Timestamps are in GMT/BST.
| 00:53:15 | | Watusimoto has joined |
| 01:10:25 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 01:10:54 | | Flynnn has joined |
| 04:44:30 | | Watusimoto Quit (Ping timeout: 246 seconds) |
| 04:56:01 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 05:01:35 | | Flynnn has joined |
| 05:02:50 | | Flynnn Quit (Client Quit) |
| 07:40:38 | | LordDVG has joined |
| 09:53:35 | | sam686 Quit (*.net *.split) |
| 09:53:42 | | Darrel Quit (*.net *.split) |
| 09:53:49 | | ChanServ Quit (*.net *.split) |
| 09:53:50 | | sheepluva Quit (*.net *.split) |
| 09:53:55 | | Nothing_Much Quit (*.net *.split) |
| 09:58:19 | | ChanServ has joined |
| 09:58:19 | | sheepluva has joined |
| 09:58:19 | | Darrel has joined |
| 09:58:19 | | sam686 has joined |
| 09:58:19 | | asimov.freenode.net sets mode +ov |
| 09:58:35 | | Nothing_Much has joined |
| 10:47:35 | | raptor has joined |
| 10:47:36 | | ChanServ sets mode +o |
| 10:47:58 | raptor | good morning! |
| 12:22:01 | | Watusimoto has joined |
| 12:57:34 | Watusimoto | hi bye |
| 12:57:38 | raptor | hi bye |
| 13:00:17 | | Flynnn has joined |
| 13:02:31 | | Watusimoto Quit (Ping timeout: 272 seconds) |
| 13:26:32 | raptor | 'kitchens for sale' strikes again! |
| 13:59:27 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 14:00:08 | | Flynnn has joined |
| 14:01:03 | | Flynnn Quit (Client Quit) |
| 14:04:00 | | Flynnn has joined |
| 14:11:45 | | watusimoto has joined |
| 14:11:45 | | ChanServ sets mode +o |
| 14:11:51 | watusimoto | hi |
| 14:11:55 | raptor | hi |
| 14:32:27 | raptor | i think the tests ran without crashing! |
| 14:33:35 | raptor | all passed! |
| 14:46:05 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 15:01:15 | | Flynnn has joined |
| 16:09:09 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 16:13:29 | watusimoto | Yes, they did! |
| 16:14:09 | | Flynnn has joined |
| 16:31:51 | watusimoto | I am all red up on shaders now... I get it. I'm excited to start using them! |
| 16:31:57 | raptor | oh goodie! |
| 16:31:59 | watusimoto | sorry... "read" |
| 16:32:13 | raptor | if you want to forge ahead - |
| 16:32:35 | | Watusimoto_ has joined |
| 16:32:36 | raptor | The GLES2 class in RenderManager needs to be implemented with all the methods |
| 16:32:55 | raptor | GLES1 is the normal one we use and is fully implemented |
| 16:33:12 | raptor | both inherit the GL class's interface |
| 16:34:09 | watusimoto | I have a general question... let's take testItems as an example. Do we create a shader for the testItem that can generate the vertices (i.e. pass a point and a radius), or generate the points in the app as we do now? The advantage of doing it all in-shader is that we can customize the rendering ore (i.e someone could switch to an 8-point testitem if they wanted). The disadvnatage is that we could potentially decouple rendering from col |
| 16:34:29 | raptor | no |
| 16:34:48 | watusimoto | no? which do you mean? |
| 16:34:49 | raptor | i think as a first pass, you'd just need to reimplment the renderVertexArray method |
| 16:35:34 | watusimoto | well, for test item, we'd need to implement a fragment or vertex shader to make the thing yellow, right? |
| 16:35:51 | raptor | fragment, yes for color |
| 16:36:10 | raptor | we have two code paths for this: |
| 16:36:19 | raptor | in GLES1 that we use now |
| 16:36:35 | raptor | 1. glColor(); then renderVertexArray(); |
| 16:37:12 | raptor | you could re-implement both, the glColor will need a fragment shader, probably a default one that you just pass the color to |
| 16:38:15 | raptor | the reimplementaion of renderVertexArray() may or may not need a shader - I think there is way to do it without, using: http://docs.gl/es2/glEnableVertexAttribArray |
| 16:38:37 | watusimoto | it would probably just be a passthrough function, I'd think |
| 16:38:38 | raptor | that is in place of the the old glEnableClient(VERTEX_blah) |
| 16:39:02 | watusimoto | ok, so were not going to be working with object-level shaders |
| 16:39:14 | raptor | negatory |
| 16:39:28 | watusimoto | I was under the impression that we could use shaders on all the platforms we support |
| 16:39:33 | watusimoto | is that wrong? |
| 16:39:36 | raptor | but we *could* abstract the data for the objects, i.e. the vertices and the colors into some loadable text file |
| 16:39:50 | raptor | "on all platforms" |
| 16:39:52 | raptor | not quite |
| 16:40:07 | raptor | windows (even 7/8) only implmenet OpenGL 1.3 |
| 16:40:23 | raptor | and shaders are OpenGL 2.0 |
| 16:40:26 | watusimoto | I was thinking about the cool effects could create for cores using shaders to do the fancy stuff |
| 16:40:44 | raptor | this means that if you want shaders on windows with opengl, you need special drivers |
| 16:40:54 | watusimoto | yeah... that would suck |
| 16:41:09 | raptor | which, probably most systems *do* have them installed, but not all... |
| 16:41:35 | raptor | i suspect even generic Intel drivers for windows support most of OpenGL 2 |
| 16:41:57 | watusimoto | maybe we could (later) have special shaders for some objects that get run where supported; otherwise the fallback is what we have today |
| 16:42:09 | raptor | yes |
| 16:42:31 | raptor | like maybe we could do the switch at runtime instead of compile time |
| 16:42:37 | raptor | based on certain things we detect |
| 16:42:44 | watusimoto | I do like the idea of using shaders with the source exposed to players so they can mod and make changes in the rendering |
| 16:42:49 | watusimoto | yes |
| 16:42:52 | raptor | yes, i agree |
| 16:43:20 | watusimoto | ok, then for now we just focus on creating shaders to support the abstractions you created |
| 16:43:36 | raptor | i'd say as a first step, that would be good |
| 16:43:58 | raptor | then we can determine if we want to do per-object loading of data somehow... maybe even as Lua! |
| 16:44:13 | watusimoto | lua! |
| 16:44:28 | watusimoto | I wonder if anyone has written a lua->glsl translator |
| 16:44:35 | watusimoto | so we could write shaders in lua |
| 16:44:54 | raptor | or, just put in vertex array data with color data |
| 16:45:00 | raptor | and done |
| 16:45:50 | raptor | i've seen some projects where they dynamically build the glsl from the lua arrays, then compile them and feed them to the graphics card at start-up |
| 16:48:36 | watusimoto | interesting... a while back, I made some changes that pregenerated geometries for some objects (like testitems), and then used gltranslate to shift them into place. doing that in lua would hardly add to the load because it is only done once for each object type |
| 16:53:58 | raptor | you've seen this game?: http://stabyourself.net/mari0/ |
| 16:54:08 | raptor | entirely done in Lua via the Love framework |
| 16:58:28 | | Watusimoto_ Quit (Ping timeout: 276 seconds) |
| 17:00:08 | watusimoto | yikes! |
| 17:00:19 | watusimoto | their not pacman game is crazy |
| 17:00:27 | raptor | hahaha, yeah |
| 17:00:41 | raptor | my wife got a good laugh from not-tetris |
| 17:04:42 | watusimoto | yes, it is a bit nuts |
| 17:04:52 | watusimoto | they like to mix the genres! |
| 17:52:38 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 17:56:08 | | Watusimoto_ has joined |
| 17:59:58 | | Flynnn has joined |
| 18:19:43 | | Watusimoto_ Quit (Ping timeout: 276 seconds) |
| 18:59:23 | | LordDVG Quit (Read error: Connection reset by peer) |
| 19:01:35 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 19:05:45 | | Flynnn has joined |
| 19:21:46 | | raptor Quit () |
| 20:03:21 | | Watusimoto_ has joined |
| 20:35:21 | | Watusimoto_ Quit (Ping timeout: 272 seconds) |
| 20:50:08 | | Watusimoto_ has joined |
| 20:53:52 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 20:59:41 | | Watusimoto_ Quit (Read error: Connection reset by peer) |
| 21:54:17 | | raptor has joined |
| 21:54:18 | | ChanServ sets mode +o |
| 22:02:43 | | watusimoto Quit (Quit: Leaving.) |
| 22:24:54 | | raptor Quit () |
| 22:29:41 | | Watusimoto has joined |
| 23:39:00 | | Watusimoto Quit (Ping timeout: 245 seconds) |