killhour
-
Posts
125 -
Joined
-
Last visited
Posts posted by killhour
-
-
It could be made like a ragdoll. Just use the same process you would to make an animated character, just put it directly in the ragdoll mode and use a differente skeleton, obviously.
Good idea. Does the skeleton have to be in a hierarchical format, or can I set it up like a grid?
-
I'm trying to create a banner that moves out of the way when you walk through it. Is this possible?
-
But if you're spending that much time and not exploring the rooftops, perhaps you need a refresher course on vertical space... Granted, 95% of first and third person games don't make good (or any) use of vertical space. Even those that do (Assassin's Creed, for example) don't prepare players for a game like Thief, because no effort is required to reach and traverse rooftops, and there are no interior spaces to explore.
That's the problem. I read the thing about leaving the lab window open, so spent ages looking for a way to get up there. Why would I think "Oh, I have to climb this seemingly pointless wall that I've been trying to jump up onto from the dark side, but can't because of this tiny ledge, so that I can get into a window of a house I really don't care about, since that will OBVIOUSLY give me a way to get on the rooftops to get into the lab"?
See where I'm coming from?
-
I might be alone in this, but I didn't really like this mission. It's extremely difficult to figure out what to do next, and it just seemed so... linear. I spent about an hour and a half walking around the streets before I got fed up and turned notarget on and found the window like 20 seconds later. Some of the scenes were great though, very... unsettling.
-
I think they already can tell the difference. That's as easy as setting a different flag on the body.
@killhour: Having an AI drag someone out from under something and then roll them over to a position where an animation could take over to get them the rest of the way up would be extremely difficult. Don't they still have contests to program robots to pick up and place simple objects and manipulate them around obstacles? And you want to add this spatial/physical reasoning to our AI to have them manipulate human bodies around/under couches and things in the game world? Be my guest if you want to do it yourself, but we have more than enough work left just meeting all our existing AI goals.
Robots in real life have to follow the laws of physics. I'm just suggesting dragging a ragdoll like the player would. Just have the AI walk to an open area and the ragdoll follow.
I might give the AI a whorl at implementation. Also to the guy that said you don't wake up from a blackjack - perhaps that could depend on how close to the "sweet spot" you hit them. Also, you should wake up from a gas arrow if someone kicks you. I guess the main problem is the animations, and I'll try to give a system for that a shot.
-
Whether the gaurds wake up by themselves or someone wakes them up, the same gameplay and coding problems (described in numerous threads on the issue) remain.
Not necessarily. Gameplay wise, it actually penalizes you less for using a blackjack than the current method. Currently, if you have no place to hide a body, and a guard patrols that area, you don't have many options. With this method, if you can come up with an 'excuse' to why the guard is unconscious (Fell asleep in a chair, sleeping in bed), the game takes that into account and doesn't raise suspiscion. It penalizes you for being stupid and leaving bodies in the middle of the floor, which obviously should be penalized. You never have to worry about some arbitrary time limit on blackjacks, and it makes killing them even more undesirable (if a guard tries to wake another to find out he's dead, there's gonna be hell).
Secondly, the coding becomes a lot simpler. Specifically, if a guard sees another guard and is able to get to them, it's a safe assumption that the other guard can be woken up without too much hastle in regards to being lost. As for the animation, the ragdoll issue can be solved by having the second guard roll the first over on his back before he wakes up (probably by kicking him or something). If the guard is under something, and can't get up, the awake guard can pull him by the arm or leg to a sufficiently open area. This allows the state of the unconscious guard to be 'zeroed' and turns a practically infinite amount of animations into 3 (dragging, rolling, and getting up).
At first glance, it may seem to be the same suggestion, but it simultaneously solves 3 problems with the origional - it makes more sense, it promotes better gameplay, and it eliminates most of the nasty fringe possibilities.
I'm not saying that it's definately the direction you want to go with this, but it at least makes the idea more palatable.
-
I totally agree that guards waking up on their own from a black jack is stupid, BUT so is the other guards assuming all blackjacked enemies are dead.
Here's a proposal for a flowchart of reactions.
Guard Unconscious | Spotted by other guard |----------------------v------------v | | | Lying on ground Slouching in chair Lying in a bed | | | v v v Rushes to side Yells to wake up Nothing and wakes guard | |--------------------->| Guard was suspicious Guard not suspicious when knocked out when knocked out | | v v Searches for thief Guard gets yelled at
This leads to guards waking up other guards if they think they should be awake, and searching for the thief if there's reason to be suspicious. Guards slouching in a chair would just get yelled at to wake up, as it is assumed they just fell asleep. Guards lying unconscious in beds would be left alone, baring an alarm of some kind, where guards in bed should be woken up anyways.
-
I just reworded it for Tels' sake, I'm actually leaning against it. It's a rather vague situation that I don't think anyone would think to take advantage of.
I used to work in BASIC and Visual BASIC myself, and I know how tedious it can get setting up multiple exceptions to those stinking If-Then scipts.
I've never seen so many "ElseIf"s all in a row!
Case functions are your friend.
-
That's probably not a great idea, unless we don't want AI to every carry anything.
We already have a few options for how to address this problem, it just hasn't made it to the top of the 'to do' heap yet.
Make them suspicious of floating objects, and check to see who's carrying them. If it's a friendly, they should remember what (s)he was carrying and no longer be alerted by it. This could be accompanied by a "Who's that?" "It's just me." conversation that calms guards down. Smacking a servent on the head and sneaking past the guard by hiding behind the wine barrel he was hauling would be pretty cool.
-
What technique would mappers recommend here?
I have a street that is made up of three brushes. The outer two slant towards the middle slightly. I'd like the street to turn at something other than a 90 degree angle, but I'm not sure how best to go about lining everything up. When I try to use the clipper tool to cut the street at an angle and then rotate other brushes to match, it never seems to line up properly. I figure there is probably an easier option I'm not familiar with to make a street like this curve?
When you use the clipper tool, you have to make the same clip on both pieces before you try to connect them. The cut side is going to be longer than the uncut side, and the cut should be half of what you want to turn it by. Remember to use the "snap selected entities to grid" option to make sure they line up properly.
-
Indeed
Even if they don't have normal maps we can create these easier than creating diffuse+normal. (Also, if I understand that correctly, the textures do have heightmaps, and creating normal maps from heightmaps is even easier than creating them from scratch).
So I see only positive things for us and would like to thank pataran for doing this!
Agreed. Again, we all appreciate the effort and I'm sure we'll find some use for them. We were just worried about the copyright issues. Are we allowed to modify them? Some of them may need tweaking to look good.
-
<br />Again you fail to read my previous posts. Those normals didn't come from them, I said I didn't understand how to combine the greyscale images when I took those pics. So I used just the regular "normal" button to make the detail.<br />
I understand what you said, I'm just saying it's strange for them not to come with normals, especially since they seem optimized to be used with shaders (no baked shadows). This means that either they were lazy, didn't know how to make them, or don't understand that those textures would look bland as hell as just diffuse. As a first attempt, however, your normals are quite good - just far too deep.
-
;I do appreciate the effort you put in. You don't need to get defensive about things other people made, though. I assumed that those normals came with the textures. Why would they charge for textures that don't come with normals or spec maps? They're alright, I guess ugly was the wrong word. They're just far from professional, and seem to be low resolution. I'm sure we all appreciate the gesture, but the license is too restrictive for us to use.<br />Well, Mr Killhour, those ugly textures looked that way because I stated earlier that I couldn't figure out how to use the height maps with the textures yet. They look way better now because I got crazy bump to do it for me. Please read what people are talking about before trying to flame. The errors in the textures were originally from me. <br />P.S, I was trying to do this for YOUR benefit to make darkmod have better textures.<br /> -
Those textures are ugly as hell anyways. They don't know how to make proper normals, and the spec is cranked up WAY too high. The rug ones are probably the only ones worth using.
-
Update:
http://shadow516.googlepages.com/killhour.zip
Lots of stuff kept me away from mapping, but I'm happy to be back to it.
-
<br /><br /><br /><br />It's the func_fracture entities, which are crashing your map during spawn time. A surface pointer is NULL somehow, which obviously should not be the case. <br /><br />Has this ever been working before? Maybe some textures got changed on your end lately?<br />Ah, I know what it is. Thanks. I made a new surface shader (included in the zip), but I guess it screwed with something.
-
Every time I try to open my level with Doom, it crashes on the load screen. Any ideas?
-
It seems that when you select a face and go to brush>create decal patches,
it creates a copy of that face as a patch. Put any decal texture on it, lets say ivy vines, and now your wall will be covered with ivy vines.
If that makes sense.
And feel free to correct me if you know about decals.
I tried that, but I can't get it to work. Mind making a video of it and uploading it somewhere?
-
How do you use decals in DR? There's nothing in the wiki about it, and I can't figure it out from other maps.
-
Yo, you totally took my idea for using that chest texture as a door.
Awesome.
-
I remember an early test of this on D3W that showed a city map made of massive skyscrapers. There was an imp on the other side of the map that was so far away that you could barely make him out.
That doesn't really say much anymore. /newmap 16 in the cube 2 engine is just over 3x3x3 miles.
-
I explored this issue extensively a long time ago, forgot some of the details, but one thing is that it doesn't like single HUGE room, but if you make smaller rooms (still large) with visportals then you can make your map extremely crazy huge.
How huge? Oblivion huge? GTA huge? /newmap 16 huge?
-
-
Okay, just for kicks I've worked on this map in DarkRadiant for a while. Some stuff is still messed up, some parts I messed up (some texture problems), and I never did get a couple of the vis-portals to work yet, but I felt like stopping. Here is my version:
http://www.shadowdarkkeep.com/files/kh1_komag.map
I reworked a lot of brushes and things, for instance I made the ground brushes only where they need to be and trimmed in the big sky portal brushes to avoid having any sky portal texture exposed to the void. I shrunk the size of most of the lights. I lowered the subdivisions on the balcony posts way down and reduced the number of them. Something I did made the map compile MUCH faster now, so that's good! Plus I get somewhat better performance, such as only going as low as 6 instead of 3 FPS, but it's still pretty low.
It appears killhour that when building, you first slap down big solid brushes, and then you do CSG Subtract within them to make rooms and such, is that right? This is a very bad way of building because it creates tons of weird long sections of brushes which get multiple light counts, extend into multiple vis-portal zones, etc. I "normalized" some of the outside walls to attack this problem a bit, but I see it's just rampant all throughout the house, ever wall and floor and ceiling is all broken up into weird long brushes and brush "splinters" that extend across multiple rooms and floors. I suspect the engine doesn't like that stuff.
After working on this for an hour I see that you have an interesting mixture of cool creative approaches and newbie crappiness! The newbie part will wear away, so it's clear to me you can become a great mapper over time.
Wow, you did a great job optimizing this. It takes a fifth of the time it did to compile. Mind if I continue working on it from yours?
Does TDM support soft body physics?
in TDM Editors Guild
Posted
Yeah, the only problem with that solution is that there's no way to stop two columns of bones from moving in totally opposite directions, stretching the model ridiculously.
That, and I'm not particularly proficient with any modeling programs.
I'd love to hear some input from whoever set up the ragdolls for the AI, if at all possible.