| --- Log opened Sun Jul 26 00:00:00 2009 |
| --- Day changed Sun Jul 26 2009 |
| 00:00 | <@Derakon> | Over 13s of runtime, 7s in PhysicsObject.update, 4.2s in rebalancing the quadtree, and the rest in loading assets. This is no good. |
| 00:08 | | You're now known as TheWatcher[T-2] |
| 00:09 | | Attilla [~The.Attil@92.1.240.ns-2748] has joined #code |
| 00:09 | | mode/#code [+o Attilla] by ChanServ |
| 00:11 | | You're now known as TheWatcher[zZzZ] |
| 00:25 | | * Derakon ponders if there's some way to get cProfile to generate some kind of colorized source code output that tells him cleanly how his time is being spent in a more heirarchical fashion than the basic profiler output does. |
| 00:26 | <@Derakon> | Like, the report tells me that I'm spending 3.566s in and under Sprite.getBounds, but how much of that is from collision detection vs. from rebalancing the quadtree? |
| 00:39 | | AnnoDomini [AnnoDomini@Nightstar-29146.neoplus.adsl.tpnet.pl] has quit [Quit: Oh hai. In teh beginnin Ceiling Cat maded the skiez An da Urfs, but he did not eated it.] |
| 00:59 | <@Derakon> | Woot, fixed it! |
| 00:59 | <@Derakon> | Went from 45 objects causing the game to choke and die a horrible death to 45 objects causing no noticeable slowdown whatsoever. :) |
| 01:05 | <@Derakon> | ...I spoke too soon. Things were fast because I was dropping objects on the floor. |
| 01:22 | | * Bobsentme stifles a giggle |
| 01:22 | < Bobsentme> | Still better than me. I can't even get a system ping to work in Qt Creator |
| 01:22 | | Consul [~Consul__@Nightstar-2425.dsl.sfldmi.ameritech.net] has quit [Ping Timeout] |
| 02:10 | | Consul [~Consul__@Nightstar-2425.dsl.sfldmi.ameritech.net] has joined #code |
| 02:10 | | mode/#code [+o Consul] by ChanServ |
| 02:27 | | Attilla [~The.Attil@92.1.240.ns-2748] has quit [Quit: ] |
| 02:39 | | Thaqui [~Thaqui@121.98.166.ns-22683] has joined #code |
| 02:39 | | mode/#code [+o Thaqui] by ChanServ |
| 02:39 | | gnolam [lenin@Nightstar-1382.A163.priv.bahnhof.se] has quit [Quit: Z?] |
| 03:01 | | * Derakon fixes the dropping of objects on the floor, goes back to figuring out why things are so slow. |
| 03:22 | <@Derakon> | Blargh. I hate the prospect of just converting all this stuff over into C/C++, but I'm not seeing any obvious sources of slowdown. |
| 03:23 | <@Derakon> | This function's taking .906s over 16073 calls, just in itself (i.e. disregarding calls to other functions), and I don't know why. http://paste.ubuntu.com/233587/ |
| 03:24 | <@Derakon> | It's also taking about half of the overall runtime, when counting calls to other functions made by it. |
| 03:31 | <@Thaqui> | Well, it may not be the function that's the problem but unnecessary calls? |
| 03:43 | <@Vornicus> | A fun little math problem: to get the coefficients etc of an arbitrary polynomial given a bunch of results from xs you choose, you need degree + 1 points: once you get a value that is expected, you can stop. How many points do you need if the coefficients are restricted to nonnegative integers? |
| 03:44 | <@Derakon> | What do you mean by "given a bunch of results from xs as you choose"? |
| 03:45 | <@Vornicus> | I have a polynomial function f(x). |
| 03:45 | <@Vornicus> | It's a black box though; you can giveit an x and it will return the y. |
| 03:45 | <@Derakon> | Ahh, empirically-derived results. |
| 03:56 | <@Vornicus> | Eh. I'll give you the answer: 2. Now all you have to do is figure out why. |
| 03:59 | <@Derakon> | Well, 2 is clearly enough to establish if the function is linear. |
| 04:00 | <@Derakon> | But it shouldn't be sufficient for anything *more* than that. |
| 04:01 | <@Derakon> | Is there a constant term? |
| 04:01 | <@Derakon> | That is, c_0 * x^0? |
| 05:18 | | * Derakon installs Cython, notes that one of the log lines that goes by is "Picking lexicon..." |
| 05:42 | <@Consul> | Yeah, Cython doesn't like installing on Windows. |
| 05:42 | <@Consul> | At least, not unless you compiled Python from scratch yourself. |
| 05:49 | <@Consul> | Ah, there it goes. I finally got it pointing to mingw |
| 05:50 | <@Consul> | Looks like it's installed. |
| 05:52 | | * Derakon looks up, notes he made a typo. It was "Pickling lexicon...". |
| 05:53 | <@Derakon> | In any event, my goal was to do some speed tests comparing the pure-Python Vector2D class and a Cython version. Sadly, Cython doesn't seem to like my use of __new__ instead of __init__. |
| 05:58 | <@Consul> | Well, my install seems to work. |
| 05:59 | <@Consul> | We'll see if this plus pyVST makes for some interesting stuff. |
| 05:59 | <@Consul> | Which is as yet unfinished, actually. |
| 06:02 | | Syloqs-AFH [Syloq@ServicesAdmin.Nightstar.Net] has quit [Connection reset by peer] |
| 06:03 | <@Derakon> | Hah. Rudimentary test has the pure-Python version being .01s *faster*, in a loop that instantiates some 1 billion Vector2D instances. |
| 06:03 | <@Derakon> | Mind you, I'm not really using Cython to its full extent here. |
| 06:11 | <@Derakon> | Argh, I have to recompile the code after changing it! >.< |
| 06:11 | <@Consul> | Welcome to the world of C. :-P |
| 06:14 | <@Derakon> | I'm not glad to be back. |
| 06:15 | <@Derakon> | The Cython docs really, really need a simple example of a class written entirely in Cython. |
| 06:25 | | Derakon is now known as Derakon[AFK] |
| 06:26 | | Derakon[AFK] [~Derakon@Nightstar-5824.hsd1.ca.comcast.net] has quit [Quit: This computer has gone to sleep] |
| 07:32 | | Vornicus is now known as Vornicus-Latens |
| 08:15 | | AnnoDomini [AnnoDomini@Nightstar-29264.neoplus.adsl.tpnet.pl] has joined #Code |
| 08:15 | | mode/#code [+o AnnoDomini] by ChanServ |
| 08:43 | | Thaqui [~Thaqui@121.98.166.ns-22683] has quit [Client exited] |
| 10:09 | | Attilla [~The.Attil@92.1.240.ns-2748] has joined #code |
| 10:09 | | mode/#code [+o Attilla] by ChanServ |
| 10:25 | | You're now known as TheWatcher |
| 10:53 | | gnolam [lenin@Nightstar-1382.A163.priv.bahnhof.se] has joined #Code |
| 10:53 | | mode/#code [+o gnolam] by ChanServ |
| 11:06 | < Orthia> | What is Cython? |
| 11:29 | | Rhamphoryncus [~rhamph@Nightstar-7168.ed.shawcable.net] has quit [Quit: Rhamphoryncus] |
| 11:36 | | * gnolam guffaws. |
| 11:36 | <@gnolam> | http://support.microsoft.com/default.aspx?scid=kb;en-us;297830 <- "More information" |
| 11:43 | | * AnnoDomini laughs out loud. |
| 14:40 | | Explorer [Explorer@88.235.124.ns-11209] has joined #code |
| 14:40 | | TarinakyKai [~Tarinaky@Nightstar-16638.plus.com] has joined #code |
| 14:40 | | Explorer [Explorer@88.235.124.ns-11209] has left #code [] |
| 14:42 | | Tarinaky_ [~Tarinaky@Nightstar-16638.plus.com] has quit [Ping Timeout] |
| 16:08 | | Syloqs_AFH [~Syloq@Admin.Nightstar.Net] has joined #code |
| 16:09 | | Syloqs_AFH is now known as Syloqs-AFH |
| 17:04 | | Derakon [~Derakon@Nightstar-5824.hsd1.ca.comcast.net] has joined #code |
| 17:04 | | mode/#code [+o Derakon] by ChanServ |
| 17:25 | | SmithKurosaki [~Smith@Nightstar-10671.dsl.teksavvy.com] has quit [Client exited] |
| 17:27 | | SmithKurosaki [~Smith@Nightstar-10671.dsl.teksavvy.com] has joined #code |
| 17:27 | | mode/#code [+o SmithKurosaki] by ChanServ |
| 17:51 | <@ToxicFrog> | Orthia: it's a language for writing Python C extensions in. |
| 17:51 | <@ToxicFrog> | It looks similar to Python, but is statically typed and lets you call C code directly; at compile time it's translated into C. |
| 18:01 | | * gnolam stabs non-free standards with a rusty, tetanus-coated shiv. |
| 18:05 | | * gnolam also stabs free-but-impossible-to-navigate standards. |
| 18:20 | | Derakon is now known as Derakon[AFK] |
| 18:25 | | * ToxicFrog joins in |
| 18:57 | | Derakon[AFK] is now known as Derakon |
| 19:11 | <@gnolam> | The City of Boston hands out free cyber security advice: http://www.liquidmatrix.org/blog/wp-content/uploads/2009/06/screencap.png |
| 19:12 | <@gnolam> | (Tip #6) |
| 19:18 | <@ToxicFrog> | Snrk |
| 19:33 | <@SmithKurosaki> | Wow |
| 20:17 | <@Vornicus-Latens> | Der: two is enough for things with positive integer coefficients - you just have to use that fact carefully. |
| 20:25 | | * Vornicus-Latens writes a q&d polynomial evaluator in five lines: def evaluate_polynomial(coefficients, x): y = 0; for coefficient in coefficients: y = y * x + coefficient;; return y |
| 20:32 | <@Vornicus-Latens> | ...which, actually, if you recognize the inner loop there, gives it away. |
| 20:37 | | Vornicus-Latens is now known as Vornicus |
| 20:48 | | Rhamphoryncus [~rhamph@Nightstar-7168.ed.shawcable.net] has joined #code |
| 22:30 | | Vornicus [~vorn@ServicesOp.Nightstar.Net] has quit [Quit: Leaving] |
| 22:34 | | Vornicus [Vornicus@Admin.Nightstar.Net] has joined #code |
| 22:34 | | mode/#code [+o Vornicus] by ChanServ |
| 23:24 | | Reiv [~82d94c4d@Nightstar-29731.dsl.in-addr.zen.co.uk] has joined #Code |
| 23:54 | | AnnoDomini [AnnoDomini@Nightstar-29264.neoplus.adsl.tpnet.pl] has quit [Quit: ACHTUNG! ALLES LOOKENSPEEPERS! Das Internet is nicht fuer gefingerclicken und giffengrabben. Ist easy droppenpacket der routers und overloaden der backbone mit der spammen und der me-tooen. Ist nicht fuer gewerken bei das dumpkopfen. Das mausklicken sichtseeren keepen das bandwit-spewin hans in das pockets muss; relaxen und watchen das cursorblinken.] |
| --- Log closed Mon Jul 27 00:00:48 2009 |