Search the Community

It happens so fast, no one will even notice if it cuts through, or you can cover the frog with the left hand to hide it. Of course, in reality what you'd do is hold the frog in the left hand and pull it away in the opposite direction to the sword, but the frog isn't set up to be animated in this model.

When you change that model path you must reload the map.

It should just be a matter of changing the cvar to the new path. @Sparhawk: Can you think of any reason why the lightgem still wouldn't work when the cvar is changed to the new model path? Does that #DEFINE have to be changed after all? @Oddity: Sorry for the trouble, but would it be possible to re-upload the broadhead model you did with the origin changed? That one was working great for testing the new projectile system.

That's always been there since the first version I uploaded. It's that way for a very good reason because I did a lot of experimenting, both with myself and with the guard model. If he tries to turn on the spot in only 2 steps, his legs are too close together and his boots intersect. I can techincally turn 90 degrees on the spot - when I lift my right foot, I need to place it at 90 degrees to my other foot so that the heel is almost touching my other heel, then when I lift my left I can place it parallel to the right one again. (I did this while standing over a tile pattern, so I knew exactly where my pivot point should be and where I should place my feet). This doesn't feel natural, and also I'm not wearing thick boots. Also, if you want to talk about "natural", try and "naturally" turn exactly on your vertical axis. I found that you naturally want to turn on an axis that exists somewhere behind you, if you want to do it in only 2 steps. So he steps out a bit wide with his right foot so that his boots won't intersect, and then turns, but his right foot is not where it should be, so then it moves back to where it should. In the end, I think it looks like a natural shuffle someone might do if they were just stepping around to face another direction, rather than a perfectly planned dance move.

What about the model used for the lightgem, was that moved? The reason I ask is that in my copy, the new lightgem is now always pitch black. Could it be failing to load the model if it was moved? [EDIT: Yeah, looks like the lightgem was moved. From the way Spar set it up, it looks like we just have to change the cvar "dm_lg_model" to the new structure, "models/darkmod/props/misc/lightgem.lwo" Then we have to change the defaults in the code when we get the chance]

I've tried programs like Terragen, but never really got into them, so to speak. I think my understanding of how erosion and stuff works is pretty limited. Also I have no tablet so drawing heightmaps is rather clumsy. So basically I'm wondering if anyone here is sufficiently proficient in any such program to be able to make me a landscape or two? I basically just need highpoly model + texture, I can make a lowpoly one pretty easily myself, and normalmap it, but the high detail just fails me somewhat. I want a valley primarily, with a river at the bottom and high hills either side, in which a city sits. The river should lead out to sea, which should be not too far away from the city, which should be overall in the centre. In fact, it doesn't need to be that large, because you won't see all of it anyway. It will be viewed from about halfway up one of the hills on one side, with most of the city below in front, and the sea over to the right, with more land to the left. And no, no city will actually be required, unless you're feeling incredibly generous. So, just wondering if anyone's up for that. Otherwise I'll do it myself and I just can't imagine how that would turn out. Just to stress: THIS IS NOT AN OFFICIAL POSITION, IT WOULD NOT ENTITLE TEAM MEMBERSHIP OR CONTRIBUTER STATUS, THIS IS MERELY A PERSONAL REQUEST, although if anyone does do it they'd be duly credited and probably invited to contribute more if they so wish. It's for my FM.

We already have a pretty comprehensive list, broken into reasonable stages. Any animations we come up with that aren't there can be added to the list. http://forums.thedarkmod.com/index.php?showtopic=1650

I just played some of the Bank level in Thief 2, and there are several grandfather clocks in it. They all have swinging pendulums while they tick, which is a nice effect. It would look kinda strange, like a very static 3-D model/texture, if it didn't swing while ticking. Considering T2's swing, I think it would look kinda bad if ours didn't swing. I think if someone can model it, it couldn't hurt to have a closed/ornate front one for variety. But I rather like watching the pendulum swing on clocks and think it'd be worth it to figure it out.

A question to someone with Lightwave: How can i easily create a poly if i have the points? i gotta cap some holes on this model and i dont want to bing it back to max.. (the low poly was old model and i didnt know of the shadow problems when i was making that)

nope you have to create a (cm) collision-model for it - and because it's a moveable this collision model must'nt have more than 16 triangles. I already created some collision-models, for example for the red dinging chair and the crates.

It's come up from time to time whether D3 uses the rendermodel for a collision model or tries to auto-generate one or what. For idMovables [EDIT: Should say static not, movables ], it looks like it does use the rendermodel. This might be okay in some cases, but we also might want to create our own simplified ones to help computation. For example, this chair probably could be pared down quite a bit around the legs. Although any simplification we make in the .cm will effect the collision physics (like when shooting arrows into the chair, stacking things on the chair, putting unconscious bodies on the chair, etc), so it should still be a pretty close match to the chair.

It's actually easier to do it manually than to redo everything else, I think. I'm fairly certain Oddity is the only person that would have updated any files in those folders. Oddity, do you remember uploading anything to the models/chars, models/md5 or model/weapons folder since the middle of December, except for the broadhead.lwo this week? If not, then I'll use my backup to restore those folders.

I just checked my logs. On 2nd of January you uploaded the banners. On 5th you started to create a lot of directories, so I guess this is when you started. So I can take the backup from 4th and replace it. The only thing that has been updated in the meantime was Oddity with a broadhead model. This will have to be resubmitted. Fortunately this means there is not that much damage done right now.

the headline basically says it: since about 3 days ago I can't access the cgtalk site+forums anymore. firefox tries to connect to the site for a few seconds than I get the "operation timed out" message, same thing with IE. I first thought they may have server-problems, but I asked some other guys and they don't seem to have any problems with the website. After that I tried it on another other PC and on a laptop here, which connect over the same router - they can't open cgtalk as well. Anyone an idea what's wrong?? They have actually no reason for banning me, since I probably have 10 posts per year and never had any argument over there - most of the time I'm just reading through the threads, which is quite interesting and I learned a lot over there.

Argh, we're going to have to change all of the def files to point to the correct paths now. None of the weapons load since the model points to the old path. Also, none of the AI work. So yeah, pretty much all the def files need changing.

Well, I also created new folders in the model structure that don't have any models in them yet, and I don't want those to be deleted as well. Maybe I should put a dummy file in them for now.

I'm slowly deleting the old models so that we don't have people updating the old ones instead of the new ones. It's a time-consuming process because (incredibly) there doesn't seem to be a way to erase entire folders from the repository. I have to do them manually. I also can't remove the actual folders, which is almost guarrunteed to cause confusion.

So yeah. I met him around new years, and talked a bit with him about the mod. Although he's a bit busy with university stuff, he'd like to help out. I know he can model, skin and animate. I think he works mostly with Maya, but also knows his 3ds. He put up some of his work here. Some of this is a bit older work, as far as I know the last thing he did was creating a virtual studio (for TV). He asked if he could have a smaller prop to model, just to see how things work out. So, what do you all think?

The current model is on the Y axis, with the head pointing up, and the origin is at the center of the shaft. I updated CVS with the tip moved to the origin.

Almost forgot, the origin MUST be at the tip of the arrow in order for this to work. The collision model for the arrow is based around the origin, and this is what stops the arrow in flight and tells it that it's collided. In order to stick arrows in by a variable amount, I need to know when the tip of the arrow hits, not when it's already halfway stuck in the wall. I think BT made a corrected broadhead model with the origin at the very tip, but I don't think this ever made it to CVS. Is it on the ftp?

While coding projectiles, I would like to write some code that pushes the arrow into the wall by a variable amount. The problem is, I need to figure out the vector to push along. I could just use the velocity, but the arrow is rotating a bit as it's fired, so the axis of the arrow is not aligned with the velocity variable. It's more correct and looks better to have the arrow push in along the direction of its axis when it hits. To do this though, I need a reference direction on the arrow to make my axis direction vector. This depends on the specifics of how the arrow is modeled. Most importantly, which local axis does the arrow axis lie along in the model? I.e., what local direction is the "length of the arrow," x, y, or z? Also, it's important whether the fletching to tip line segment lies along +x or -x directions. First of all, I don't have any modeling program myself; can someone open up the broadhead.lwo and tell me what axis the existing broadhead model is aligned on? Second, either we'll have to be sure to either always use the same axis for the length of the arrow and hardcode it in, or I can add a key/value to define the axis in the def file, so we can correct it easily in the def file. I'm inclined to go with the latter, since it's easier to correct one text entry than re-model the arrow, but I still need to know which axis the current arrow lies along.

Update: I've also had a go at the material files, creating some new ones and moving some materials around so they more closely match the model folder structure. There are still some materials I can't find, however, like those for crates.

Reposting this since the other thread went a different direction.... I was looking around the models we have available, and I noticed that there are models showing up in the editor that are not in the darkmod directory tree. I can select a lamp called "models/props/lights/lantern/squarestreetlamp.lwo", but when I look in our models/props/lights/lantern folder, there is no such model. So where is this model? I think it's one of Dram's, does that mean it's in a pk file somewhere? Shouldn't it be put in the darkmod folder? It seems like a number of models are missing like this. Without knowing where the models and their material files are located, I can't make additional skins for them.

Ok, as far as I can tell the uploading of the new texture organization is done. I've updated the material entries (since the props/textures folder was moved) and the skins files. *Hopefully* everything will now work as planned, but I'm prepared to find out I missed something. Please let me know if you find anything amiss. I started to remove the no-longer-needed original model files, but I can't seem to remove folders. Anyone know how to do that? I'd rather not have to go through every sub-folder to remove the files individually. I'll also post the new structure and guidelines for future models in the INformation section. One problem still remains though...there are models present in the editor that I can't find in our mod. Mostly Dram's models, which I imagine might be inside his pk4 file. They need to be put into the proper folder, but Dram will have to do that. So his models won't work until then. Any other models that were not in our original darkmod/models structure will not be in the new structure and may not work. Please report any problems to me and I'll look into it.

Brown dwarves are planets that didn't quite become stars. Jupiter is a brown dwarf (a brown dwarf is a gas giant planet that puts out more energy than it receives - a net energy emmiter - , which is exactly what Jupiter is), and Saturn is not far off it either. Perhaps we should modify oDDitys planetary inventory down to 6 planets, to exclude the two almost-stars... Perhaps you misunderstood me, but I was pointing out that by your definition, a planet wandering around deep space was not a planet, which is just silly. I didn't say anything about it having to be captured by another star, merely that it could be flung off it's parent star and drif though interstellar space, though you are wrong on that count too, the chances of any individual planet being flung from one star to another might be small, but the sample space is huge, so you can still expect quite a lot of them to be flung off one star and captured by another star. I don't think you understand probability too well. Anyway, the point is not how common or rare that event is, the point is that you shouldn't change the definition of a planet just because it happens to no longer be orbiting the star it originated from. A planet is a planet, whether it has a regular orbit, or an irregular one, or no discernable orbit at all. I prefer to define things by what they are, not where they are or how they happen to move. A planet should be definable without needing to put it in the context of particular orbital patterns, or particular stars, or its point of origin, which is what you are doing. Would the Earth cease to be a planet if it was knocked out of orbit from the Sun and flung far off out of the Galaxy? If so, what would you then propose we call it? If Mercury can be considered a planet, why exclude pluto? Sorry, defining something by its current orbit is just stupid. If Pluto was knocked into a regular orbit closer to the Sun would it suddenly become a planet in your view? That's not what most experts at NASA seem to think - the most probable explanation for the current state of Venus' and Uranus' rotational axes is that a planet roughly the size of Mars wandered very close too them at great speed at some point in the past and kept going, eventually being flung out of the solar system and into intestallar space beyond the heliopause. Whether or not this happened early in the formation of the Solar system or late is utterly irrelevant. The key is the huge objects whizing around - some of them will have no doubt been flung from the solar system, and there is no reason to think that some may have arrived from other sources. And with sufficient super-computing power you should be able to model what happend and when by analysing the current orbital patterns of the planets, which has been done already to a limited extent. My point is not erroneous, and you are the one clutching at straws by coming up with some convoluted, abstruse definition that is obviously tailored to specifically exclude Pluto as a planet, and serves no other purpose, and cannot be applied in a general way to any object regardless of its position, habit or origin. EDIT: I see from Max's post that Pluto does indeed pass your atmosphere test for planethood, certainly to the extent that Mercury does, so now you are really clutching at straws to scratch Pluto off the list of planets....