| --- Log opened Thu Oct 06 00:00:07 2011 |
| 00:17 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has quit [Ping timeout: 121 seconds] |
| 00:26 | | RichardBarrell [mycatverbs@Nightstar-3b2c2db2.bethere.co.uk] has quit [Client closed the connection] |
| 00:43 | | You're now known as TheWatcher[T-2] |
| 00:45 | | You're now known as TheWatcher[zZZZ] |
| 00:45 | | You're now known as TheWatcher[zZzZ] |
| 00:55 | | Derakon[AFK] is now known as Derakon |
| 01:07 | | Vash [Vash@Nightstar-f03c5637.sd.cox.net] has joined #code |
| 01:25 | | Vash [Vash@Nightstar-f03c5637.sd.cox.net] has quit [Ping timeout: 121 seconds] |
| 01:29 | | Vash [Vash@Nightstar-f03c5637.sd.cox.net] has joined #code |
| 01:40 | | Vash [Vash@Nightstar-f03c5637.sd.cox.net] has quit [[NS] Quit: I <3Lovecraft<3 Vorn!] |
| 02:22 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has quit [[NS] Quit: Z?] |
| 02:32 | | Kindamoody[zZz] is now known as Kindamoody |
| 04:18 | < McMartin> | Ugh. So, I've been having to work with COM and GObject lately, both of which are terrible. |
| 04:18 | < McMartin> | Very common, and kind of annoying, is the IUnknown::QueryInterface() method, which COM uses to do type interrogation and shift interfaces around, since it's both reflective and has strong typing |
| 04:19 | < McMartin> | It also doesn't believe in return values, which makes it very inconvenient to use, so I'm making some wrapper methods to make life easier |
| 04:19 | < McMartin> | It is requiring every fiber of my will to not call the QueryInterface wrapper "COMcast" |
| 04:56 | < Vornicus> | D: |
| 05:37 | | kwsn is now known as kw-sleep-n |
| 05:48 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has joined #code |
| 06:20 | | celticminstrel [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 06:54 | | Derakon is now known as Derakon[AFK] |
| 07:04 | | Kindamoody is now known as Kindamoody|afk |
| 09:19 | | Attilla [Some.Dude@Nightstar-f29f718d.cable.virginmedia.com] has joined #code |
| 09:28 | | You're now known as TheWatcher |
| 09:39 | < jerith> | http://thedailywtf.com/Articles/Caught.aspx |
| 09:44 | < McMartin> | I disbelieve~ |
| 09:45 | < jerith> | I can see that being carefully calculated and then all the parameters hardcoded or something. |
| 10:37 | | Rhamphoryncus [rhamph@Nightstar-14eb6405.abhsia.telus.net] has quit [Client exited] |
| 10:48 | | You're now known as TheWatcher[afk] |
| 11:29 | | kw-sleep-n is now known as kwsn |
| 11:56 | | Kindamoody|afk is now known as Kindamoody |
| 13:27 | | Kindamoody is now known as Kindamoody|out |
| 13:29 | | gnolam [lenin@Nightstar-202a5047.priv.bahnhof.se] has joined #code |
| 13:32 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has quit [Connection closed] |
| 13:35 | | celticminstrel [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has joined #code |
| 14:03 | < gnolam> | So many Steve Jobs jokes, so little time... |
| 14:10 | | teuanjo [teuanjo@Nightstar-5b7a5549.rev.vodafone.pt] has joined #code |
| 14:11 | | teuanjo [teuanjo@Nightstar-5b7a5549.rev.vodafone.pt] has quit [[NS] Quit: IRC webchat at http://irc2go.com/] |
| 14:15 | < gnolam> | "Do not joke about Steve Jobs's death, it's not PC." |
| 15:01 | < jerith> | Nice. |
| 15:02 | < kwsn> | well... PC did do him in >_> |
| 15:03 | < kwsn> | (Pancritic cancer) >_> |
| 15:09 | | AnnoDomini [annodomini@60F158.737D66.CA1918.CC7562] has joined #code |
| 15:11 | < gnolam> | ... cancer that transcends criticism? |
| 15:15 | < Lingerance> | So hipster cancer? |
| 15:18 | < AnnoDomini> | I have nothing against giving hipsters cancer. |
| 15:19 | < kwsn> | hehe |
| 15:38 | | * gnolam stabs Python and Python installers. |
| 15:41 | < kwsn> | :P |
| 15:46 | < gnolam> | They fail at supporting multiple Python installations. :P |
| 15:48 | | celticminstrel [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 16:10 | | * AnnoDomini hates on websites which require absurdly diverse passwords. |
| 16:11 | | * ToxicFrog hates on websites that limit passwords to 8 characters and reject email addresses with + in them |
| 16:11 | | * Tamber hates on websites that 'validate' anything to the left of the @ |
| 16:17 | | AnnoDomini [annodomini@60F158.737D66.CA1918.CC7562] has quit [[NS] Quit: leaving] |
| 17:21 | < gnolam> | At least it's not mandating absurdly diverse /and less secure/ passwords *cough*Handelsbanken*cough*. |
| 17:23 | < gnolam> | Specifically: a password, 8-12 characters long, with at least one lowercase letter, one uppercase letter and at least two digits. |
| 17:23 | < gnolam> | >_< |
| 17:24 | < Attilla> | but digits are less secure than letters :O |
| 17:41 | < gnolam> | Mandating "must contain digits" is ok (/sort of/) since it increases the search space per character. But someone must've thought "hey, if mandating one makes it more secure... then mandating /two/ makes it twice as secure!". |
| 17:58 | < Fantastic_Phox> | There should be no mandation at all. Strong recommendations, yes |
| 18:04 | < gnolam> | Thus the "sort of". |
| 18:08 | | * gnolam glaraghaeuoagls. |
| 18:14 | < gnolam> | It shouldn't be this hard to distribute a goddamned program. |
| 18:15 | < gnolam> | FOR FUCK'S SAKE |
| 18:15 | < gnolam> | Fuck it, looks like I'm going to have to ditch Python altogether and rewrite the entire goddamn thing in C++. |
| 18:16 | < gnolam> | N GANGEAHGHGE |
| 18:22 | < Attilla> | you have seen the light |
| 18:25 | | You're now known as TheWatcher |
| 18:25 | < gnolam> | No, I have seen the RED MIST OF FUCK YOU AND DIE PYTHON |
| 18:42 | < gnolam> | I AM INCLUDING THE GODDAMN MSVC DLLS SO WHAT THE HELL ARE YOU COMPLAINING ABOUT |
| 18:42 | | * gnolam rages |
| 18:51 | < gnolam> | Install Python on the target machine and it works. Uninstall it and it ceases to work. W.T.F. |
| 19:00 | < gnolam> | And as usual, there is no actually /helpful/ help available. |
| 19:00 | < gnolam> | It's just "you're not bundling the MSVC DLLs, perform these steps to do that". Which I'm doing, but it doesn't fucking work. |
| 19:14 | < gnolam> | FGNNSH |
| 19:20 | < kwsn> | FGNNSH? |
| 19:21 | | Kindamoody|out is now known as Kindamoody |
| 19:24 | < gnolam> | Sound of extreme frustration as expressed by randomly mashing the keyboard. |
| 19:28 | < gnolam> | FINALLY |
| 19:29 | < gnolam> | It works inside the Matrix at least. |
| 19:29 | < gnolam> | That just took... let's see... 5 hours. |
| 19:33 | | * ToxicFrog fiddles with LuaPilot |
| 19:34 | <@ToxicFrog> | I really need to get around to implementing scatter and reduce. |
| 19:34 | <@ToxicFrog> | And possibly loosening the restrictions on gather, if I can. |
| 20:00 | | Kindamoody is now known as Kindamoody[zZz] |
| 20:41 | | Derakon [chriswei@Nightstar-f68d7eb4.ca.comcast.net] has joined #code |
| 20:41 | < Derakon> | Thoughts on this thing? http://pastebin.com/J5prq6Na |
| 20:41 | < Derakon> | Basically I have a library I have to talk to that accepts a handle and returns an error for every single call. |
| 20:42 | < Derakon> | So I want to automate the process of including the handle and processing the error. |
| 20:43 | < Derakon> | But it feels pretty hackish to examine a module like this and toss its functions into a class instance by directly modifying its __dict__. |
| 20:51 | < Fantastic_Phox> | Oh, that's cool. |
| 20:52 | < Fantastic_Phox> | Instructor for an introductory C++ programming class mandates that everyone use Netbeans. He fails to give any directions on how to install it, and the necessary compilers. |
| 20:52 | | * Fantastic_Phox goes to read shit |
| 20:53 | < Fantastic_Phox> | If I've got MinGW installed for a different IDE, I should be able to use the existing installation, no? |
| 20:53 | < Derakon> | Presumably. |
| 20:54 | < Fantastic_Phox> | Ah, but I'd probably have to get additional libraries |
| 20:54 | < Fantastic_Phox> | Damn |
| 20:54 | | * Fantastic_Phox tries, anyways |
| 20:55 | | * Derakon takes a moment to be amused, again, at the fact that he's routing around the university's IRC block via SSH. |
| 20:55 | < Fantastic_Phox> | I've done it a few times |
| 20:55 | < Derakon> | (SSHing onto my home computer and using irssi from there) |
| 20:56 | < Fantastic_Phox> | Even if they knew you were doing it, I don't think they would care. Policy exists to allow loopholes |
| 20:56 | < Fantastic_Phox> | Not that they could block SSH traffic, anyways, without causing serious collateral issues elsewhere |
| 20:59 | < Derakon> | Yeah. |
| 20:59 | < Derakon> | It's mostly just irritating. |
| 20:59 | < Derakon> | I'd rather use a different client. |
| 21:00 | < Fantastic_Phox> | Configure your home computer as a proxy and use something like Putty to access it? |
| 21:00 | < Fantastic_Phox> | I never had any luck doing that, but it's a little outside of my field of experience |
| 21:01 | < Derakon> | Uh. Isn't PuTTY just another telnet/SSH client? |
| 21:02 | < Derakon> | I'm using the built-in SSH in OSX. |
| 21:03 | < Fantastic_Phox> | Ah. Then, stick with that, I'd say |
| 21:03 | < Derakon> | (Also PuTTY would imply that Windows was getting involved at some point, which it is'nt) |
| 21:03 | < Derakon> | Er, isn't. |
| 21:03 | < Derakon> | Heh. |
| 21:03 | < Fantastic_Phox> | Ha, success. Looks like I even had the c++ libraries already. |
| 21:13 | | * Derakon eyes his PG&E bill. |
| 21:13 | < Derakon> | Apparently there's a minimum bill on gas usage. |
| 21:13 | < Derakon> | I could increase my gas use by 70% and still pay the same bill... |
| 21:14 | < Derakon> | I think. |
| 21:14 | < Fantastic_Phox> | Ha, do so. |
| 21:14 | < Fantastic_Phox> | Use the excess gas to power an electrical generator, and cut your electrical bill |
| 21:14 | < Derakon> | I also apparently used 243kWH of electricity... |
| 21:14 | | * Derakon goes to see if that's realistic. |
| 21:16 | < Derakon> | I guess so. |
| 21:17 | < Derakon> | Surprising that that's still $30. |
| 21:24 | < Fantastic_Phox> | 8 cents a kilowatt hour, here. Sounds about the right price, after distribution |
| 21:24 | < Fantastic_Phox> | My parents' place is rural. 75% of their bill is distribution |
| 21:31 | < gnolam> | 243 kilowatt henry? |
| 21:32 | | * Tamber throws an hour at gnolam |
| 21:33 | < Derakon> | Oh, right, lowercase 'h'. My mistake. |
| 21:35 | | celticminstrel [celticminst@Nightstar-5d22ab1d.cable.rogers.com] has joined #code |
| 21:35 | < gnolam> | http://niemann.blogs.nytimes.com/2011/09/30/log-in/ |
| 21:56 | | AnnoDomini [annodomini@60F158.737D66.CA1918.CC7562] has joined #code |
| 21:57 | | * Derakon snerks as he names a variable "numNums". |
| 21:57 | < Derakon> | (It's the number of numbers to allocate!) |
| 21:58 | < Namegduf> | That makes perfect sense |
| 21:58 | < Namegduf> | I don't mind seeing it or working with it |
| 21:58 | < Namegduf> | But whenever I'm naming a variable and something like that, where a variable is named after its own prefix, happens |
| 21:58 | < Namegduf> | I feel compelled to somehow break it |
| 21:58 | < Namegduf> | numNumbers or something |
| 21:58 | | * Derakon nods. |
| 21:58 | < Derakon> | This is just for some test code anyway. It won't be in production. |
| 21:58 | < Namegduf> | Ah. |
| 22:01 | < gnolam> | Coding in Ewok? |
| 22:02 | < AnnoDomini> | He's obviously not coding in Tusken Raider. |
| 22:04 | < AnnoDomini> | http://chanarchive.org/4chan/tg/9646 |
| 22:09 | < McMartin> | I'm not sure numNums is any worse than COM_cast |
| 22:09 | < Namegduf> | COM_cast offends me due to mixing naming styles |
| 22:10 | < Namegduf> | Although naming your varables COM_CAST I guess would be even worse |
| 22:10 | < Namegduf> | Really I guess there's no good way to use something non-constant prefixed with COM_ |
| 22:15 | < McMartin> | Well, it's dynamic_cast, but for COM objects, and COM is an acronym |
| 22:16 | | * AnnoDomini names all his variables COM_cast. All of them. |
| 22:16 | < Namegduf> | McMartin: So com_cast |
| 22:16 | < Namegduf> | Or at least COM_Cast |
| 22:16 | < Namegduf> | Either you're capitalising or you're not |
| 22:17 | < Derakon> | Or expand the COM acronym~ |
| 22:18 | < McMartin> | As a rule you have to do this a lot, and it's already a template -_- |
| 22:19 | | * McMartin hasn't yet decided whether COM or GObject is worse |
| 22:19 | < AnnoDomini> | GObject sounds like a class name. |
| 22:19 | < McMartin> | More or less. |
| 22:19 | < McMartin> | It's the completely insane object system that Gtk and GNOME use built mostly out of the C Preprocessor |
| 22:19 | < AnnoDomini> | In one Uni class, we had classes named CStar, CCircle, CWhatever. |
| 22:20 | < Namegduf> | I take prefixing significant numbers of things with stuff as a hint that either the prefix is redundant or I need to use better namespacing |
| 22:20 | < Namegduf> | It still happens sometimes, though. |
| 22:21 | < Derakon> | So is COM::cast better? |
| 22:22 | < gnolam> | The one decent use of that awful Hungarian 'C' prefix is when you're writing a HUD class. |
| 22:23 | < McMartin> | Derakon: Not really. |
| 22:23 | < McMartin> | Since the whole point here is to replace IUnknown::QueryInterface |
| 22:23 | < celticminstrel> | Yay, everything works... except for exporting to bitmap, that is. |
| 22:24 | < Derakon> | What's your renderer, CM? |
| 22:24 | < celticminstrel> | Hm? |
| 22:24 | < Derakon> | SDL? OpenGL? |
| 22:24 | < celticminstrel> | Oh, OpenGL. |
| 22:24 | < Derakon> | Jetblade has a functional export-rendered-view-to-image system. A moment. |
| 22:25 | < celticminstrel> | I haven't investigated the actual cause of the issue, but while the export "works" the resulting file is invalid. |
| 22:25 | | Rhamphoryncus [rhamph@Nightstar-14eb6405.abhsia.telus.net] has joined #code |
| 22:25 | < Derakon> | Oh, okay. |
| 22:29 | | celticminstrel is now known as celmin|away |
| 22:31 | | AnnoDomini [annodomini@60F158.737D66.CA1918.CC7562] has quit [[NS] Quit: leaving] |
| 22:32 | < McMartin> | Elch. Actually, com_cast isn't any less inconvenient. |
| 22:32 | | * McMartin mutters at C++'s lack of type inference. |
| 22:35 | | * Derakon demonstrates to himself that he could still pass the C portion of the CS60 course he took in college. >.> |
| 22:36 | < Derakon> | Well, the first assignment or so anyway. |
| 22:36 | < Derakon> | (Read: I made a linked-list struct) |
| 22:36 | < Derakon> | This seems like such a hack, but argh. |
| 22:37 | < Derakon> | The problem is that one of the third-party functions I want to call takes a char** argument, by which it returns a 1D array of data that I, the client, had previously handed to it. |
| 22:37 | < Derakon> | And getting this to mesh cleanly with Python seems to be beyond me. |
| 22:38 | < Derakon> | So instead I'm trying to write my own C library that handles allocation and deallocation of buffers and wraps around these functions. |
| 22:40 | | * Fantastic_Phox facedesks at his lab. spacing points equally along a curve, and finding the length and angle from one end. Kind of a classic problem. The math's not hard, but my description of the steps required is incomprehensible. |
| 22:48 | < Fantastic_Phox> | Ugh, and it doesn't even reflect how I'm really going to be doing it |
| 22:48 | < Fantastic_Phox> | What a hassle. I hate writing things in plain english. |
| 22:49 | < Fantastic_Phox> | I'll just code, then describe. Much easier |
| 23:16 | | celmin|away is now known as celticminstrel |
| 23:31 | < celticminstrel> | So apparently the compiler is inserting gads of padding between all the elements of the bitmap header, for some reason. |
| 23:31 | < celticminstrel> | s/elements/fields/ |
| 23:32 | < celticminstrel> | There should be no padding at all, correct? |
| 23:32 | < McMartin> | celticminstrel: Aha, struct alignment |
| 23:32 | < McMartin> | No, without compiler commands, stuff in a struct gets spread out so that each pointer is aligned with its pointer width |
| 23:32 | < celticminstrel> | I even had #pragma pack(push,2) and it didn't do anything. |
| 23:32 | < McMartin> | Pretty sure you want pack 1. |
| 23:33 | < celticminstrel> | ...well, it was 2 in wingdi.h or whatever the header was. |
| 23:33 | < celticminstrel> | But I'll try 1 then. |
| 23:33 | < celticminstrel> | And I suppose apply it to both structs instead of just the one. |
| 23:34 | < celticminstrel> | I'm assuming that if using the Windows structs on a Windows compiler the alignment will be correct. |
| 23:35 | < celticminstrel> | GLUI annoyingly has backspace and delete reversed. |
| 23:37 | < celticminstrel> | Pack 1 didn't seem to help; I'll try __attribute__((packed)). |
| 23:37 | < celticminstrel> | Or whatever it is. |
| 23:41 | < celticminstrel> | I think it's just ignoring my instructions... |
| 23:42 | < celticminstrel> | Wait a second, I'm doing something wrong here. Maybe it's working better than I thought. |
| 23:44 | < celticminstrel> | Okay, there's less padding than I though, but there's still padding. |
| 23:44 | < Lingerance> | Try ordering them from largest size'd element to smallest. The compiler will usually put them on an address that's divisible by their size if it can. |
| 23:46 | < Derakon> | Gotta love how you can add a new file to a project, thereby causing an old file to stop compiling correctly. |
| 23:46 | < Derakon> | Go go accurate compilation error codes. |
| 23:47 | < celticminstrel> | Wait, there's no padding. |
| 23:47 | < celticminstrel> | I think I'm using the wrong size for a long. |
| 23:47 | < celticminstrel> | Either that or every long is being allocated double space, which seems infinitely less likely. |
| 23:47 | < Lingerance> | ... and that is why I use stdint.h for all my integers now |
| 23:49 | < celticminstrel> | 32 bits = 4 bytes doesn't it? |
| 23:49 | < McMartin> | Yes |
| 23:49 | < Lingerance> | Y |
| 23:50 | < celticminstrel> | ...it looks like the long is 8 bytes. |
| 23:50 | < celticminstrel> | Except Windows makes it 4. |
| 23:50 | < Derakon> | The definition of "long" is, IIRC, up to the compiler. |
| 23:50 | < celticminstrel> | Well, MinGW on Windows. |
| 23:50 | < McMartin> | int and long are Whatever The Fuck They Want. |
| 23:50 | < Derakon> | What do you get from "sizeof(long)"? |
| 23:50 | < celticminstrel> | DWORD and LONG are supposed to be four bytes, right? |
| 23:50 | < McMartin> | Uh |
| 23:51 | < McMartin> | long is 4 or 8 depending. |
| 23:51 | < McMartin> | int is 2, 4, or 8, depending |
| 23:51 | < McMartin> | The only requirement is that sizeof(short) <= sizeof(int) <= sizeof(long) |
| 23:51 | < Derakon> | And sizeof(char) == 1. |
| 23:51 | < McMartin> | Often 8 is "long long", but sometimes not! |
| 23:51 | < McMartin> | Especially when targeting 64-bit architectures, which MinGW does not do, but. |
| 23:52 | < Derakon> | I remember some friends who were writing a program to calculate interstellar distances as part of some programming competition. |
| 23:52 | < Derakon> | They discovered that the distances involved exceeded the precision they had allocated for their distance numbers. |
| 23:52 | < celticminstrel> | Well, Wikipedia uses uint32_t for the bmp headers, so I'll follow its example. |
| 23:52 | < celticminstrel> | . |
| 23:52 | < Derakon> | So they put "typedef int long long;" at the top of the program. |
| 23:52 | < celticminstrel> | ...that's not actually possible, is it? |
| 23:53 | < McMartin> | I don't know if you can do it with typedef, but you can certainly do it with #define. |
| 23:53 | < Derakon> | Maybe it was #define. |
| 23:53 | < celticminstrel> | Ah yes, #define could do it. |
| 23:53 | < Derakon> | And you can do all kinds of horrible things to the language if you really want to. |
| 23:55 | < celticminstrel> | Looks like my header is working. |
| 23:55 | < Derakon> | \o/ |
| 23:55 | < celticminstrel> | And for Windows I'm just typedef'ing the windows.h stuff (and made sure to use the same field names in my version, of course), so it should be fine there as well. |
| --- Log closed Fri Oct 07 00:00:02 2011 |