Boothand

It's a cool feature vs troopers and bots. Or if you had to pick it up like Yoda does. But in a standalone MP mod based on fencing, having saberthrowing enabled would go very much against the purpose and breed many harsh feelings, as it does currently in JK2, for example What role it has in JKG, however, is up to the developers, as the outlooks are that I will wait for that instead. If someone is up for it though, it would still be a much embraced mod for JK2.

Actually if this system was realized, there wouldn't be any saberthrow. It would be a gametype fit for no force. Either way, I'm one of those that think saberthrowing is unfit for Jedi knight in either case

Yeah, actually this could go well into the system they're making at JKG. They are currently going for 8 directional triggered-by-keymove animations. Which I think is great. A little wiggle room could make it more interesting if done right.

I think it just needs to be there. I haven't tested this in a while, but just test it out on small brushes close to the floor or even patch meshes maybe. I suppose you should make the area small so that players won't do what @ described.

In your systems folder, there is a shader called "cushion". Place this below the ladder, and falldamage won't happen here. It also won't play footstep sounds I think.

I guess by the time the opponent gets 90 degrees to your side, you will have released the block enough to turn towards him, or you would step out of range so that you can turn towards him. The camera moving at half speed would work in a way that lets the player turn a bit while doing hits and blocks. The "camera" is really just another word for "the player's facing direction".

Both JK2 and JKA's saber systems mostly base itself on timing hits, using fast/slow styles, having reach and pulling off moves. Which is great. It should be played. This would be a mod for people who are dying to fight like they do in the movies, with a more risky approach, and much more reaction based. Instead of saber styles, every two opponents would have their own pace of fighting and any hit would be decisive. Yeah, I hope others are interested. Or if I could get these suggestions over to the JKG community I will try that.

Really interesting that JKG is working on something like this. Maybe I can get some suggestions over on how to do some things. I should have specified this is for MP. The reason why it's (kind of) simple is that you wouldn't have to deal with AI code, in my case. I've never seen any real inconsistencies in saber collision in JK2 with two players with normal ping, actually. And judging by the old game Die by the sword, which also had netplay, it shouldn't be a real problem I think. It worked fine in terms of latency. Mount & Blade is great, but some things about it annoy me so much. It has no dynamics in the movement, and the attacking and blocking is based on what you click, not what happens on the screen. You can't deal with two opponents since you can only hit one person per swing and the lack of movement in that game makes it almost impossible to deal with them. I'm really excited to try JKG, maybe I even can involve myself in the making. However, what I describe offers some very interesting (in my point of view) aspects, which I would love to see. Let me explain the blocking in more detail this time . When you right click and pull the saber in a direction, you activate an animation sequence which corresponds to the mouse movement. Pulling to the side will activate a different animation sequence than pulling upwards or pulling down + a side. Video with incredibly bad animations to demonstrate what I mean: http://www.youtube.com/watch?v=cTiHmA85mlI Attacking would be based on the same concept - holding mouse1 + free move within an animation triggered by what direction you choose.

I'm not sure how to put this, to seperate "this would be kinda nice" from "this would be the greatest thing in the world". It would be the latter. I will start this off as basic as I can. I have an idea that seriously keeps me from falling asleep. The reason is because I believe it shouldn't be too difficult. The idea, in its shortest statement, is: Having an absolute saber-collision system with some basic dynamic attack directions and equivalent block directions, controlled by the mouse. If anyone has played Mount & Blade, it will be related. On the other hand, if someone played Die by the sword long ago, it would also be related. How it would/could work: Swinging: We make a simple animation sequence going from left to right. Every animation corresponds to a cursor coordinate. Holding mouse1 and dragging left and right swings left and right, with a set speed limit. The same happens for vertical swings. Letting go of mouse1 (or not pressing it) will make the camera move as usual. When you are holding mouse1 and swinging, the camera moves at half speed or something that makes it possible to navigate while swinging the sword with the mouse. If this works, we can add more depth to it by having a zone it can move into, aside from its base-trajectories. If you ever played Die by the sword, this was totally chaotic and it was almost impossible to control the sword with excellence. So if you could have a base-path and then possibly expand it, that would have been a nice feature. Blocking: Right clicking will trigger animations that also correspond to cursor coordinates. Pulling right would block a swing coming to your right. And left for left. Pulling the mouse up would block swings from above. Same deal with the camera I suppose. Attacking or blocking shouldn't make it very hard to face your opponent if he or she moves around. Winning/losing: I'd want it to be close to the realistic Star Wars style (realistic in a Star Wars-sense). Attacking and blocking is no longer random since you see exactly what's happening and realize exactly what you must do. So it's fair to instantly die when you get sliced by a lightsaber, at least in the upper body area. For starters, we could increase the saber damage a lot and enable chopping. This is the essence of it. It would of course fit into a no force setting. I would also love features like dodging hits in an accurate way. Imagine using a key + mouse movement to duck at the speed you choose or making tiny jumps and so on. If anyone would be interested in trying to help me make this, I would be more excited than all the school girls at all their first days. Even just making basic basic prototypes to see if it could be realized would be great. I have yet to learn programming, so I wouldn't be able to compile much. I do understand a bit of the language though, so I wouldn't be impossible to communicate with. With some guidance I could probably try to contribute with animations and eventual visual necessities. Anyone else out there who thinks this would be extremely cool and fun? Please let me know!

I threw this together in a hurry just now, because I have to go, but it might be something to play around with. It's not the solution however, as far as I know because I can't seem to cap it. But I can't think of any other way to do it, than to actually bend the texture with the brush. Maybe you could use the cap you already have and build this around it. Don't know how it will look in the middle though. http://www.youtube.com/watch?v=xL3c6H2YHdc PS. I hope I didn't pronounce "bevels" the wrong way. I might have. Haha!

Hmm, it doesn't do anything differently for me. It's almost identical with the shader I posted. It gives me this anyhow: https://dl.dropboxusercontent.com/u/58757568/noblend.jpg This was made with a square image using a faded round alpha channel from black to white, like the one in my first post.

Wow, thanks for this one! It took me a (good!) while to crack this one, but now I've got it. I'll definitely use this for terrain blending. Thanks again! I guess for simple decals (like on walls, or for variation in very small areas of terrain) it would be more expensive for the engine to do it this way, because of the extra segmentation required. Opposed to an extra texture. Or what do you think?

You could save websites so that you don't need the internet later to read tutorials. I think if we somehow give even clearer instructions on how to do these things, there's a big probability it will just as well bear many follow-up questions, specifications and troubleshooting (which I believe could be solved by reading tutorials. It probably takes a lot less time than waiting for answers here, especially when it's on the very basic level). You could also download these mapping tutorials by Mug and watch them later: http://www.youtube.com/user/JediKnightHub Mapping tends to be the stuff you don't do in a hurry. I can spend 4-5 hours at a time trying to figure out how some things work. It needs patience.

If you haven't yet, read through Richdiesal's mapping university. It has most of the basics in a very comprehensible way. For all the things that have to do with messages - target_print will print a message in the middle of the screen. A trigger_multiple can define an area where something will happen if someone walks into it - for example this can trigger you to teleport. Basically, everything the trigger is connected to will *happen*. A target_activate deals with whether or not something is usable yet. Combined with the trigger_multiple, you should be able to put something together to make it work the way you want. Once you learn these things, you'll also understand how to make many other things work, including some of the other things you mentioned.

I've seen programs installed in the program files directory behave like this when they believe they don't have administrator rights. Have you tried saving after running GTK as administrator?

Yeah well, I understand them enough to never copy one without research. But some things seem very inconsistent when combining with other methods, thus creating a lot of experimentation even when it contradicts my sense of logic But be sure to tell me if you make a tutorial, or if you have any other references I could use to make it. Or if it's easy to do but hard to explain, I could just observe instead if you have an example pk3 or so. Either way, too bad it's so unsupported, as I had great ambitions with terrain blending and the likes. At least I can still make good use of decals with the all-or-nothing method.

What kind of nolightmap shader do you mean? (Special, as in one for every texture it lies upon?) Would it defeat the purpose of having decals in the first place (opposed to making different versions of textures you already have, and gaining map size)? If it's not either of these, couldn't it just be a stage in the shader to cancel out the transparent areas being lit twice? That said, I didn't quite understand what you meant by changing the texture coordinates seperately. Either way, could you give an example?

After trying all I can think of and some more, I can't seem to simply fade in a decal I'm making, while keeping it lightmapped. In-game what is showing up is the brightest area of the alpha channel, I think. It is at least a hard round border, instead of a faded one. Shader: textures/Bootland/terrain_decal1 { qer_editorimage textures/Bootland/terrain_decal1.tga surfaceparm nonsolid { map textures/Bootland/terrain_decal1 alphaFunc GT0 depthWrite } { map $lightmap rgbGen identity blendFunc GL_DST_COLOR GL_ZERO depthFunc equal } } Decal: https://dl.dropboxusercontent.com/u/58757568/terrain_decal1.jpg Alpha channel: https://dl.dropboxusercontent.com/u/58757568/alphachannel.jpg In-game: https://dl.dropboxusercontent.com/u/58757568/shot0055.jpg (placed upon an irrelevant mud texture for clarity) With blendFunc blend in the first stage: https://dl.dropboxusercontent.com/u/58757568/shot0056.jpg with alphaFunc GE128: https://dl.dropboxusercontent.com/u/58757568/shot0057.jpg How would I do this?

Yeah, I'm sorry about that I have already posted so many topics lying around, and it feels a bit uncomfortable pushing all my questions (they are many) onto the forums. The mapping section in lucasforums is just a wasteland of old unresolved threads of mine, as an example... Yeah, I forgot to specify I'm mapping for multiplayer. Everything would have been so much easier in singleplayer... You resolved my shader problem! Thank you! I wasn't aware you could use 192. I thought 128 was max. Now that I try with 128, that works equally well though. As for the train - with or without the origin, the problem remains the same. It spawns at the center of the map (at least in multiplayer). Step by step: 1. https://dl.dropboxusercontent.com/u/58757568/train1.jpg (door + origin becomes a train) 2. https://dl.dropboxusercontent.com/u/58757568/train2.jpg (I make the path corners and connect them - train > path > path) 3. https://dl.dropboxusercontent.com/u/58757568/train3.jpg (I make a trigger to toggle the train going up and down) 4. https://dl.dropboxusercontent.com/u/58757568/train4.jpg (I compile, and train spawns at center coords of map) 5. https://dl.dropboxusercontent.com/u/58757568/train5.jpg (I click the trigger and train goes from path1 to path2 and is now ready to be toggled by the trigger). This is why I'm trying to work around it - all I need is a way to make two valid triggerings from worldspawn. That would set the door to its correct position. Hope that clears it up!

1. As some may have noticed, func_trains in JK2 are very bugged. If you want them to be togglable (using a trigger), you cannot connect the last waypoint back to the first one, or else it will always run (using a trigger then will make it bug and teleport between the points). To make it togglable, the setup looks like this: trigger_multiple > func_train > point1 > point2. This will cause the train to spawn in the (0 0 0) center of the map. It will work once you use the trigger, but it will move from point1 to point2. The train has no origin brush, but having one causes even more problems. Workaround: A trigger_always (runs targets on worldspawn) connected to a target_relay which triggers the func_train. So when the map starts, the train starts moving, so that it's not stuck in the center of the map. However, I need to make it (almost) immediately be triggered a second time, so that it moves back towards point1. During testing, it works using a trigger manually with use button (it does not need to reach point2 before going back to point1, but has a 500 ms delay between triggerings or so). What I've tried: Trigger_always > > target_relay & target_delay (target_delay has wait value of 2000) > > func_train > point1 > point2. Target_delays seem not to work at all, not even when triggered manually. Unless there is a way to set a delay for the relay (rhymes not indenday), I don't know how I should make the func_train go back to point1. Ultimately what I want to achieve is just to get the train to spawn in its starting position as it's supposed to. _______________________________________________________________________________________________________________________________ 2. I have made a plant, with alpha channels. It looks good in-game, but when putting two next to each other, one of them keep flipping "infront" of the others, in a way. Here's the shader: textures/Bootland/sapling1 { qer_editorimage textures/Bootland/sapling1.tga cull disable surfaceparm nonsolid q3map_nolightmap // <div> <func> <base> <amplitude> <phase> <freq> deformVertexes wave 10 sin 0 0.8 0.5 0.4 { map textures/Bootland/sapling1.tga blendFunc GL_SRC_ALPHA GL_ONE_MINUS_SRC_ALPHA alphafunc GT0 } } Removing the blendfunc altogether makes it cellshaded and looks messy on a distance (however fixes the fighting issue). Here's a video showing the problem: http://no.twitch.tv/boothand/c/3087720 (what I said about there being a problem on my other computer, about the hanging sign - was incorrect. The problem has solved itself. I do however still need help with the bugging sprites (the chains) above the sign - which I have another thread for). Any ideas?

Hey. Did this have any results yet? Also, any ideas if this could possibly work in multiplayer, in JK2?

I've been trying to do this for many hours... I'm quite close, but something really weird is going on. I have a chains.tga I've made myself, and want to use it as a sprite (autosprite2). Here is the shader: textures/Bootland/chains { qer_editorimage textures/Bootland/chains.tga deformVertexes autosprite2 q3map_onlyvertexlighting q3map_nolightmap { map textures/Bootland/chains blendFunc GL_SRC_ALPHA GL_ONE_MINUS_SRC_ALPHA alphaFunc GT0 depthWrite rgbGen identity } } Problem: The sprite's texture twists and breaks off in a way, if I'm using a simple patch mesh (in-game). Here are some pictures to show what happens. In this picture, you see the chain to the left, a patch mesh. To the right is a sprite on an ordinary square brush. It has nodraw on the other sides but the front. This sprite doesn't follow the camera 360 degrees around, but stops after 180. The patch mesh follows all the way, but remains bugged. https://dl.dropboxusercontent.com/u/58757568/sprite1.jpg Here is a big match mesh with the same sprite, just in case it makes it clearer. That's supposed to be straight, and whole. https://dl.dropboxusercontent.com/u/58757568/sprite2.jpg

I've been trying to create a corridor where I want the lighting to be completely even. I've tried to do this by using light-emitting shaders, such as this: textures/Bootland/bricks2_emit { qer_editorimage textures/Bootland/bricks2.tga q3map_lightimage textures/Bootland/bricks2.tga q3map_surfaceLight 30 { map $lightmap rgbGen identity } { map textures/Bootland/bricks2 blendFunc filter rgbGen identity } } It does, however, not become even. Some areas of the wall are dark, some have round bright spots, as if a light was placed randomly around. https://dl.dropboxusercontent.com/u/58757568/surfacelight4.jpg Does anyone know a way to fix this?

Onysfx, haha would love that. Zelda (and mario with the never-ending stairs) had some great concepts. Zack!!!!! You've brought me back in business. The roll did the trick. As for the patch mesh, I have tried that a couple of times, but it leaks the mirror it seems. For now I'll just use a clip (patch mesh) to hold the player in. When trying it out in game after doing the roll-parameter, I actually scared myself. Like, I got chills, surprised by how it worked. Still need a few tweaks, but expect a little pk3 showcase of it for those who play JK2 here... soon enough! Thanks!

You mean a window-like reflection of a room? (As in, a jpg texture of what I want to reflect). eezstreet - I'll look up tcgen environment! What I'm creating is a sort of paradox in a corridor, where you walk through a mirror which is invisible on one side (but you can't walk back), and shows a reflection of the corridor if you try to go back. A reflection, as if the path had actually changed and going back would appear to be the same way as going onward. I've made a slight alteration in the idea though. I want to repeat this effect with two mirrors facing each other (from some distance). Don't worry, I'm aware that mirrors can't mirror other mirrors in this game. But I'm making a work-around using cameras. I have run into a new problem though (and quite silly it is). Problem: The camera the portal surface is pointing to shows the world up-side down. I tried using one earlier in a different part of the map, and it was normal. What I ultimately want to achieve is to have the camera in the height that makes it look exactly like the textures on the wall (etc) repeats along the camera view.