mrwonko

Huh? Why would that be frowned upon? If you release under GPL you grant rights for commercial use. See the Lugaru fiasco on how not to do it though. Maybe because you don't want to share revenues or like open source? Or you're a long time Jedi Academy modder that is already comfortable with the engine. Maybe you just want the nice mod support of idtech 3. There are a couple of reasons I could think of.

I'm not eezstreet, may I still answer? If not, screw you I'm doing it anyway. No, this is not built on top of Open Jedi Project, it's largely orthogonal to it. You can still play Open Jedi Project in OpenJK. OPJ is a Mod, OpenJK the engine (although an improved modbase is included with the source).

Yeah, so since the Jedi Academy source got released, we can go nuts with high detail now. Provided this project takes long enough we'll have normalmaps, parallax occlusion mapping, dynamic lighting and higher possible polygon counts. Maybe even tesselation. So it would be fairly cool to create a reference level of what is technically possible. And if that does not happen in time, we can just bake a full render - lightmaps & mega textures, Blender style!

Improving AI is not a focus at OpenJK at the moment, but it might be something to investigate in the future. I agree that better AI is better.

I'm fairly sure MB2 has custom code to make it animate, that won't work in base Jedi Academy.

You hardly need to do any coding. You need to remove the CD check and that's about it. Mostly you just need to recreate the assets since you can't use Jedi Academy's, most importantly the animations. And if you simplify movement outside the water to walking, crouching and jumping, you only need very few since most of your game is in the water, with what I believe are already custom animations. That requires access to Hydroball's source code though - do you have that? Otherwise you'll either need to recreate all animations or accept glitchy animations.

If you created the UV by marking seams and running unwrap, it's trivial since the seams are highlighted. If you chose some kind of auto unwrapping, it's more difficult. Try using a gradient texture like this one, you should see the seams fairly clearly.

Yes it could. First person and third person use different models. Check if it works with the default model.

Probably an error in the shader, disabling the depth test or something. Did you change anything there? Post it, please.

If there ever was any rush, the release of the source code changed that. I'll be busy poking the code now, but that doesn't mean we can't do this as well.

Wrong. Splitting is required if there are any seams on your uv map, i.e. if you have to make cuts in your model to be able to map it to the texture. Learn what UV Seams are.

I don't need to tell you how amazing this is. Give me a year. Raven is aware of the mixup by the way.

Well shit.

Justified, since I actually missed your post somehow. Why did I only get a notification about the bump? Regarding your actual problem, try what redsaurus said.

Feel free to stick around. I'm thinking about making this fairly open anyway, with some shared storage (via dropbox, github or whatnot) where everybody can contribute. Some preliminary planning/style previs would probably be good, but once that's out of the way anybody should be able to contribute.

Then welcome aboard! If someone just uses it lightly (chairs and other smaller objects) s/he's welcome as well, we'll need some objects as well, after all. Also, ping @@Pande.

Maybe you should've thought about that before participating in a collaborative project.

Yes, that's obviously important in any collaborative project. Be aware that others may change your work. It's usually for the better, though if you disagree try to find a solution by discussion instead of changing it back and forth. Also, you kind of lose the rights to your work as in you can't later decide "nah, you mustn't use what I created." (You can of course still use it for personal projects, but what's contributed becomes ours, not just yours.) As for the scale, at least that's easy to change when modelling since you don't have to adhere to some grid.

Hey, I've briefly talked about this in the JKHUB Pass-Me-Around planning thread, but it should really be its own thing, hence me opening this thread. Basically, I'd like to create a whole map in Blender. And I'd like you to join me. One of the nice things about Blender is that collaboration is easy. You can link objects from other .blend files and they'll even be updated when the original is changed, allowing for parallel work. So this wouldn't even necessarily have to be a sequential thing, we could all work on it in parallel. Who would be interested in joining me? Some experience with Blender would obviously be appreciated, but you might be able to contribute even if you're just starting out. Pretty much every imaginable environment has some simple objects in it, after all. I'm not too good myself - I know the technical side, so if you need to do a certain operation I can probably tell you, but I'm not very good at using those operations to come up with a cool model. And what could we do? We should probably keep it small, duel map sized, so we have a better chance of finishing it. And gameplay is pretty hard to get wrong in a duel map, so that's another thing we wouldn''t have to worry about too much. As for the theme, I'd like something that highlights the strengths of modelling, something that would be hard to do with brushes. (Yes, there's hardly anything you can't do with brushes, but square rooms are inherently easier.) Terrain or something organic would be obvious choices. Talking about organic - it could even be animated, since we can use md3 models, which support animations. (They get huge though, since it's per-vertex animation.) Or something highly destructible, akin to Fearis Incident - .rof files allow for some nice rigid body animations and combining that with Blender's physics baking capabilities leads to interesting possibilities. But I get ahead of myself - first we should see who's interested and think about the kind of map we'd like to make. So, who's with me? So long, mrwonko

I guess it would help if you uploaded the .blend for us to take a look at...

Since we're passing the map around anyway, how about just publicly uploading it for everybody to judge? That way the current mapper would get way more feedback.

For the common description only, since they were too lazy to copy it to each weapon. It doesn't actually exist.

You did not read the tutorial, did you? In edit mode, select all (A) and Triangulate the mesh (Ctrl+T) - you can't have any faces with more than 3 vertices, glm does not support that.

Are you using a custom skeleton or creating a new playermodel? Either way, this is where the problem comes from: A glm file only refers to bones (for weights) by number. To be able to get those numbers right, the exporter looks them up in the .gla file, which contains the actual bone names. In order for this to work, the exporter needs to load said gla file. So you specify it when exporting (in the export settings, below the bookmarks in the export file choice) - the setting in question is ".gla name". You specify that relative to the base folder, just the way Jedi Academy does (because this path gets saved directly into the .glm), so for playermodels the default of "models/players/_humanoid/_humanoid" is just right. (Notice the lack of an extension - that's correct and important!) The exporter has to be able to locate that file, of course. So it has to be unpacked to the correct directory, and the exporter needs to be able to locate your base folder. In order to also support mods, the exporter searches for "/GameData/" in the chosen file name and assumes the folder below that is the base folder (e.g. "/GameData/base/" or "/GameData/mycoolmod/"). If you're not actually working in your base folder, or need to manually specify the path for some other reason, you can do so via the "Base Path" setting - so if you've unpacked the "_humanoid.gla" to "D:/JKA-Models/models/players/_humanoid/_humanoid.gla", you'd set it to "D:/JKA-Models/". (Not sure about the trailing slash, and I don't think I've ever actually tested exporting to folders outside of base.) Hope that clarifies it. Also take a look at the custom vehicle tutorial and the manual for further information. [sharedmedia=tutorials:tutorials:127]

There's the deformvertexes autosprite shader keyword for camera-facing planes, used e.g. for flares. That what you're looking for?