| --- Log opened Tue Oct 16 00:00:43 2012 |
| 00:10 | | himi [fow035@D741F1.243F35.CADC30.81D435] has joined #code |
| 00:10 | | mode/#code [+o himi] by ChanServ |
| 00:14 | | You're now known as TheWatcher[t-2] |
| 00:18 | | You're now known as TheWatcher[zZzZ] |
| 00:43 | | Derakon[AFK] is now known as Derakon |
| 00:45 | | syksleep is now known as sykwerk |
| 01:07 | | OrthiaLap [orthia@3CF3A5.E1CD01.C6689C.33956A] has joined #code |
| 01:07 | | * gnolam tries to figure out which direction is which for a TO-220 package. |
| 02:23 | | Kindamoody[zZz] is now known as Kindamoody |
| 02:23 | | Attilla [Obsolete@Nightstar-836bfbe1.as43234.net] has quit [Ping timeout: 121 seconds] |
| 04:42 | | Kindamoody is now known as Kindamoody|breakfast |
| 05:28 | | Vash [Vash@Nightstar-3ba4108e.wlfrct.sbcglobal.net] has joined #code |
| 05:28 | | mode/#code [+o Vash] by ChanServ |
| 05:44 | | ErikMesoy|sleep is now known as ErikMesoy |
| 05:50 | | Kindamoody|breakfast is now known as Kindamoody |
| 05:52 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 06:19 | | himi [fow035@D741F1.243F35.CADC30.81D435] has quit [Ping timeout: 121 seconds] |
| 06:31 | | Kindamoody is now known as Kindamoody|out |
| 06:42 | < Azash> | http://25.media.tumblr.com/tumblr_mb3hokfupJ1raaf5co1_1280.png |
| 06:46 | | Derakon is now known as Derakon[AFK] |
| 07:00 | | Syloq_Home [Syloq@NetworkAdministrator.Nightstar.Net] has quit [Client closed the connection] |
| 07:06 | | You're now known as TheWatcher |
| 08:04 | | rms [rstamer@genoce.org] has quit [Ping timeout: 121 seconds] |
| 08:11 | | rms [rstamer@Nightstar-e681a855.org] has joined #code |
| 08:12 | | You're now known as TheWatcher[afk] |
| 08:35 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 08:35 | | mode/#code [+o himi] by ChanServ |
| 08:43 | | Vash [Vash@Nightstar-3ba4108e.wlfrct.sbcglobal.net] has quit [[NS] Quit: I lovecraft Vorn!] |
| 09:16 | | Attilla [Obsolete@Nightstar-c0d703de.as43234.net] has joined #code |
| 09:45 | | RichyB [richardb@Nightstar-3b2c2db2.bethere.co.uk] has joined #code |
| 09:48 | | sykwerk is now known as Syk |
| 10:05 | | You're now known as TheWatcher |
| 10:56 | | * TheWatcher eyes this co-worker |
| 10:57 | <@TheWatcher> | If you're getting unexpectedly missing or partial data in files downloaded by wget, perhaps the /first/ think you should have done is log the messages generated by wget |
| 11:12 | < Syk> | hah |
| 11:31 | < gnolam> | TheWatcher: come on, that would be like asking for help with a compilation error and actually saying what the error is and showing code. |
| 11:31 | < gnolam> | Which everyone knows you should never do under any circumstances. |
| 11:35 | < AnnoDomini> | (Well, code showing is sometimes inadvisable because of NDAs, but yes.) |
| 12:43 | | OrthiaLap [orthia@3CF3A5.E1CD01.C6689C.33956A] has quit [[NS] Quit: ] |
| 13:07 | <&McMartin> | TheWatcher: Is this that thing from that one site from the other day? |
| 13:11 | <@TheWatcher> | Yeah |
| 13:23 | | * Azash waves at the channel at large |
| 13:24 | < Azash> | So apparently there's a draft RFC for OAuth 2.0, I wonder if this would be a good time to write a tool for it |
| 13:25 | <&McMartin> | What's your goal in writing the tool? |
| 13:26 | < Azash> | Learn how OAuth 2.0 works, practice my currently-neglected C |
| 13:26 | < Azash> | The latter of which wouldn't be set in stone depending on what I have to do, but a man can dream |
| 13:27 | <&McMartin> | I'd guess OAuth uses a lot of strings in it, which means I'd steer clear of C if possible |
| 13:27 | <&McMartin> | Favoring C-linked C++ |
| 13:27 | < Azash> | Or maybe learn perl |
| 13:27 | <&McMartin> | It's probably not bad as scripting language practice |
| 13:28 | < Azash> | I'm also not quite sure if it's something that's worth doing as a separate tool or as a browser addon |
| 13:43 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
| 13:44 | <&ToxicFrog> | If it's a browser addon, you're probably working in JavaScript anyways. |
| 13:45 | < Azash> | Mm |
| 13:49 | < rms> | It isn't |
| 13:50 | <@TheWatcher> | Trying to make a perl OAuth 2 module would probably be a good project, if potentially a bit challenging for new perl user |
| 13:51 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 14:04 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 14:04 | | mode/#code [+o himi] by ChanServ |
| 14:43 | | You're now known as TheWatcher[afk] |
| 15:08 | | Syloq_Home [Syloq@NetworkAdministrator.Nightstar.Net] has joined #code |
| 15:31 | < iospace> | man i have to make everything similar to EFI's standard in code style |
| 15:31 | < iospace> | bleh :< |
| 15:35 | <&McMartin> | When people told you open source was about freedom they lied ;-) |
| 15:36 | < iospace> | http://visualstudiogallery.msdn.microsoft.com/59ca71b3-a4a3-46ca-8fe1-0e90e3f793 29/ |
| 15:36 | < iospace> | that's so going on my computer when i get home |
| 15:36 | < iospace> | :D |
| 15:37 | <@Tamber> | McM: The freedom to choose who gives you orders, perhaps? :p |
| 15:37 | <&McMartin> | Tamber: The precise mechanism for lock-in. |
| 15:37 | | * McMartin hands iospace Vimperator while he's at it |
| 15:38 | <&McMartin> | MAKE YOUR COMPUTER UNUSABLE BY EVERYOEN |
| 15:40 | < Syk> | i had a fun bug a while ago |
| 15:40 | < Syk> | there were some intel drivers that had a bugged right click hookin |
| 15:41 | < Syk> | so whenever anyone would right click their desktop or Explorer, explorer.exe would instantly crash |
| 15:41 | < Syk> | I unfortunately didn't remember to save an image of the PC for use for annoying people |
| 15:46 | < iospace> | McMartin: i already have Vimperator on here :D |
| 15:46 | < iospace> | and on my netbook and desktop |
| 16:00 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 17:09 | | ErikMesoy is now known as Harrower |
| 17:21 | | Syk is now known as syksleep |
| 17:32 | | EvilDarkLord is now known as Maze |
| 17:44 | < gnolam> | http://ruby-doc.org/core-1.9.3/Object.html#method-i-instance_variable_set |
| 17:47 | | Kindamoody|out is now known as Kindamoody |
| 18:14 | <&McMartin> | Hm, Quantal Quetzal is supposed to hit in two days |
| 18:14 | <&McMartin> | I'm not sure if I want to rush over to a system actually named QQ |
| 18:17 | <&McMartin> | ("Hey, the next one could be Raving Rabbid! And the one after that Solid Snake!") |
| 18:25 | < froztbyte> | haha |
| 18:25 | < froztbyte> | my mind instantly went gutter with that latter name |
| 18:25 | <&McMartin> | At this rate, 17.10 will be Yuzzic Yuzz-a-ma-tuzz |
| 18:25 | < froztbyte> | also, thanks for the notice |
| 18:25 | <&McMartin> | froztbyte: This is yet another reason Kojima must be destroyed |
| 18:25 | < froztbyte> | I've been meaning to check when the next release is |
| 18:26 | < froztbyte> | (since my mirror is going to get hit like mad) |
| 18:26 | < froztbyte> | McMartin: I'm okay with the gutter, it's interesting here |
| 18:26 | <&McMartin> | This is off wiki; I can't find where the official data is. |
| 18:27 | | McMartin [mcmartin@Nightstar-3895ee8e.pltn13.sbcglobal.net] has quit [[NS] Quit: reboot] |
| 18:35 | | McMartin [mcmartin@Nightstar-3895ee8e.pltn13.sbcglobal.net] has joined #code |
| 18:35 | | mode/#code [+ao McMartin McMartin] by ChanServ |
| 18:54 | | Derakon [chriswei@Nightstar-a3b183ae.ca.comcast.net] has joined #code |
| 18:55 | | mode/#code [+ao Derakon Derakon] by ChanServ |
| 18:55 | | * Derakon mutters at his image-viewing code, which is oddly unresponsive. |
| 18:55 | <&Derakon> | As in, it takes about 1.5s to page through 40 528x528 images. |
| 18:55 | <&Derakon> | Despite them all being in RAM. |
| 18:56 | <&Derakon> | Main costs appear to be in glTexSubImage2D, numpy.histogram(), and numpy.tostring(). |
| 18:56 | <&Derakon> | If anyone more familiar with OpenGL than I would be willing to look over the "load new image data into a texture" code, I'd appreciate it: http://pastebin.com/HyhZMQhy |
| 18:57 | | Kindamoody is now known as Kindamoody[zZz] |
| 18:58 | <&McMartin> | If it's *in* glTexSubImage2D the bottleneck is RAM->VRAM communication, most likely |
| 18:59 | <&Derakon> | Well, if I comment out the glTexSubImage2D call then total time spent in this function goes from .8s to about .05s. |
| 18:59 | <&McMartin> | But that is also most likely the part that Does The Work, right? |
| 18:59 | <&Derakon> | Yes, without it there's nothing to see. |
| 19:00 | <&McMartin> | That's a single API call in every implementation of OpenGL that I know of |
| 19:00 | <&McMartin> | Unless the Python data marshalling is super-slow or something? |
| 19:01 | <&Derakon> | I wouldn't think so. |
| 19:01 | <&Derakon> | But I have no idea how I'd check. |
| 19:01 | <&McMartin> | I'm also not sure if that wouldn't just move the cost |
| 19:01 | <&McMartin> | glTexSubImage2D, like all the glTex functions, expects to be handed a contiguous block of RAM win a structured format (usually RGBA quartets) |
| 19:02 | <&McMartin> | *in a |
| 19:02 | <&McMartin> | Getting from "array of python numbers" to that might be costly, maybe. |
| 19:02 | <&Derakon> | Well, that's what the numpy.tostring is doing. |
| 19:03 | <&Derakon> | Converting the array into a format that OpenGL can process. |
| 19:05 | <&Derakon> | (Note that converting the numpy array [8] to string gives '\x08\x00', given we're using unsigned 16-bit ints) |
| 19:05 | <&McMartin> | Aha |
| 19:05 | <&McMartin> | How large an array are we talking, here? |
| 19:05 | <&Derakon> | 528x528. |
| 19:06 | <&McMartin> | Yeah, that "should" be taking about 5 milliseconds. |
| 19:06 | <&McMartin> | If that. |
| 19:07 | <&Derakon> | 40 calls of the pastebinned code takes .8s, so we're off by an order of magnitude here. |
| 19:07 | <&Derakon> | Er, factor of 4. |
| 19:07 | <&Derakon> | Got my division backwards. |
| 19:07 | <&McMartin> | Right |
| 19:08 | <&McMartin> | By comparison, UQM uses that function to update its 640x480 screen and can hit 60 FPS without trying very hard. |
| 19:08 | <&Derakon> | Considerably more pixels. |
| 19:10 | <&Derakon> | Hm, if I strip out all of the GL calls except for glBindTexture and glTexSubImage2D then we drop to .3s |
| 19:10 | <&Derakon> | Or rather, if I comment out the glPixelTransferf calls. |
| 19:10 | <&Derakon> | Those are meant to provide multicolor image overlays. |
| 19:11 | <&Derakon> | But that's not essential functionality. |
| 19:13 | <&McMartin> | Ah, OK |
| 19:13 | <&McMartin> | I thought those were a configuration option but I don't know it well enough to say |
| 19:14 | <&Derakon> | ...wait, no, that breaks display too. Damn. |
| 19:19 | <&McMartin> | Hrm |
| 19:20 | <&McMartin> | Those values aren't default for your use case? |
| 19:20 | <&McMartin> | That might be where the slowness is coming in |
| 19:20 | <&McMartin> | (That is, f is not always 1 and fBias is not always 0?) |
| 19:21 | <&Derakon> | The pixel values are not scaled so that the max value is maxint and the min value is 0, no. |
| 19:21 | <&Derakon> | Basically this code is setting the white and black points for the image. |
| 19:21 | <&McMartin> | Rght |
| 19:21 | <&McMartin> | Hm |
| 19:21 | <&McMartin> | Also, I notice you are using GL_LUMINANCE |
| 19:22 | <&McMartin> | That might not be accelerated to the same degree |
| 19:22 | <&Derakon> | I think that is to allow for texture overlays... |
| 19:22 | <&Derakon> | So you can view two images layered on top of each other. |
| 19:22 | <&McMartin> | GL_LUMINANCE is "how do I interpret this data" |
| 19:22 | <&Derakon> | Oh, right. |
| 19:22 | <&McMartin> | IIRC, it is "alpha is always 1, and copy the single value in the array to R, G, and B" |
| 19:24 | <&Derakon> | All else being equal, using GL_RED instead makes no appreciable difference. |
| 19:25 | <&Derakon> | Using GL_RGBA would require me to tweak the input array so that it generates three times the string data. |
| 19:25 | <&Derakon> | I can't imagine that would be faster than whatever OpenGL has cooked up already. |
| 19:26 | <&McMartin> | One would like to think that |
| 19:27 | <&McMartin> | Um |
| 19:27 | <&McMartin> | When you first specified this texture, what was its internalFormat? |
| 19:27 | <&Derakon> | GL.glTexImage2D(GL.GL_TEXTURE_2D,0, GL.GL_RGB, tex_nx,tex_ny, 0, GL.GL_LU |
| 19:27 | <&Derakon> | MINANCE, dtypeToGlTypeMap[imgType], None) |
| 19:28 | <&McMartin> | Try replacing that GL.GL_RGB with another GL.GL_LUMINANCE. |
| 19:28 | <&McMartin> | That way it can stay a Luminance Texture on the hardware too. |
| 19:29 | <&Derakon> | Makes sense; seems to gain about .15s or so. |
| 19:29 | <&McMartin> | And if they're 16-bit luminance values in the data, maybe GL_LUMINANCE16 will be better still? |
| 19:30 | <&McMartin> | And then you can make the type be GL_UNSIGNED_SHORT |
| 19:30 | <&McMartin> | Oh, that latter is handled by dtypeToGlTypeMap |
| 19:30 | <&Derakon> | The type is already GL_UNSIGNED_SHORT in this case. |
| 19:30 | <&Derakon> | Yeah. |
| 19:31 | <&Derakon> | Hm, using GL_LUMINANCE16 gives an "invalid enumerant" error. |
| 19:31 | <&McMartin> | That one's only valid for the first LUMINANCE in the call, not the second. |
| 19:31 | <&Derakon> | Right. |
| 19:31 | <&Derakon> | Okay. |
| 19:32 | <&McMartin> | (The first is "what is the format of this texture you are creating", the second is "how do I interpret the data you're passing in" and is modulated further by the dtypeEtc) |
| 19:32 | <&Derakon> | Runtime difference is less than the error bars from that, I'm afraid. |
| 19:32 | <&Derakon> | Maybe a slight improvement. |
| 19:38 | <&Derakon> | Well, thanks for the assistance. |
| 19:38 | <&Derakon> | I think I may end up posting to gamedev.net's OpenGL forums, or something along those lines. |
| 19:53 | | * iospace wonks around in PCI space |
| 19:56 | <&Derakon> | Hm, dramatic improvement if I get the texture size down below 512x512. |
| 19:56 | <&Derakon> | Down to .154s for the 40 calls. |
| 19:58 | <&Derakon> | Of course, then I can't show the entire thing. |
| 19:58 | <&Derakon> | I actually wrote a tiler system so I could keep the texture size sane, but it pads the image with zeros prior to slicing it up -- and doing 4 512x512 textures is worse than doing 1 528x528 texture. D'oh. |
| 20:03 | | * iospace slaps her co-worker |
| 20:03 | < iospace> | comment better hor! |
| 20:04 | <&McMartin> | Derakon: If I had to hazard a guess, your hardware only accelerates up to 512x512 |
| 20:05 | <&McMartin> | What if you mix a 512x512 and four 256x256 ones~ |
| 20:06 | <&Derakon> | I'm trying something similar -- 512x512 + 32x512 and 512x32, or something like that -- basically removing the zero-padding. |
| 20:06 | <&Derakon> | Unfortunately I seem to have broken something somewhere in my code. |
| 20:07 | < iospace> | ... |
| 20:07 | < iospace> | ok then |
| 20:07 | < iospace> | "Returns: EFI_SUCCESS - Because failure is not an option" |
| 20:08 | <&McMartin> | Failure is always an option |
| 20:08 | <&McMartin> | ("It's time to reconsider the 'failure' option") |
| 20:08 | <&Derakon> | Nergh, windows! |
| 20:09 | <&Derakon> | OpenGL error spew is aggravating to deal with. |
| 20:10 | < iospace> | McMartin: i didn't write it! |
| 20:14 | | * Derakon eyes numpy.ceil(). |
| 20:15 | <&Derakon> | Why the fuck are you returning a floating point value? |
| 20:15 | <&McMartin> | That is what ceil does |
| 20:15 | <&Derakon> | ceil is "round up to the next integer", so it should always generate an integer. |
| 20:15 | <&McMartin> | http://linux.die.net/man/3/ceil |
| 20:16 | <&McMartin> | ceil never generates an integer in any typed language I've ever programmed in. |
| 20:16 | <&Derakon> | ...okay, then the entire world is mad. Oh well. |
| 20:18 | | * Derakon sighs. width = int(numpy.ceil(float(newImage.shape[0]) / self.tileSize)) |
| 20:20 | <&Derakon> | On the plus side, variable-sized textures does seem to work. .29s to do 80x calls to that function without any zero-padding. |
| 20:20 | <&Derakon> | Still not at 60FPS but we're over 40FPS, which is something like twice where we were when I started. |
| 20:21 | <&McMartin> | Also, you're doing this 40 times where UQM did it once. |
| 20:22 | <&Derakon> | No, I'm doing it twice per refresh. |
| 20:22 | <&Derakon> | The 40x is just to have enough data to be measurable. |
| 20:22 | <&McMartin> | Ah |
| 20:22 | <&Derakon> | (Twice because I have one 512x512 texture and one 512x16 texture) |
| 20:23 | <&McMartin> | (Don't you also need a 16x512 and a 16x16 one?) |
| 20:23 | <&Derakon> | No, I lied about the image size earlier; it's 512x528, not 528x528. |
| 20:23 | <&McMartin> | Aha. |
| 20:23 | <&McMartin> | That is still an odd size |
| 20:24 | <&Derakon> | Cameras. *shrug* |
| 20:26 | <&McMartin> | Hee |
| 20:26 | <&McMartin> | [kickstarter] Gunther | Also at $4.0m Chris Avellone will be forced to play Arcanum. |
| 20:26 | <&McMartin> | [kickstarter] Gunther says, "chip in all you can, people" |
| 20:28 | < Harrower> | I think I am missing some context. |
| 20:29 | <&Derakon> | Next order of business: numpy.histogram is taking .56s in that same time period, thus now representing over half the total runtime for this function. |
| 20:30 | <&Derakon> | (I need the histogram to display to the user and help them manipulate the black- and white-points) |
| 20:36 | <&McMartin> | Harrower: That's because I mispasted |
| 20:36 | <&McMartin> | But this is apparently part of the new stretch goals for Obsidian's Kickstarter |
| 20:37 | <&McMartin> | http://www.kickstarter.com/projects/obsidian/project-eternity |
| 20:37 | <&McMartin> | If you liked Planescape:Torment or the early Fallout gamse, this might be relevant to your interests~ |
| 20:41 | | * Moltare reads up |
| 20:42 | < Moltare> | RuneScript doesn't have ceil or floor, but if it did they'd generate integers |
| 20:42 | < Moltare> | Because we only have integers in the first place! :D |
| 20:42 | < Moltare> | ;_; |
| 20:42 | <&Derakon> | Ahh, non-floating-point systems. |
| 20:43 | <&Derakon> | Angband's C code doesn't use floating point. |
| 20:43 | <&Derakon> | But it was first written in the early 90's. |
| 20:45 | <&Derakon> | Okay, down to .63s overall to refresh 40 times, which is almost evenly split between OpenGL loading the texture, and Numpy telling me how the data is distributed. |
| 20:45 | <&Derakon> | I think that's acceptable. |
| 20:45 | < Harrower> | What is wrong with Arcanum? |
| 20:45 | < Harrower> | I disliked it, but it didn't seem to have major obvious stupids |
| 20:45 | <&Derakon> | ...except that I somehow broke display again, dammit. |
| 20:46 | < Azash> | Arcanum~ |
| 20:53 | <&McMartin> | Harrower: Avellone helped *design* Arcanum |
| 20:53 | <&McMartin> | And he's also been forced to play Icewind Dale 2 already |
| 20:54 | < Harrower> | McMartin: It still feels as though I'm missing something. Like, the joke I'd *expect* is that Avellone helped design it, never played it, praised it somewhat cluelessly, there was something terribly wrong with Avernum, and now he will have to experience the end user's suffering for his mistakes. |
| 20:56 | <&McMartin> | Arcanum, not Avernum |
| 20:57 | < Harrower> | Um, yes. |
| 20:57 | | * Harrower blames the guy being named Ave- |
| 21:01 | | Harrower is now known as ErikMesoy |
| 21:05 | <&ToxicFrog> | Avellone was not actually involved in the development of Arcanum |
| 21:06 | <&ToxicFrog> | I could in fact believe that he hasn't played it |
| 21:06 | <&ToxicFrog> | ErikMesoy: it does have a few, er, balance issues |
| 21:06 | <&ToxicFrog> | Like mages being overpowered all day every day and the first hour or two being utterly miserable as a technologist |
| 21:10 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 21:21 | | Maze is now known as EvilDarkLord |
| 21:22 | < iospace> | yay ACPI Dx |
| 21:23 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 21:23 | | mode/#code [+o himi] by ChanServ |
| 21:26 | < gnolam> | A derivative of ACPI? |
| 21:27 | < iospace> | no |
| 21:27 | | * iospace bops gnolam |
| 21:51 | | Derakon [chriswei@Nightstar-a3b183ae.ca.comcast.net] has quit [[NS] Quit: leaving] |
| 21:58 | <&McMartin> | ToxicFrog: Talking to others about Arcanum, I've gotten the impression that Mages start getting their faces eaten once melee-heavy enemies show up unless you start playing them more carefully |
| 22:01 | < iospace> | ... dear nmake, when you error out, please give a clearer error as to why you errored out |
| 22:01 | | You're now known as TheWatcher |
| 22:03 | | OrthiaLap [orthia@3CF3A5.E1CD01.C6689C.33956A] has joined #code |
| 22:04 | <&ToxicFrog> | McMartin: it's true that eventually, you may reach a point where spamming Harm is not guaranteed victory. |
| 22:05 | | * Azash enters realtime and does Fireflash as quickly as his fingers will allow |
| 22:05 | | celticminstrel [celticminst@Nightstar-05d23b97.cable.rogers.com] has joined #code |
| 22:06 | <&ToxicFrog> | But by then you have Dweomer Shield and Disarm to neuter enemy mages and fighters, respectively, and will shortly - if you don't already have it - have Disintegration for those occasions when you just need to instantly, automatically kill something. |
| 22:07 | <&ToxicFrog> | Or you could just cast Tempus Fugit and then figure out some way to take advantage of your entire party getting 3-6 actions per enemy action. |
| 22:07 | < Azash> | I liked getting summoning spells and abusing the shield glitch |
| 22:07 | <&ToxicFrog> | Also, at this point you'll have Teleport, which lets you blink around and collect all the best party members and whatnot really easily. |
| 22:08 | <&ToxicFrog> | Technologists are still viable (especially once you hit 100 tech and spells just stop working against you entirely), but mages are sick. |
| 22:08 | < Azash> | I liked playing a neutral rogue personally |
| 22:09 | < Azash> | Also for warriors getting some tech at the start was great so you could make balanced swords, afaik |
| 22:10 | | Binem [b1n@5FDE33.2FB5BF.3CFADD.7F10C5] has joined #code |
| 22:10 | <&ToxicFrog> | Yeah, there's some nice fightery stuff in the low tech levels |
| 22:11 | <&ToxicFrog> | Herbology has Heal and Restore Fatigue at levels 1 and 2, Electricity has the Charged Ring at level 2, and Smithy has the Balanced Sword and Ultralight Axe at 2 and 3. |
| 22:11 | | * Binem slaps Binem around a bit with a large trout |
| 22:11 | <&ToxicFrog> | I think Jayna and Magnus can do Herbology and Smithy respectively, too, and you get both of them early on |
| 22:39 | | Binem [b1n@5FDE33.2FB5BF.3CFADD.7F10C5] has quit [Ping timeout: 121 seconds] |
| 22:46 | < celticminstrel> | ...? |
| 23:06 | | Rhamphoryncus [rhamph@Nightstar-cc6253d6.abhsia.telus.net] has joined #code |
| 23:06 | < iospace> | time to break everything \o/ |
| 23:07 | | OrthiaLap [orthia@3CF3A5.E1CD01.C6689C.33956A] has quit [Ping timeout: 121 seconds] |
| 23:07 | | * iospace blinks |
| 23:07 | < iospace> | build... worked? |
| 23:08 | | Orthia [orthianz@3CF3A5.E1CD01.C6689C.33956A] has quit [Ping timeout: 121 seconds] |
| 23:09 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 23:11 | | * gnolam ponders Lua. |
| 23:11 | < iospace> | ok, now time for more breaking :3 |
| 23:13 | < iospace> | orrrrrrrrrrr not~ |
| 23:23 | | himi [fow035@Nightstar-5d05bada.internode.on.net] has joined #code |
| 23:23 | | mode/#code [+o himi] by ChanServ |
| 23:28 | | ErikMesoy is now known as ErikMesoy|sleep |
| 23:31 | | Orthia [orthianz@Nightstar-6ca59a6f.callplus.net.nz] has joined #code |
| 23:32 | | OrthiaLap [orthia@Nightstar-6ca59a6f.callplus.net.nz] has joined #code |
| 23:36 | < Rhamphoryncus> | augh. Updating an android mod is a friggen nightmare. Too many different things by different people with different names, covering different components, none of it coming together in a coherent way |
| 23:37 | < Reiv> | what's the mod? |
| 23:37 | < Rhamphoryncus> | cyanogenmod |
| 23:37 | < Rhamphoryncus> | Currently a d2tmo nightly of CM10 for galaxy S3 |
| 23:38 | < Reiv> | Whosawhutnow |
| 23:38 | < Reiv> | I got, uh, the device name. :P |
| 23:38 | < Rhamphoryncus> | SGH-T999? |
| 23:40 | <@TheWatcher> | I believe the good chap is saying that the only part of your sentence he understood is "galaxy S3", the rest remaining confusing and acronym-laden |
| 23:40 | < Rhamphoryncus> | ah |
| 23:43 | < Rhamphoryncus> | d2tmo is sort of a branch name. It could actually be a branch name, but all I get access to is a tarball with that in the name :P |
| 23:44 | < Rhamphoryncus> | CM10 is cyanogenmod 10 |
| 23:48 | | Kindamoody[zZz] is now known as Kindamoody |
| --- Log closed Wed Oct 17 00:00:59 2012 |