| --- Log opened Tue May 01 00:00:38 2012 |
| 00:33 | | Derakon[AFK] is now known as Derakon |
| 00:53 | | You're now known as TheWatcher[T-2] |
| 00:57 | | You're now known as TheWatcher[zZzZ] |
| 01:22 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has quit [[NS] Quit: Z?] |
| 01:58 | | Noah [maoranma@Nightstar-93e1b6e0.pools.spcsdns.net] has quit [[NS] Quit: ] |
| 02:04 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has quit [Ping timeout: 121 seconds] |
| 03:57 | | Noah [maoranma@Nightstar-93e1b6e0.pools.spcsdns.net] has joined #code |
| 04:42 | < Noah> | Hmm, that was clever |
| 04:43 | < Noah> | Instead of storing the cards of a deck I've popped off, I instead have a method that makes two sets, one with a full normal deck of cards, and one with the current classes card list, then return the diff to know what's missing |
| 04:44 | < celticminstrel> | Huh? |
| 04:45 | < Noah> | def missing(self): |
| 04:45 | < Noah> | return list(set([rank + suit for suit in "cdhs" for rank in "A23456789TJQK"]) |
| 04:45 | < Noah> | - set(self.deck)) |
| 04:46 | < celticminstrel> | That'll return any cards not in self.deck... |
| 04:49 | < Noah> | Right |
| 04:53 | | celticminstrel is now known as celmin|zZz |
| 04:55 | | celmin|zZz [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has quit [[NS] Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.] |
| 05:03 | < Noah> | Help me out here, what's a good naming convention for methods in a class |
| 05:03 | < Noah> | I know isSomething or hasSomething are good for things that return true or false |
| 05:04 | < Noah> | I'm thinking doSomething for things that do stuff without returning anything |
| 05:05 | < Noah> | I'm getting getSomething for things that return lists and stuff? |
| 05:05 | < Noah> | guessing( |
| 05:10 | < Noah> | But it does break down when I have something that both does something and returns a value |
| 05:17 | <&Derakon> | Generally speaking, my functions that exist solely to check the instance's state are "getFoo", so e.g. "getIsPrime", "getMembers", etc. |
| 05:18 | <&Derakon> | Functions that return stuff while also doing stuff should indicate that if possible, so e.g. "generateCollisions", "makeMergedEffect". |
| 05:18 | <&Derakon> | Straight-up mutator functions can then use the remaining verbs. |
| 05:52 | | Rhamphoryncus [rhamph@Nightstar-5697f7e2.abhsia.telus.net] has quit [Client exited] |
| 06:18 | <~Vornicus> | How to tell your server overloads poorly |
| 06:18 | <~Vornicus> | You get linked on fark, and when i click that link I get an "install drupal!" page. |
| 06:18 | <~Vornicus> | I wasn't aware it was possible to configure a server so that happens. |
| 06:18 | < Noah> | lol |
| 06:18 | < Noah> | Maybe someone found a hole and broke it |
| 06:24 | | Derakon is now known as Derakon[AFK] |
| 06:30 | < Noah> | Post-modern Exceptionally Rickety Language |
| 07:07 | | ErikMesoy|sleep is now known as ErikMesoy |
| 07:22 | | Netsplit *.net <-> *.split quits: ErikMesoy, ShellNinja, McMartin, @rms, Reiver|AFK, jerith |
| 07:23 | | Netsplit over, joins: &jerith, &McMartin, Reiver|AFK, ErikMesoy, @rms, ShellNinja |
| 07:36 | | himi [fow035@D741F1.243F35.CADC30.81D435] has quit [Ping timeout: 121 seconds] |
| 07:39 | | Netsplit *.net <-> *.split quits: @McMartin, ShellNinja, ErikMesoy, Reiver|AFK, @jerith, @rms |
| 07:40 | | Netsplit over, joins: &jerith, &McMartin, Reiver|AFK, ErikMesoy, @rms, ShellNinja |
| 07:46 | | * McMartin finishes reading the OpenGL "New Testament" |
| 07:46 | <&McMartin> | That was fairly brief, actually. |
| 07:47 | <&McMartin> | There are very few places I could meaningfully plug that into Sable. |
| 07:47 | <&McMartin> | Replacing the old-style fog and texture/height color mapping, I guess, and making specular reflections be done as a fragment shader instead of as vertex lighting. |
| 07:47 | | * Noah head-bobbles cluelessly |
| 07:48 | | eckse [eckse@Nightstar-1c1741ed.dsl.sentex.ca] has quit [Client closed the connection] |
| 07:49 | <&McMartin> | "My old OpenGL game doesn't have a lot of room for improvement by shifting to programmable, modern graphics hardware over the old-school stuff, because it was designed to hew very closely to the old-school hardware limitations. I'd just be carefully simulating those." |
| 07:49 | < Noah> | I head |
| 07:49 | < Noah> | Ahem |
| 07:49 | <&McMartin> | "But I could make everything look a little more like plastic toys, and a couple of the hacks I did would no longer be necessary." |
| 07:49 | < Noah> | I heard |
| 07:49 | < Noah> | "Something something game something" |
| 07:50 | <&McMartin> | Not much good would come from rewriting the graphics engine for my game. |
| 07:51 | < Noah> | Demix it |
| 07:51 | < Noah> | Go from opengl to 8bit tiles |
| 07:51 | <&McMartin> | Heh |
| 07:51 | <&McMartin> | Not really plausible, though it would have been in days of yore. |
| 07:52 | < Noah> | I think about 90% of modern games would be better this way |
| 07:52 | <&McMartin> | That was the original plan |
| 07:52 | <&McMartin> | However, Sable - despite being named as a weak pun on "Gradius" - is way closer to After Burner |
| 07:52 | <&McMartin> | And let's be honest here, Space Harrier and After Burner itself both prove that trying to 8-bit trench-run gameplay is a bad idea. |
| 07:52 | < Noah> | Space Harrier |
| 07:52 | < Noah> | I used to have the Sega 32X version |
| 07:53 | <&McMartin> | Yeah, so |
| 07:53 | <&McMartin> | The Sega Master System version? |
| 07:53 | <&McMartin> | Not So Much (tm) |
| 07:53 | < Noah> | It was kind of lame, I mostly played Doom 32X instead |
| 07:53 | <&McMartin> | Heh |
| 07:53 | <&McMartin> | That said |
| 07:53 | <&McMartin> | Those kids appear to be mowing the lawn, so I guess I shouldn't shake my cane at them so much |
| 07:53 | <&McMartin> | A lot of the stuff that I focused on back in the day turns out to still be relevant |
| 07:54 | < Noah> | Heh |
| 07:54 | <&McMartin> | (And gets handed to the Vertex Shader as gl_ModelViewProjectionMatrix) |
| 08:01 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has joined #code |
| 08:03 | | Eri [Eri@Nightstar-3e5deec3.gv.shawcable.net] has joined #code |
| 08:05 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has quit [Ping timeout: 121 seconds] |
| 08:49 | | Netsplit *.net <-> *.split quits: @McMartin, ShellNinja, ErikMesoy, Eri, Reiver|AFK, @jerith, @rms |
| 08:50 | | Netsplit over, joins: &jerith, &McMartin, Reiver|AFK, ErikMesoy, @rms, ShellNinja, Eri |
| 09:13 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has joined #code |
| 09:31 | | You're now known as TheWatcher |
| 11:32 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 11:32 | | mode/#code [+o himi] by ChanServ |
| 13:22 | <@ToxicFrog> | McMartin: oh, god. Sabre, right? |
| 13:33 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has joined #code |
| 13:35 | | * Tamber wonders what the difference is between "x86_64" and "x86. with CONFIG_64BIT" |
| 13:35 | <@Tamber> | Aside from a few more letters~ |
| 13:37 | <@Tamber> | (Presumably x86 with CONFIG_64BIT is a 64-bit kernel, expecting the rest of the system to be 32-bit.) |
| 14:01 | <@ToxicFrog> | No. |
| 14:01 | <@ToxicFrog> | x86_64 as a kernel config option is deprecated; it (and all other 64bit options like sparc64) have been merged into a common CONFIG_64BIT flag |
| 14:02 | <@ToxicFrog> | This means that the same option will toggle 64-bit support regardless of what architecture you are building for. |
| 14:02 | <@Tamber> | Well, that makes things a lot clearer. |
| 14:04 | | * Tamber now feels rather silly. |
| 15:31 | | Rhamphoryncus [rhamph@Nightstar-5697f7e2.abhsia.telus.net] has joined #code |
| 15:44 | | celticminstrel [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has joined #code |
| 15:58 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has quit [Operation timed out] |
| 16:00 | | ToxicFrog [ToxicFrog@ServerAdministrator.Nightstar.Net] has joined #code |
| 16:00 | | mode/#code [+o ToxicFrog] by ChanServ |
| 16:50 | | Noah [maoranma@Nightstar-93e1b6e0.pools.spcsdns.net] has quit [Client closed the connection] |
| 16:50 | | Noah [maoranma@Nightstar-93e1b6e0.pools.spcsdns.net] has joined #code |
| 16:57 | | ErikMesoy is now known as Harrower |
| 17:23 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has quit [[NS] Quit: ] |
| 17:25 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has joined #code |
| 17:26 | | eckse [eckse@Nightstar-1c1741ed.dsl.sentex.ca] has joined #code |
| 17:26 | | mode/#code [+o eckse] by ChanServ |
| 17:50 | | Attilla [Obsolete@Nightstar-4d709978.as43234.net] has quit [Ping timeout: 121 seconds] |
| 18:01 | | Attilla [Obsolete@Nightstar-5589485e.as43234.net] has joined #code |
| 19:01 | < Noah> | Hmm, so, I'm thinking I should implement a players hand as an object |
| 19:02 | < Noah> | And the player too, obviously |
| 19:02 | < Noah> | s/object/class |
| 19:02 | < Noah> | Everything is an object |
| 19:07 | <&jerith> | Noah: What is it you're building? |
| 19:08 | < Noah> | Ultimately, an IRC bot that does Texas Hold'em |
| 19:08 | <&jerith> | Then you probably want something like a CardSet that represent a deck or a hand. |
| 19:08 | < Noah> | I have a deck class |
| 19:09 | <&jerith> | There are a few different ways to structure this. |
| 19:09 | < Noah> | And I can draw X cards off of it and count them as missing form the deck |
| 19:09 | < Noah> | Oh, there are tons of ways to structure everything |
| 19:09 | < Noah> | I'm trying to think more object oriented though |
| 19:09 | <&jerith> | Yes, but there are a few different /sane/ ways to do this one. ;-) |
| 19:09 | < Noah> | Yes, sane, that does narrow it down a bit |
| 19:10 | <&jerith> | One way is to have separate Deck and Hand classes, and move cards between the two. |
| 19:10 | < Noah> | that was my idea |
| 19:10 | <&jerith> | Another way is to have a single CardSet class that can represent either a hand or a dekc. |
| 19:10 | <&jerith> | *deck |
| 19:10 | < Noah> | Classes and ... contain?... other classes, right? |
| 19:11 | < Noah> | can* |
| 19:11 | <&jerith> | Another way is to have Deck class that always holds all the cards, but tracks where they are. |
| 19:11 | < Noah> | That sounds less sane |
| 19:11 | <&jerith> | Depends what you mean by "contain". |
| 19:11 | < Noah> | Say I do a hand class |
| 19:11 | <&jerith> | It has useful properties. |
| 19:11 | < Noah> | Players could have hands |
| 19:12 | <&jerith> | For example, you're guaranteed never to have duplicate or missing cards, because moving a card from the deck to a hand means changing its location attribute instead of removing it from one thing and putting it in another. |
| 19:13 | <&jerith> | An instance of an object is just a value. |
| 19:13 | <&jerith> | Err, what language is this? |
| 19:13 | < Noah> | Yea, that makes since |
| 19:13 | < Noah> | Python |
| 19:13 | <&jerith> | Ah, cool. |
| 19:14 | <&jerith> | So you can have a PlayerHand class that is given a Deck instance when it's constructed. |
| 19:14 | <&jerith> | Then hand.draw(5) could take the top 5 cards off the deck and hold onto them. |
| 19:15 | < Noah> | True, but keeping the deck the same for each player? |
| 19:15 | < Noah> | Each player having their own deck makes since for say, Magic or some other CCG |
| 19:15 | <&jerith> | Do you have your code on the webnets anywhere? |
| 19:15 | < Noah> | Webnets? No? |
| 19:16 | < Noah> | Lemme get a paste bin up |
| 19:16 | <&jerith> | Something like deck = Deck(); hand1 = PlayerHand(deck); hand2 = PlayerHand(deck) |
| 19:16 | <&jerith> | That way both hands share the same deck. |
| 19:17 | < Noah> | Hmm, true |
| 19:17 | <&jerith> | But it would probably make more sense to have the deck separate and say hand.draw(deck, 5) or hand.add(deck.draw(5)) or something. |
| 19:19 | < Noah> | Yea, I was thinking the latter |
| 19:19 | <&jerith> | Now, since this is Python, you /could/ just make each hand a set(). |
| 19:20 | < Noah> | http://pastebin.starforge.co.uk/516 |
| 19:20 | <&jerith> | But you probably want a class that has scoring methods on it. |
| 19:20 | < Noah> | Yea |
| 19:21 | <&jerith> | Which is a good reason not to use the single PileOfCards class for both. |
| 19:21 | <&jerith> | (Although is there's enough duplicate functionality, you might want a common base class.) |
| 19:22 | < Noah> | I figured it would be a method of the gameplay class say, gameplay.checkWinningHand(listofplayersandtheirhandsinasaneway) |
| 19:23 | <&jerith> | What revision control system are you using? |
| 19:23 | < Noah> | Uh...my brain |
| 19:23 | < Noah> | I really should upgrade to something better |
| 19:23 | <@ToxicFrog> | :gonk: |
| 19:24 | <@ToxicFrog> | Go. Now. Install mercurial or git. |
| 19:24 | < Noah> | Anything that I should consider, using PyDev? |
| 19:25 | <&jerith> | I don't know PyDev. |
| 19:25 | < Noah> | I heard that in Lilu's voice from 5th Element |
| 19:27 | < Rhamphoryncus> | argh. I know there's at least vertex arrays, display lists, and VBOs, as well as ARB and non-ARB functions, and that at least some of VBO is depreciated |
| 19:27 | <&jerith> | Oh, it's an Eclipse thing. |
| 19:27 | | * jerith dislikes Eclipse. |
| 19:28 | < Rhamphoryncus> | I have no idea what I should be using. Some indication that vertex arrays and display lists are old/depreciated/discouraged too.. |
| 19:28 | <&jerith> | Anyways, I'd go with git or hg. The Eclipse bzr plugin makes me sad. |
| 19:28 | <&jerith> | But get one and install it now. |
| 19:29 | < Noah> | On it |
| 19:29 | < Noah> | Well |
| 19:29 | < Noah> | I would be |
| 19:29 | < Noah> | But hg or git... |
| 19:29 | < Noah> | Why is there never an obvious choice T_T |
| 19:29 | <&jerith> | Once that's done, I'll show you how you should structure your project into lib, script and test bits. :-P |
| 19:29 | <&jerith> | If you use git, you can use github and make it easier for me to look at your code. |
| 19:30 | < Noah> | that's obvious enough for me |
| 19:30 | <&jerith> | (If you use hg, you can use bitbucket and make it slightly less easier for me to look at your code.) |
| 19:30 | <&McMartin> | That reminds me |
| 19:30 | <&McMartin> | Now is not a good time |
| 19:30 | < Noah> | Shh, you're complicating it |
| 19:31 | <&McMartin> | But jerith, I want to pick your brain later re: the current state of what used to be distutils (and may still be). |
| 19:31 | <&jerith> | McMartin: The quick answer is "horrible". |
| 19:31 | <@ToxicFrog> | Noah: it looks like you're on windows, so you may prefer hg, which has better windows integration |
| 19:31 | <&jerith> | The long answer is that you can do a bunch of nifty things, but you sometimes have to be careful. |
| 19:32 | < Noah> | Uhg |
| 19:32 | < Noah> | Damnit you two, git or hg!? |
| 19:32 | <&jerith> | Oh, Windows. |
| 19:33 | < Noah> | This is why I've never picked one yet, lol |
| 19:34 | <&jerith> | http://help.github.com/win-set-up-git/ |
| 19:34 | <&jerith> | No idea how useful that is. |
| 19:35 | <&jerith> | But if you live in Eclipse, all you really care about is Eclipse integration and the existence of a functioning implementation. |
| 19:35 | < Noah> | I don't |
| 19:35 | <@ToxicFrog> | Noah: like I said, if you're on windows, you probably want hg, unless you have a pre-existing preference for github or bitbucket as a hosting site |
| 19:36 | < Noah> | The hell |
| 19:36 | <@ToxicFrog> | Since hg has an actual windows port, whereas git has a windows build of the linux tools (which is to say it is heavily command line driven) |
| 19:36 | < Noah> | hg downloaded the 64bit version |
| 19:37 | < Noah> | I'm not too bothered by command line |
| 19:38 | <&jerith> | Good point. hg it is. |
| 19:38 | < Noah> | Okay, so we've come to a concensus? Hg is it? |
| 19:38 | <&jerith> | Yes. |
| 19:39 | < Noah> | because if I download this, and McMartin says some shit about git, all of you are flamable |
| 19:39 | <&jerith> | McMartin: Say some shit about git. :-P |
| 19:39 | <@ToxicFrog> | (I am a hardcore git user, but I also do all of my development on linux, and my attempts to teach my advisor how to use it on windows have been fraught with suffering) |
| 19:39 | <&jerith> | Noah: It really doesn't matter all that much in the grand scheme of things. |
| 19:39 | <&McMartin> | Yeah, that |
| 19:39 | <@ToxicFrog> | (granted this is about 50/50 "git behaving oddly on windows" vs "he is damaged from years of svn"~) |
| 19:39 | <&McMartin> | git on Windows still involves installing MSYS and opening a Bash shell. |
| 19:40 | <&jerith> | I like git mostly because (a) it's what I'm more familiar with and (b) github. |
| 19:40 | <@ToxicFrog> | And yeah, if it really becomes an issue at some point I'm pretty sure git and hg can import/export from each other anyways, so |
| 19:41 | <&jerith> | I have some specific reasons to dislike hg, but there are at least as many comparable (but different) quirks in git that I don't dislike as much simply because I'm used to them. |
| 19:41 | < celticminstrel> | Is bitbucket not as good as github, then? |
| 19:41 | <&jerith> | celticminstrel: I don't think so. |
| 19:41 | <@ToxicFrog> | McMartin: in its defence, this worked flawlessly for me, also, command line supremacy |
| 19:41 | <&jerith> | But then I don't really use bitbucket much. |
| 19:41 | < celticminstrel> | Though, there are other hosting options for hg. |
| 19:41 | <@ToxicFrog> | On the flip side it didn't work for my advisor (none of the context menu entries are showing up) and neither of can figure out why >.< |
| 19:42 | <@ToxicFrog> | (on the gripping hand he is running a decade-old OS so who the fuck knows) |
| 19:42 | <&McMartin> | TF: I have MSYS installed three times on my system, and I think they're all different versions, it's beginning to make me wish they wouldn't be so allergic to just linking against Win32 =P |
| 19:42 | < celticminstrel> | Which OS is that? |
| 19:42 | <@ToxicFrog> | XP. |
| 19:42 | < Noah> | *headdesks* I just used ls in windows command prompt |
| 19:42 | < celticminstrel> | I have an XP too. |
| 19:42 | <@ToxicFrog> | McMartin: er, msys tools do link against the win32 libraries |
| 19:43 | <&McMartin> | git is the only one with an excuse here, being written in Shell |
| 19:43 | <@ToxicFrog> | The reason git4win uses it is that git is heavily dependent on bash and perl and a decent terminal emulator, and MSYS is the most convenient way to get all three at once on windows. |
| 19:43 | < Noah> | Okay, hg installed, teach me oh sensei |
| 19:43 | | * jerith receives a failing test from his boss, begins to do the big chunk of work required to make it pass. |
| 19:44 | <@ToxicFrog> | epsilon holy shit, do not test gravity with glass tableware, demon cat |
| 19:44 | <&jerith> | Noah: Install whatever Eclipse hg magic you want. |
| 19:44 | <@ToxicFrog> | ...he's using Eclipse? |
| 19:44 | <&jerith> | ToxicFrog: This is the one who has been installing updates and deleting your desktop, right? |
| 19:44 | <&jerith> | ToxicFrog: PyDev, which is Eclipse. |
| 19:44 | <@ToxicFrog> | jerith: actually no, that was Suzie |
| 19:45 | <@ToxicFrog> | Epsilon is less about lying on the keyboard and more about chewing on the monitor. |
| 19:45 | < Noah> | yes, PyDev |
| 19:45 | <&jerith> | I decided against introducing him to Emacs at this point. :-P |
| 19:45 | < Noah> | I've met Emacs, I found him rather obtuse |
| 19:45 | | Harrower is now known as ErikMesoy |
| 19:45 | <@ToxicFrog> | Neither emacs nor vi has ever clicked for me. |
| 19:45 | < Noah> | I like vim though, but setting up vim crap in windows is some kind of nightmare |
| 19:46 | <&jerith> | Noah: I never found it to be so, but then it's been about a decade since I last tried. |
| 19:46 | < Noah> | Anyway, PyDev is up and running and it seems to work well enough |
| 19:47 | < Noah> | So, apparently, I'm doing something with Eclipse for hg |
| 19:47 | <@ToxicFrog> | Emacs kind of strikes me as the Elder Scrolls of editors, which is to say, it's terrible out of the box but becomes excellent after weeks of modding and fine-tuning |
| 19:47 | <&jerith> | Do you have a suitable hg plugin? |
| 19:47 | < Noah> | Not yet |
| 19:47 | < Noah> | http://javaforge.com/project/HGE |
| 19:47 | < Noah> | I'm seeing this |
| 19:47 | <&jerith> | ToxicFrog: Emacs is less an editor than a toolkit with which to build an editor. |
| 19:47 | <@ToxicFrog> | Honestly I'd just use the hg interface outside of Eclipse; I've never used an IDE-integrated versioning system that I liked |
| 19:47 | <&jerith> | Noah: I can't really help you there, I'm afraid. |
| 19:48 | < Noah> | Emacs is less an editor and more of an operating system, complete with preinstalled games |
| 19:48 | <&jerith> | But using hg outside Eclipse is doable. |
| 19:48 | <@ToxicFrog> | Then again, I haven't used Eclipse plus any of them because I dislike Eclipse to begin with~ |
| 19:48 | <&jerith> | Noah: It doesn't technically meet the requirements of an OS, since it can't run on bare metal.~ |
| 19:49 | <&jerith> | Eclipse has some pretty decent VCS integration. |
| 19:49 | < Noah> | Win9x needed DOS, so did it not count as an OS? |
| 19:49 | <&jerith> | Noah: DOS counts as "dare metal".~ |
| 19:49 | <&jerith> | *bare |
| 19:50 | < Noah> | okay, so emacs needs linux, otherwise known as the emacs-loading-station for you guys |
| 19:50 | <&jerith> | Although when I last seriously used it (about 5 years ago), the VCS integration was biased toward svn-alikes. |
| 19:51 | <&jerith> | Noah: Maybe you should set up a bitbucket repo now. |
| 19:51 | < Noah> | Okay, and I just do that at bitbuckets site? |
| 19:51 | <&jerith> | Yup. |
| 19:51 | <&jerith> | There should be instructions. |
| 19:52 | < Noah> | On it |
| 19:53 | < Noah> | Great, ToxicFrog, you made my Eclipse angry |
| 19:53 | < Noah> | Woah |
| 19:54 | < Noah> | I...apparently...already have a bitbucket? |
| 19:54 | <&jerith> | Noah: Can you log in? |
| 19:54 | < Noah> | About to find out |
| 19:55 | <&jerith> | It's possible someone else has your username. |
| 19:55 | < Noah> | And email? Unlikely |
| 19:55 | <&jerith> | (Unless it uses an email address. I forget.) |
| 19:55 | < Noah> | I think I just made one and forgot all about it |
| 19:55 | <&jerith> | Possible. |
| 19:56 | <&jerith> | Maybe last time we had this conversation. ;-) |
| 19:56 | < Noah> | no, I think it had something to do with minecraft, but I'll never remember why |
| 19:56 | < Noah> | Okay, in bitbucket, looking into ide integration with eclipse |
| 19:59 | < Noah> | I do like how eclipse handles plugin stuff |
| 20:00 | < Noah> | What the fuck am I watching? Some assholes are looking for Hilter's ghost |
| 20:00 | <&jerith> | Noah: ... |
| 20:01 | <&jerith> | I find the plugin management to be the least likeable bit of Eclipse. |
| 20:01 | < Noah> | They're taunting him too "You're a war criminal. You know what you did, it's in the history books" |
| 20:01 | <&jerith> | (For starters, it's bloody impossible to figure out how to update or uninstall anything.) |
| 20:02 | | * Noah points at the update button? |
| 20:02 | <&jerith> | It may have improved in the past few years. |
| 20:02 | <&jerith> | I know they were madly working on it. |
| 20:03 | < Noah> | Probably for people like you, you ungrateful heathen |
| 20:03 | <&jerith> | (I know this because they apparently reimplemented the entire frontend with every release.) |
| 20:03 | <&jerith> | Anyways, enough rant. :-) |
| 20:11 | <&jerith> | My boss' 8 week old daughter keeps finding bugs in our software. |
| 20:11 | < Noah> | HgE installing |
| 20:11 | <&McMartin> | jerith: \o/ |
| 20:11 | < Noah> | haha |
| 20:11 | <&jerith> | Her name is "Zo?". |
| 20:11 | < Noah> | Oh crap! Gotta change the channel, hot english bitch is cooking and I don't want to drown in saliva |
| 20:12 | <&jerith> | And he uses that as test input a lot. |
| 20:12 | < Noah> | How do you pronounce Zo... .e.? e:? |
| 20:13 | < ErikMesoy> | I would figure that it's "Zo-e" and the dots are to make sure "oe" doesn't become a single sound. |
| 20:13 | <&McMartin> | Yeah |
| 20:13 | <&McMartin> | Rhymes with "glowy" |
| 20:13 | < Noah> | Ah, Zoey |
| 20:13 | <&McMartin> | Dieresis! |
| 20:13 | <&McMartin> | \m/ |
| 20:14 | <&jerith> | Same as https://en.wikipedia.org/wiki/List_of_Firefly_characters#Zoe_Washburne |
| 20:15 | <&jerith> | McMartin: His second son also has r?ckd?ts. |
| 20:16 | <&jerith> | The eldest shares a more normal name with someone in-channel. |
| 20:16 | < Noah> | Oh! is it me? |
| 20:16 | < Noah> | Is it Noah?! |
| 20:16 | <&jerith> | Noah: Well, you're a bit older than any of his kids. ;-) |
| 20:17 | < Noah> | I meant his name |
| 20:18 | <&jerith> | I'll concede that it is a biblical name and leave it at that. |
| 20:18 | <&jerith> | I've probably already talked about his kids more than I really should. |
| 20:19 | < Rhamphoryncus> | Clearly named Rhamphorhynchus |
| 20:19 | <&jerith> | Rhamphoryncus: You do actually have a biblical name. |
| 20:19 | < Rhamphoryncus> | I know :P |
| 20:21 | < Rhamphoryncus> | but Rhamphorhynchus would be cooler |
| 20:21 | <&jerith> | Noah: Are you all set up with your code in bitbucket yet? |
| 20:22 | < Rhamphoryncus> | although Tyrannus would be more mainstream |
| 20:22 | < Noah> | No |
| 20:23 | < Noah> | Figuring out how to get my exsisting code into a repo through HgE |
| 20:23 | <&jerith> | When a Mister Brown I worked with had a baby on the way, we all tried to convince him that the best name for her would be "Octothorpe". |
| 20:24 | <&jerith> | Noah: "hg init" in the working directory. |
| 20:24 | <&jerith> | Then tell Eclipse that you have an existing repo. |
| 20:24 | < Noah> | Yea, guess I'll do that instead |
| 20:24 | <&jerith> | Or ignore Eclipse for now and do whatever bitbucket says you should do to get the project up there. |
| 20:25 | <&jerith> | Then point Eclipse at your repo. |
| 20:25 | <&jerith> | Or point it at bitbucket, if that looks like an easier path to victory. |
| 20:30 | < Noah> | Hmm, tried pushing and it says no changes |
| 20:31 | <&jerith> | What's the bitbucket URL? |
| 20:31 | < Noah> | https://bitbucket.org/mao42ranma/pyholdem-irc/ |
| 20:32 | | * Rhamphoryncus thwacks whoever did the naming in opengl |
| 20:32 | <&jerith> | Noah: Would you like to either make it public or add me to it? |
| 20:33 | < Noah> | And public |
| 20:34 | <&jerith> | No commits. |
| 20:34 | <&jerith> | Have you set up the link to the remote repo properly? |
| 20:35 | | * jerith doesn't remember how to do that in hg. |
| 20:38 | <&McMartin> | Rham: VBOs, it turns out, are Old Testament, but I think they have some use still |
| 20:38 | <&McMartin> | It's glDrawElements that's deprecated, if I'm reading it right |
| 20:38 | < Rhamphoryncus> | McMartin: it seems like they're still the preferred way, they've just been gutted |
| 20:41 | | * McMartin nods |
| 20:41 | <&McMartin> | Yeah, so |
| 20:41 | <&McMartin> | Not counting the introductions |
| 20:41 | <&McMartin> | The Old Testament is 14 chapters long |
| 20:41 | <&McMartin> | The New Testament is... 3 |
| 20:42 | <&jerith> | Are we talking about OpenGL here still? |
| 20:42 | <&McMartin> | But a whole bunch of stuff (notably: transformation matrices and all that surround them) are shared |
| 20:42 | < Rhamphoryncus> | jerith: yes |
| 20:42 | <&McMartin> | Yeah |
| 20:43 | < Noah> | Sucess |
| 20:43 | <&McMartin> | Also, much belated: |
| 20:43 | <&McMartin> | TF: Yes, by analogy with "Gladius" |
| 20:44 | <&jerith> | Noah: Cool. Now for the next bit. |
| 20:44 | <&jerith> | Create a directory called "pyholdem", move "deck.py" into it, create an empty "__init__.py" in it. |
| 20:45 | < Noah> | okay, sec |
| 20:45 | <&jerith> | You now have a package called pyholdem, which contains a module called deck. |
| 20:45 | <&jerith> | (A package is just a module made out of a directory containing other modules.) |
| 20:49 | < Rhamphoryncus> | If I understand correctly a vertex array object is.. not a vertex array. It is a group of buffers (often but not always used to store vertexes), a parent buffer (often (always?) used to index the child buffers), and a bunch of associated attributes |
| 20:50 | < Rhamphoryncus> | it should really be called VertexBufferObjectArrayObject :P |
| 20:50 | < Rhamphoryncus> | +ObjectObject |
| 20:51 | <&jerith> | Rhamphoryncus: +ProxyFactoryFactory? |
| 20:51 | < Rhamphoryncus> | there ya go! :D |
| 20:52 | <&jerith> | And now we're in Java! \o/ |
| 20:52 | <&McMartin> | Yeah, it does seem things are generalized |
| 20:52 | <&McMartin> | I haven't worked out how this translates to shaders, and if you can keep some stuff persistent between runs |
| 20:53 | | * McMartin is *particularly* wondering if you can generate an infinite-extent plasma fractal in a shader and use it as a height map |
| 20:53 | < Rhamphoryncus> | jerith: well these are all handles. As integers even |
| 20:53 | < Rhamphoryncus> | Except for ones stored as void *.. which are also just integers now |
| 20:53 | <&McMartin> | Woo, C interfaces |
| 20:54 | < Rhamphoryncus> | Yeah, this pretty much defines a "C interface" |
| 20:55 | < Noah> | Add, commit, push |
| 20:55 | < Noah> | This is complicated |
| 20:55 | < Noah> | When do I code? |
| 20:56 | < Noah> | Anyway, folder and __init__ made |
| 20:56 | <@ToxicFrog> | Noah: now that it's all set up, you can just get into an add/commit cycle and that's all you need to do |
| 20:57 | <&jerith> | Noah: Sorry, juggling work work at the moment. |
| 20:57 | <&jerith> | Be with you again in a bit. |
| 20:57 | < Noah> | that's fine |
| 20:58 | < Noah> | between two kids while doing this anyway |
| 20:59 | < Noah> | All the smallest isn't much of an issue, she's happily gumming on her plush hippo |
| 21:00 | <&McMartin> | Yay |
| 21:00 | < Noah> | The boy is a brat though |
| 21:00 | | * McMartin would suggest lego jet fighters but the little one is probably little enough that this would be a hazard |
| 21:00 | < Noah> | Yea, I try not giving her anything that isn't at least as big as her head |
| 21:00 | < Noah> | Because she likes to TRY to eat everything |
| 21:01 | < Noah> | Elizabeth and Watts btw |
| 21:01 | < Noah> | But I hate that name, so I call her "Z" |
| 21:04 | <&McMartin> | Heh |
| 21:04 | <&McMartin> | Yeah, though, I've been breaking out LEGO again for a silly wargame coming out this month |
| 21:05 | <&McMartin> | THe new kits are actually really quite slick while not being 'this piece is 50% of the final model' |
| 21:05 | <&McMartin> | https://hkn.eecs.berkeley.edu/~mcmartin/01/jet_front.JPG |
| 21:07 | < Noah> | What game? |
| 21:08 | <&McMartin> | "Mobile Frame Zero: Rapid Attack" |
| 21:08 | < Noah> | I've never heard of that |
| 21:08 | <&McMartin> | As noted, it hasn't actually come out yet |
| 21:08 | < Noah> | Oh |
| 21:08 | < Noah> | Cool jet anyway |
| 21:09 | <&McMartin> | Way bigger than stuff I'll need to design for the game (the resulting robots need to be only about the size of normal wargame minis) |
| 21:09 | <&jerith> | Noah: So, you now have your code in a package. |
| 21:10 | < Noah> | So it's wargaming, but is uses legos? that's ingenous |
| 21:10 | <&jerith> | There are two things you can do, and you should probably do both of them. |
| 21:10 | <@ToxicFrog> | McMartin: this has been my main complaint about recent lego kits - so many of the pieces are special forms useful only in that model. |
| 21:14 | <&jerith> | Noah: First, create run.py in your project root (next to pyholdem) and put your whole main() and if __name__ thing in there. |
| 21:14 | <&jerith> | At the top, "from pyholdem.deck import DeckOfCards". |
| 21:15 | <&jerith> | Also, rename your class to DeckOfCards. The initial capital is convention, but it's a very strong convention. |
| 21:15 | < Noah> | Hm, okay |
| 21:16 | <&jerith> | Then "python run.py" should do what "python deck.py" did before, only with neater code organisation. |
| 21:16 | < Rhamphoryncus> | Define array as GLshort, tell opengl it's GL_UNSIGNED_SHORT. *twitch* |
| 21:20 | < Noah> | jerith: sure enough |
| 21:22 | < Noah> | Add, commit, push |
| 21:22 | < Noah> | Sounds like something I'm going to be saying a lot |
| 21:27 | < Noah> | McMartin: This game sounds pretty fun, I want to play! :P |
| 21:34 | | Vash [Vash@Nightstar-fb5b40e5.wlfrct.sbcglobal.net] has joined #code |
| 21:34 | | mode/#code [+o Vash] by ChanServ |
| 21:37 | < Rhamphoryncus> | yay, indexed drawing of a triangle |
| 21:37 | <&McMartin> | ToxicFrog: I'm having good luck with kits in the "Creator" line. It's got slick results, but the subcomponents are pretty generic. |
| 21:37 | <&McMartin> | The most specialized piece in that jet model was the 1x10 curved slope piece |
| 21:37 | <&McMartin> | And, well, "1x10 curved slope piece" is still pretty generic. |
| 21:40 | <&jerith> | Noah: Go look up unit testing in Python while I'm busy. :-) |
| 21:41 | < Rhamphoryncus> | Two! Two triangles! |
| 21:41 | < Rhamphoryncus> | Using shared vertexes, so I could now expand on that |
| 21:48 | | * Noah googles unit testing in python~ |
| 21:54 | | ErikMesoy is now known as ErikMesoy|sleep |
| 22:00 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 22:15 | < Noah> | mmm |
| 22:15 | < Noah> | I don't get it |
| 22:22 | <&jerith> | Hmm? |
| 22:24 | < Noah> | Unit testing |
| 22:25 | <@TheWatcher> | What don't you get about it? |
| 22:25 | < Noah> | It |
| 22:25 | <@TheWatcher> | Why it's needed, how it works...? |
| 22:26 | < Noah> | yes |
| 22:26 | <@TheWatcher> | Because the former is easy: if you don't have unit tests, you have no way to ensure that your code behaves as it should in all situations. |
| 22:29 | <&jerith> | Noah: Have you read http://docs.python.org/library/unittest.html#basic-example ? |
| 22:29 | < Noah> | Okay, but creating test code for every little situation seems, tedious? |
| 22:30 | <@TheWatcher> | Yes, and? |
| 22:30 | <@TheWatcher> | You're welcome to break out VDM++ instead~ |
| 22:31 | <&jerith> | Noah: Less tedious than manually testing ALL THE THINGS. |
| 22:33 | < Noah> | Perhaps |
| 22:33 | <&jerith> | Once you write a test asserting that deck.draw(5) returns five cards and that those five cards are no longer in the deck, you never have to worry about breaking that again. |
| 22:33 | <&jerith> | If you tweak something that results in the cards not actually being removed or whatever, the test will fail and let you know. |
| 22:34 | < Noah> | I see |
| 22:34 | < Noah> | Do I just make a test package? or just a test file? |
| 22:34 | <&jerith> | That way you don't spend the next hour trying to figure out why there are five aces floating around later. |
| 22:35 | < Noah> | I don't just ban the guy with 5 aces? |
| 22:35 | <&jerith> | I generally have foo/tests/test_blah.py for each foo/blah.py |
| 22:35 | <&jerith> | I also use a smarter test runner than the default one. |
| 22:36 | <&jerith> | Like http://pypi.python.org/pypi/nose/1.1.2 |
| 22:39 | <&jerith> | Noah: VCS gives you the freedom to change your code without worrying that you'll break it and not be able to get back to a working state. |
| 22:40 | <&jerith> | Noah: Unit tests give you the freedom to change your code without worrying that you'll break it and not know that you've broken it. |
| 22:40 | < Noah> | Hmm, well, I suppose I'll have to try it |
| 22:40 | <&jerith> | It's more complicated than that for both, but those two things are at the core. |
| 22:41 | <&jerith> | Also, there's a reason I'm telling you to test stuff now rather than later. |
| 22:41 | <&jerith> | It's *much* easier to add tests when you're writing the code than it is to go back and add them later. |
| 22:42 | <&jerith> | A good set of tests influences your architecture and design. You end up with stuff that's more modular, because it's easier to write tests around that stuff. |
| 22:44 | < Eri> | But, how do you know what makes a good test? |
| 22:44 | < Eri> | Like, lets take something like the quadratic equation as an example |
| 22:45 | <&jerith> | Eri: How do you know what makes good code? |
| 22:45 | < Eri> | Say we were solving for x. There's four cases. Linear, two identical real roots, two real roots, and complex roots |
| 22:45 | <&jerith> | A test is just a piece of code. |
| 22:45 | < Eri> | Experience, I guess |
| 22:45 | <&jerith> | You generally want more small tests rather than fewer big ones. |
| 22:45 | < Noah> | Well, PyDev gives me a unittest template |
| 22:45 | <&jerith> | But it depends. |
| 22:46 | < Eri> | Like, would it be sufficient to throw in a test for each case, and maybe one on the boundaries? |
| 22:46 | <&jerith> | Noah: Actually, PyDev might have a thing that will run them for you. If so, you might not need nose or whatever. |
| 22:46 | <&jerith> | Eri: Maybe. It depends. |
| 22:47 | | * TheWatcher does admit to being pretty bad with unit tests |
| 22:47 | <&jerith> | I generally start with a few simple tests, including whatever pathological cases I can come up with. |
| 22:47 | <@TheWatcher> | (then, 90% of my code also happens to need databases, CGI, sessions and stuff, so it gets really nasty to set up really quick ;.;) |
| 22:48 | < Noah> | So, should I do all my tests in one file? |
| 22:48 | <&jerith> | Almost invariably, I find new edge cases (and often bugs) that aren't covered by my tests when I use my code in the next bit. |
| 22:48 | <&jerith> | So then I go back and add more tests for that stuff. |
| 22:49 | <&jerith> | Noah: I tend to have a test module for each non-test module. |
| 22:49 | <&jerith> | But different people do it differently. |
| 22:51 | <&jerith> | Noah: Writing tests can be a bit painful, but it's invariably worth it. |
| 23:20 | < Noah> | Kay, made a unit test and pushed it |
| 23:20 | < Noah> | Also, what's the point of that __init__.py file? |
| 23:34 | < Noah> | Ah, Adventure Time, so silly |
| 23:37 | <&McMartin> | Noah: __init__.py is a signal to python that this is a package, basically |
| 23:37 | <&McMartin> | When the whole package is imported, it runs any code in there as a global initializer, more or less. |
| 23:37 | <&McMartin> | You usually don't need that, so it's usually empty |
| 23:38 | < Noah> | Ah |
| 23:40 | <&jerith> | Actually, __init__.py is the content of the module. |
| 23:41 | <&jerith> | If you "import foo" where foo is a package, you get the contents of foo/__init__.py. |
| 23:42 | | himi [fow035@D741F1.243F35.CADC30.81D435] has joined #code |
| 23:42 | | mode/#code [+o himi] by ChanServ |
| 23:42 | <&jerith> | You can also "import foo.bar", which imports foo/__init__.py and then imports foo/bar.py. |
| 23:43 | <~Vornicus> | note that generally it doesn't add the non-bar components of foo to the namespace |
| 23:44 | <&jerith> | It doesn't add them to your namespace, but it does add them to sys.modules. |
| 23:44 | <&jerith> | (But you don't have to worry about sys.modules unless you're doing really horrible import mangling, in which case you'd already know all about it.) |
| 23:44 | < gnolam> | One hard drive down, two to go... |
| 23:45 | <~Vornicus> | At that point though, import foo is pretty much free |
| 23:45 | <&jerith> | Yes, because it's just a sys.modules lookup under the hood. |
| 23:46 | <&jerith> | Also, mucking about with sys.path can lead to some excitement. |
| 23:46 | <~Vornicus> | EXCITEMENT |
| 23:46 | <~Vornicus> | but what about excitebike? can it lead to that? |
| 23:46 | < gnolam> | Man. I should've bought one of these external HDD enclosures years ago. |
| 23:46 | <&jerith> | I once had 'foo' and 'foo/bar' in there for reasons that I won't go into now, but which were even horribler than you might imagine. |
| 23:47 | <~Vornicus> | That's horrible |
| 23:47 | <&jerith> | Because of this, I managed to import foo/bar/baz.py as both foo.bar.baz and bar.baz. |
| 23:48 | <&jerith> | foo.bar.baz.SomeClass and bar.baz.SomeClass are different things, despite both having come from the same source file. |
| 23:49 | <&jerith> | Which led to lots of exciting debugging. |
| 23:50 | <&jerith> | (This whole thing was Django's fault, by the way. With its yucky path mangling (fixed in 1.4, thank Eris) and some implicit-relative imports in code I'd inherited...) |
| 23:55 | < Noah> | Heh |
| 23:56 | | Vash [Vash@Nightstar-fb5b40e5.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| --- Log closed Wed May 02 00:00:19 2012 |