
| 10:09 | <&McMartin> | Specifically, SpaceChem |
| 10:09 | <&McMartin> | Which is trying to run in the system's mono, but can't because that 32-bit steam .so won't load into 64-bit mono |
| 10:10 | <@Syk> | i would call that Mono being absolutely useless |
| 10:10 | <&McMartin> | No, DLLs work like that everywhere |
| 10:10 | <@Syk> | i mean the package situation |
| 10:10 | <&McMartin> | This is why you would need a 32-bit version of mono to make this work |
| 10:13 | <&McMartin> | http://steamcommunity.com/app/221410/discussions/1/882965737405647921/ |
| 10:18 | <&McMartin> | Ha ha, yeah, the package situation is just fucked |
| 10:19 | <&McMartin> | It's failing because Debian has to break package names on upgrades. |
| 10:19 | <&McMartin> | It's a COMPULSION |
| 10:19 | <&McMartin> | mono-gac:i386 is uninstallable because now it is mono-4.0-gac |
| 10:19 | <&McMartin> | And mono-runtime doesn't know that in the 32-bit case |
| 10:19 | | * McMartin golfclaps |
| 10:20 | <@TheWatcher> | .... what |
| 10:23 | <&McMartin> | mono-runtime:i386 is not installable because it wants mono-gac:i386, which no longer exists because Mono 4.0 came out and so ALL PACKAGE NAMES MUST CHANGE because Debian is full of psychopaths |
| 10:24 | <&McMartin> | mono-runtime:amd_64 works fine (and is what I have), but that cannot load the Steam version of SpaceChem because it cannot link against the relevant libraries. |
| 10:24 | | * TheWatcher has this mental picture of a team of Mad Hatters, all yelling "CHANGE PACKAGES!" in unison |
| 10:25 | <&McMartin> | I take a lot of shit for preferring Yum, but it's really more that I prefer that Fedora *doesn't do this* |
| 10:25 | <@Azash> | Please stop using what I don't use, McMartin |
| 10:25 | <@Azash> | :p |
| 10:25 | <&McMartin> | Debianists swear up and down that it's a feature because *of course* it would be madness to automatically update people to Python 2.7 from Python 2.6 |
| 10:26 | <@TheWatcher> | Of...course. |
| 10:26 | | * McMartin sets up downloads for everything else he can run, to give it a whirl |
| 10:26 | <&McMartin> | Or rather, to see if I can in fact run them |
| 10:26 | | * McMartin also pokes Syk |
| 10:26 | <&McMartin> | You never tried to friend me! |
| 10:27 | <&McMartin> | (The Steam client itself has been doing a lot better than I expect from Beta software, tbh) |
| 10:28 | <&McMartin> | (And I'm not even running the edge drivers, just sticking with the stock ones for Intel in Precise) |
| 10:29 | <&McMartin> | Azash: There are various technical arguments for why yum is inferior to apt, in a vaccuum |
| 10:29 | <&McMartin> | I don't begrudge or indeed even gainsay this |
| 10:29 | <@Azash> | There are, sure |
| 10:29 | <&McMartin> | But my computers are not IN SPAAAAAAAAACE |
| 10:29 | <@Azash> | Haha |
| 10:30 | <@Azash> | Still, mostly it strikes me as clutching at straws to justify the One Distro That I Use |
| 10:30 | <@TheWatcher> | NOYUMINSPACE doesn't quite have the same ring to it... |
| 10:30 | <&McMartin> | Oh |
| 10:30 | <&McMartin> | Yeah, so, I use three distros constantly |
| 10:30 | <&McMartin> | That's just how it goes |
| 10:31 | <&McMartin> | The text server is Fedora, the laptop's Ubuntu, and the phone is Android. :D |
| 10:34 | | * Tamber uses Debian on servers because meh, it works; Gentoo on the desktop because "all the sources! o/"; and FreeBSD on the laptop, because it's something slightly different to play with. |
| 10:35 | <@Tamber> | And the One True Distro is, of course... *drumroll* ...whatever you damn well feel like and your crap runs on. :) |
| 10:36 | <&McMartin> | Well |
| 10:36 | <&McMartin> | That's part of why the laptop is Ubuntu - it's what everyone else seems to target |
| 10:36 | <&McMartin> | I try to be triplatform whenever feasible, so. |
| 10:36 | <@Tamber> | *noddle* |
| 10:38 | <@Tamber> | I vaguely recall not getting along well with Ubuntu for some reason; and Debian was playing silly buggers last time I tried to install All The -dev packages! that I needed for some code I was tinkering with. Works fine for most other things, though. Ho-hum. |
| 10:38 | <&McMartin> | Yeah |
| 10:38 | <&McMartin> | Ubuntu led the charge on the LOOK LIKE A TABLET thing |
| 10:38 | <&McMartin> | They've gone off into a new crazytown now in 12.10 |
| 10:39 | <@Tamber> | Oh, last time I used it, it was before that. |
| 10:39 | <&McMartin> | But 12.04 I find actually fairly usable because Unity Dash is generally exactly what I want from a start menu |
| 10:39 | <&McMartin> | That is, I want a command line with autocomplete on all installed applications and recently used documents. =P |
| 10:39 | <@Tamber> | :) |
| 10:39 | <&McMartin> | This also means that Unity is a whole lot less LIKE A TABLET OMG than it used to be |
| 10:40 | <@Tamber> | Sudden outbreak of sanity? |
| 10:41 | | * Azash will still stick to xubuntu |
| 10:41 | <&McMartin> | It's honestly hard to tell |
| 10:42 | | * Tamber sighs at debian on the home server. |
| 10:42 | <@TheWatcher> | Tamber: could be random chance, really |
| 10:42 | <@Tamber> | TheWatcher, stopped clock, right twice a day, and all that? |
| 10:43 | <@TheWatcher> | Yeah |
| 10:43 | <&McMartin> | It doesn't "feel" like that to me |
| 10:43 | <@TheWatcher> | Except the town hall clock here, that somehow manages to avoid even that. |
| 10:43 | <@Tamber> | ha |
| 10:43 | <&McMartin> | It's one thing to be stopped, it's another to be six hours fast. |
| 10:43 | <&McMartin> | A clock six hours fast is wrong *all the time* |
| 10:44 | <&McMartin> | Tamber: The reason "sudden outbreak of sanity" doesn't work is because nobody else makes the Dash so primal |
| 10:44 | <&McMartin> | Er, primary |
| 10:44 | <&McMartin> | The closest I can come up with is the windows 7 fast search, because Spotlight is total shit~ |
| 10:44 | <&McMartin> | As is GNOME's attempt at it |
| 10:47 | <@Tamber> | In better news, E17 is nearing release; and the Bling theme works again :D |
| 10:48 | | * Tamber pelted with rotten fruit |
| 10:49 | | * McMartin forges the Bling into Blingots, pelts Tamber with those |
| 10:49 | <@Tamber> | *crunch* |
| 10:49 | <&McMartin> | Azash: I shifted the Fedora system over to LXDE, which I don't use much (it's all but headless) but when I need it it's not bad |
| 10:49 | <&McMartin> | (That machine is noticably underpowered these days) |
| 10:50 | <&McMartin> | Hrm |
| 10:50 | <&McMartin> | How does one go about reporting bugs in packages? |
| 10:50 | <&McMartin> | I might as well make the packager complaint |
| 11:03 | | * McMartin blinks at his Clojure code |
| 11:03 | <&McMartin> | ... I can make this lazy |
| 11:03 | <@auREAX> | I prefer to be eager |
| 11:05 | <@Syk> | McMartin: why are you using unity, ever |
| 11:05 | <@Syk> | GNOME 3.6 + Ubuntu 12.10 = love |
| 11:06 | <&McMartin> | GNOME is awful |
| 11:06 | <&McMartin> | Ubuntu 12.10 crashed its default compositor within 5 minutes of my first login |
| 11:06 | | * Tamber borrows from above: "[...] to justify the One DE|WM That I Use" |
| 11:06 | <&McMartin> | I was Not Impressed. |
| 11:07 | | * Tamber replaces all of Syk's UI with twm; let's see how quickly she hunts me down and beats me to death for it. |
| 11:08 | <@Syk> | owo |
| 11:08 | <@Syk> | McMartin: gnome is great |
| 11:09 | <@Syk> | it's a bit derp to start with |
| 11:09 | <@Syk> | but great |
| 11:09 | <@Syk> | also brb |
| 11:09 | | Syk [the@Nightstar-7d752098.lnk.telstra.net] has quit [[NS] Quit: leaving] |
| 11:10 | | Syka [the@Nightstar-7d752098.lnk.telstra.net] has joined #code |
| 11:11 | <@Tamber> | "Opinions are like arseholes |
| 11:11 | <@Tamber> | Everyone has one, and they're all full of shit." |
| 11:11 | | * Tamber wonders where the newline came from, goes to investigate. |
| 11:11 | < Syka> | lul |
| 11:11 | < Syka> | oh fuck, i forgot to put out my washing |
| 11:11 | < Syka> | brb |
| 11:12 | <&McMartin> | I always heard that as ending with "... and nobody wants to see yours" |
| 11:12 | <@Tamber> | hehe |
| 11:14 | < Syka> | so, today i woke up at 4pm. was great c: |
| 11:15 | <&McMartin> | auREAX: In this case it lets me make two currently different functions be replaced with one lazy function and two trivial functions, one of which operates on the whole seq and one of which only operates on the head |
| 11:15 | <&McMartin> | So it's a win here, I think |
| 11:15 | <&McMartin> | But to know for sure we must have SCIENCE |
| 11:15 | | * McMartin times the run |
| 11:16 | <&McMartin> | This is not going to be very good science, as Steam is doing Steamy things in the background. |
| 11:17 | <&McMartin> | But even with that my execution time was cut nearly in half. |
| 11:17 | <&McMartin> | So, laziness wins here. |
| 11:18 | <&McMartin> | But it's still too slow, I'll have to reimplement the core transform entirely so that it's not made of the stuff that happened to already be implemented. |
| 11:19 | < Syka> | http://reddrgn.net/fullscreen.zip < for anyone that cares |
| 11:20 | < Syka> | it's a QT app that opens a full screen instance of qtwebkit |
| 11:20 | < Syka> | syntax is "fullscreen http://<url>" |
| 11:21 | < Syka> | have fun |
| 11:39 | | Orthia [orthianz@Nightstar-6ca59a6f.callplus.net.nz] has quit [Ping timeout: 121 seconds] |
| 11:54 | <@froztbyte> | haha, cool |
| 11:54 | <@froztbyte> | you have a nice penchant for single-shot things |
| 11:54 | <@Tamber> | "Do one thing, and do it sorta okay, I guess." |
| 11:55 | <@froztbyte> | lulz |
| 11:55 | <@froztbyte> | the almost-unix way |
| 11:57 | | Orthia [orthianz@Nightstar-6ca59a6f.callplus.net.nz] has joined #code |
| 11:57 | | mode/#code [+o Orthia] by ChanServ |
| 11:59 | < Syka> | Tamber: :U |
| 11:59 | < Syka> | froztbyte: well it builds into a bigger product |
| 12:00 | < Syka> | a display board, for instance - I can get a raspi, load this on it, and then bam, I have a display driver |
| 12:00 | <@froztbyte> | no, sure |
| 12:00 | < Syka> | if my logic is done server side and the refreshing/etc is done on the client side through JS, I can load a webpage and then there it is |
| 12:00 | <@froztbyte> | that's the nice thing with unix stuff |
| 12:00 | <@froztbyte> | you basically don't need to fuck around with more than what you want |
| 12:01 | < Syka> | i once tried doing something the UNIX way on Windows |
| 12:01 | < Syka> | as part of my helpdesk system |
| 12:01 | < Syka> | it sucked pretty hard |
| 12:01 | < Syka> | but on unixlikes the unix way actually works :3 |
| 12:01 | <@froztbyte> | yup |
| 12:01 | <@froztbyte> | windows doesn't really have the ecosystem quite right |
| 12:01 | <@froztbyte> | they're trying to get it that way |
| 12:02 | <@froztbyte> | but when you basically spent 35+ years entrenching everything with everything else, you've got a lot to do |
| 12:03 | < Syka> | ok so i am going to rip 4 series of seinfeld |
| 12:03 | < Syka> | it looks like the disc layouts are a mess |
| 12:03 | < Syka> | argh :C |
| 12:05 | | Attilla [Obsolete@Nightstar-5c0fce37.range109-151.btcentralplus.com] has joined #code |
| 12:07 | <@froztbyte> | why not just use handbrake or something? |
| 12:07 | <@froztbyte> | that can rip by TOC |
| 12:11 | < Syka> | i am |
| 12:11 | < Syka> | but the ToCs are usually fucked |
| 12:11 | < Syka> | eg. this one has a duplicate entry |
| 12:11 | | Orthia [orthianz@Nightstar-6ca59a6f.callplus.net.nz] has quit [Ping timeout: 121 seconds] |
| 12:11 | <@froztbyte> | lulz |
| 12:12 | <@froztbyte> | full duplicate, or duplicate-with-other-lang ? |
| 12:12 | < Syka> | full duplicate |
| 12:12 | <@froztbyte> | durrrr |
| 12:12 | < Syka> | the red dwarf discs had fucking ep - ep - both eps in one - ep - ep - both eps in one |
| 12:12 | < Syka> | but the both in one were duplicate?? |
| 12:14 | <@froztbyte> | yeah that sounds weird |
| 12:14 | <@froztbyte> | must be some or other arb reason |
| 12:15 | < Syka> | yeah |
| 12:15 | < Syka> | i might try using the other dvd lib |
| 12:30 | <@gnolam> | http://thankstextbooks.tumblr.com/post/26341553264/at-least-theyre-honest |
| 12:33 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 13:22 | | Kindamoody|out is now known as Kindamoody |
| 13:25 | < Syka> | oh thank god |
| 13:26 | < Syka> | the other seinfeld discs are set out sanely |
| 13:37 | | You're now known as TheWatcher[afk] |
| 14:03 | <@Tarinaky> | Does Java have a stock Singleton container? |
| 14:07 | | Kindamoody is now known as Kindamoody|out |
| 14:16 | <@gnolam> | Eww, singletons. |
| 14:17 | <@Tarinaky> | Better question. |
| 14:17 | <@Tarinaky> | How the fuck do I pull the contents of a file in to a String? |
| 14:18 | < Syka> | singletons? sounds like my relationships |
| 14:18 | < Syka> | :P |
| 14:20 | <@Tamber> | *badum-tish* |
| 16:35 | | Vornicus [Vorn@Nightstar-35eb62f8.sd.cox.net] has joined #code |
| 16:35 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 16:57 | <@Tarinaky> | The group project group had a big "lets work on the project" meeting today. |
| 16:57 | <@Tarinaky> | They produced, between themselves, one small class. |
| 17:06 | <@gnolam> | How many of you are there, and how many are actually contributing something of value ot the project? |
| 17:07 | <@froztbyte> | gnolam: or, otherwise asked, how many general-flavour managers are there? |
| 17:08 | | * gnolam ponders starting up the refactor tractor. |
| 17:13 | | * gnolam first ...s at Microsoft. |
| 17:13 | < Syka> | fffff |
| 17:13 | < Syka> | Steam has defcon |
| 17:13 | <@gnolam> | This is what they're calling release notes for Visual Studio 2012 Update 1: http://support.microsoft.com/kb/2707250/en |
| 17:13 | < Syka> | go to install it |
| 17:13 | < Syka> | it's only the fucking soundtrack |
| 17:13 | <@gnolam> | You mean it's an update? For Visual Studio 2012? My god, so much information I didn't already have! |
| 17:19 | <@gnolam> | ... and it's a fucking downloader as well. |
| 17:24 | | celmin|sleep is now known as celticminstrel |
| 17:46 | | Kindamoody|out is now known as Kindamoody |
| 19:18 | <&ToxicFrog> | Syka: er? Steam definitely does have Defcon the actual game. |
| 19:19 | < Syka> | ToxicFrog: not steam for linux |
| 19:19 | <&ToxicFrog> | Aah |
| 19:35 | | Vornotron [vorn@Nightstar-35eb62f8.sd.cox.net] has joined #code |
| 19:58 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [Ping timeout: 121 seconds] |
| 20:04 | | Kindamoody is now known as Kindamoody[zZz] |
| 20:09 | | Vornicus [Vorn@Nightstar-35eb62f8.sd.cox.net] has quit [Connection reset by peer] |
| 20:11 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 20:14 | | Vornotron [vorn@Nightstar-35eb62f8.sd.cox.net] has quit [[NS] Quit: Leaving] |
| 20:16 | | Syka is now known as syksleep |
| 20:16 | | celticminstrel [celticminst@Nightstar-e83b3651.cable.rogers.com] has quit [[NS] Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.] |
| 20:16 | | celticminstrel [celticminst@Nightstar-e83b3651.cable.rogers.com] has joined #code |
| 20:16 | | mode/#code [+o celticminstrel] by ChanServ |
| 21:15 | | VirusJTG_ [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has joined #code |
| 21:15 | | VirusJTG [VirusJTG@Nightstar-09c31e7a.sta.comporium.net] has quit [Ping timeout: 121 seconds] |
| 21:38 | | mac [mac@Nightstar-fe8a1f12.il.comcast.net] has joined #code |
| 21:53 | <@froztbyte> | ahahhahahahahahhahahahaa: http://imgur.com/OxtdL |
| 21:54 | <@Tamber> | Smells like... freedom. ...and toe jam. |
| 22:34 | | Derakon[AFK] is now known as Derakon |
| 22:43 | | You're now known as TheWatcher |
| 22:44 | | Orthia [orthianz@Nightstar-6ca59a6f.callplus.net.nz] has joined #code |
| 22:44 | | mode/#code [+o Orthia] by ChanServ |
| 22:44 | | Vornicus [vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 22:44 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 22:53 | | * Vornicus is now at full power again and can do things. |
| 22:54 | <&McMartin> | Syka: As far as I can tell, only the games listed on the Steam for Linux forum actually have working installers right now |
| 22:54 | <&McMartin> | https://support.steampowered.com/kb_article.php?ref=8495-OKZC-0159 |
| 23:25 | <@Azash> | http://www.6uk.org.uk/2012/12/6uk-powerless-to-encourage-ipv6-adoption-board-res igns/ |
| 23:32 | <@gnolam> | Hah. |
| 23:33 | <@gnolam> | "Many factors impact the uptake of IPv6 and clearly free-market incentives are insufficient. Yet at a country level, delayed adoption significantly impacts national competitiveness, innovation and skills deleteriously. It may also hobble UK based companies facility to compete internationally." <- O RLY? |
| --- Log closed Sun Dec 09 00:00:55 2012 |