DopeFishhh

Well, perhaps magical torches? burn with a different flame colour for that eerie feeling. How about a lighter? I don't think it would be terribly hard for a hammerite or mechanist (or any other mechanically inclined individual) to create a lighter of sorts. They had the gas lamps in thief 2. Or you could make the lighter magical, though it wouldn't be much of a difference. Thinking about fire, would it be going too far to make a semi realistic fire system? where fire can spread (given fuel). I'd like to see missions where the player goes in to steal something and uses a fire as cover to escape or something like it. And it would be cool seeing fire arrows set the guards clothing alight and them running about screaming, trying to find water or something to put it out. or the player being traped in a burning building and needing to stumble his way through the flames and smoke to escape. The only problem is that if you were to make it realistic in terms of buildings burning down, well that obviously puts a huge strain on modeling a building, and could also unbalance the game with the player going pyromaniac on the mission. And IMO flames as currently rendered in games today are pretty crapily done, and from what i can tell a particle system that does fire justice would suck alot of cycles, so it would either chug or look like shit. Just for a bit of a giggle factor if the lights are out both the player and any human guards should have big trouble seeing and hitting each other, any sound should give the player (or the guards) position away (so guards try to be quiet as well). on the nightvision stuff, in Dungeons and Dragons there are a few things creatures/characters can have to improve their vision/detection: Low light vision - light is amplified. Dark vision - magical, can see in total darkness but only in greyscale. True seeing - magical, everything looks as though it's lit by candle light and there are no shadows, the characters eye's glow and no creature in line of sight can hide from them. Scent - obviously for creatures with a good sense of smell, ie dogs. but it could add that little bit extra to the game, like guard dogs, if you come close enough they will be able to smell you even if they can't see you and start seeking you out. Perhaps it could be possible to add those as something a mission maker can enable with specific magic items or on other creatures/guards/NPC's as default. ah shit i ranted again...

Well i haven't got the full specifics of the algorithm figured out but basically from each node you attenuate to an ai and record the volume of the sound. Which ever is the highest is where the sound is being heard from. This holds true in any situation as far as i've figured out. If an ai is in the original broadcast zone and a rebroadcast zone the original broad cast zone will be of higher volume because the rebroadcast zone has to travel further and thus attenuate more. On the issue of using frequency as a part of attenuation you could simply use an equalizer (like in winamp) for the attenuation field settings, and apertures have their settings linked to their size. the apertures would have rather dramatic settings that totally cut off the frequencies with wave lengths that exceed the aperture dimensions. of course this is a nice to have... I think that you are going to have to make a simulation system for the sound attenuation modeling as well. Mostly because if you tried to dynamically measure sound volume and then dynamically attenuate it and then dynamically measure the attenuated volume thats going to REALLY suck performance. You will need a system that assesses a waveform as it is imported into the toolset or loaded into the game. The assesment would be a heck of alot simpler piece of data to work with and thus save you quite a bit of time processing it all. When you attenuate the sound you record all the modifications it underwent and thus modify the sound the player hears accordingly. If you were to implement frequency specific attenuation then this system would be absolutely essential to not make the game chug. a few other questions: What does directX/openAl offer in the sound department other than a standard to work with? Will you implement a sound randomizer (ie multiple versions of the same sound or a system that dynamically alters the sound)?

well if you ask me saving/loading breaks immersion, unavoidable but true. the player gets shook up by having to reload because he died or did something silly or the game did something silly. look at movies for an example of immersion, the main character may not win all the time, however his defeats are not major, he rolls with the blows and moves on. saving and reloading pretty much means that the main character has control of time and can rewind back to a suitable point and redo that failure. now of course a movie is much more controlable than a game ever could be, and with games becoming more detailed and flexible controlling things becomes harder. A good thief mission would not lock a player into something, ie loot requirements. Making the player hunt around for loot is silly it should be optional, but taking the option up means the player can buy more equipment. a good thief mission will give the player an end objective and let him figure out how he wants to do it, the player cannot save or reload to any other point than the start of the mission though he can pause it indefinately where he currently is (basically if he chooses to pause it does a save, but he has no option to reload only to resume or restart the whole thing). finally if the player encouters something critically bad there should still be options left open to him, if he's dead there arent many options other than restart though it may seem a bit extreme but it would be the closest representation to what things are like in real life. it is also the most entertaining being the smoothest style of gameplay coming close to a movie in it's flow. because this is an extremely long thread already i might repost this in another thread for discussion. P.S. I don't necessarily think that the dark mod should do this but they could if they wanted to.

well if you ask me saving/loading breaks immersion, unavoidable but true. the player gets shook up by having to reload because he died or did something silly or the game did something silly. look at movies for an example of immersion, the main character may not win all the time, however his defeats are not major, he rolls with the blows and moves on. saving and reloading pretty much means that the main character has control of time and can rewind back to a suitable point and redo that failure. now of course a movie is much more controlable than a game ever could be, and with games becoming more detailed and flexible controlling things becomes harder. A good thief mission would not lock a player into something, ie loot requirements. Making the player hunt around for loot is silly it should be optional, but taking the option up means the player can buy more equipment. a good thief mission will give the player an end objective and let him figure out how he wants to do it, the player cannot save or reload to any other point than the start of the mission though he can pause it indefinately where he currently is (basically if he chooses to pause it does a save, but he has no option to reload only to resume or restart the whole thing). finally if the player encouters something critically bad there should still be options left open to him, if he's dead there arent many options other than restart though it may seem a bit extreme but it would be the closest representation to what things are like in real life. it is also the most entertaining being the smoothest style of gameplay coming close to a movie in it's flow. because this is an extremely long thread already i might repost this in another thread for discussion.

That sounds ok, but if my understanding of how it would be implemented is correct it might have issues. I guess the area of a wall/floor would have a rather strong loss characteristic to it. would you make the area sound characteristics dynamic? The funky part is where you have an open door the walls are 1 foot thick stone and a dude in the other room, without some sort of pathing system the sound would be nullified even though the person could quite clearly hear through an open door. But i agree with the ray tracing problem, it is similar to path finding. but how about sound nodes instead? Basically at key sound refraction points eg doorways open windows etc... you place a node, it will act like a proxy for the first emanation point of the sound by picking up sound intensity and re broadcasting it. it's a hell of alot simpler than ray tracing around obstacles and is much faster as well. the only drawback is that mission authors have to place the node in the toolset (unless you made it automatic some how). if you implemented the node system you could also alter the ai to go to the emanation point if it's what it is hearing from and go hunting from there.

well what about the echo? and will walls/floors/ceilings properly impede sound? people said that the thief series sound propagation was fantastic, it was but it also was quite inaccurate at times. and will the ai be a bit more realistic in its responses to sound?

a grid of bones within the material. when it simulates a cut the bones connecting points are cut and then gravity pulls on the free hanging bits. there is a maximum distance for each bone (ie between the two points it connects) gravity is constantly acting on each point, in more detailed material physics they factor in obstructions (like laying a cloth on a ball). bloodraynes system didn't look that great, you could only cut it on specific spots. ie where the bones were. this meant there were minimum distances between cuts, this was mainly to avoid the cpu having to crunch the numbers on 1000's of bones instead of 100's and still maintain a nice looking floppy material. i don't know how easy it would be to implement the physics for floppy and cut materials in the doom 3 engine but there's only one way to find out.

what about a karma factor (listen before you laugh) the way things tend to go in games is once you've got the technique or map down you can run through it without much challenge. if you throw something new at the player it only lasts as long as it takes for them to figure it out (if they are good then probably not long at all) what the karma system does (doesn't have to be offical or displayed on screen) is assess how the player has gone thus far, if they've done well then their score is a high positive number, if they've been scraping through by the skin of their teeth it would be a negative number (range +100 to -100). medium difficulty tries to keep the score at 0, difficult tries to keep it at -30, and easy at 30. how does it effect game play? the engine goes by the karma to nudge things infavor of or against the player. say the player is sneaking past a guard, if he's been doing well (positive number) the engine will slightly increase the guards senses to make it harder to sneak past and more likely that the thief is spotted. negative the guards senses will be dulled. if the thief is spotted thats bad, and thus his karma is lowered, if the thief gets into combat thats bad too so his karma is lowered. if he does well and say pick pockets someone without getting caught his karma is increased, if he snipes a guard or steals a nice piece of loot the same happens. the net effect is that through multiple plays of the same mission it may be near impossible to get through it the same way and without some sort of challenge. the mission maker can adjust the score and check the score, using them to do a more mission specific adjustment of the difficulty. the creativity of which is upto the mission maker, they can also adjust the difficulty settings so the engine based karma kicks in earlier or later. when the karma is high the mission maker could do things like: - send the guards on a surprise patrol - add an extra guard to a key spot - lock a door or increase it's lock difficulty (happens when the player gets to it) when the karma is low the mission maker could do things like: - make a guard fall asleep - send a guard to the kitchen to get something to eat - add extra random loot (when the player would find it) - unlock a door or decrease it's lock difficulty (happens when the player gets to it) this isn't aiming to make things any more difficult for a player or easier, just make it fair. play testing can determine if it needs to be harder or easier.

Well coming back to the argument that it makes things easier, that isn't necesarily true. You could give the player 1000 options about how he wants to do his mission, but if you increase the guard and make them *all* do patrols or searches for the player once alerted and they also shift their guard positions to cover some crucial spots then that would make things very hard. i didn't like the thief 1/2/3 approach where if you leave em alone for a while they will forget about you, even if you got into a knife fight with him and two of his buddies are dead. once you alert them, they remain alert, will get help to their spot, either from asleep guards or town guards, and effectively making things very difficult, next to impossible i think. that said if you guys have an agree'd upon design don't make major changes to it despite what i or others suggest, get the mod done first then you are in a position for requests.

well how about making it possible for a guard to try and block your blackjack attempt if he is aware of you? you can still blackjack him but it's no longer a certainty. And how about making it so that a guard can wake up after a random amount of time? and when he does wake up he runs about alerting others so things become even more difficult. The only way to prevent them from running about is to tie them up, or putting them somewhere so they can't get back to the rest of the guards, or kill them. that makes it easier to blackjack but makes 'jackin less effective long term.

i love the stealth genre and i think thief tops the list. but there are aspects that are either don't make sense or just plain silly. like guards instantly knowing that those footsteps he just heard was a thiefs as opposed to one of his comrades. An example would be in t3 in the hammer factory trying to steal the chalice, alot of guards and metal plating, you don't see them jumping about if they hear their brothers making a noise but as soon as you do they go nuts. basically it's an ai issue, but there are other things than just the sound. visually as soon as they've seen you they know you to be a thief, ok sure the bow and arrow pointed at their head might give garret away but there are situations where guards should be a little more human. likewise you don't see guards attacking each other the minute they see each other. so guards shouldn't necessarily see garret as hostile unless he's doing something suspicous or hostile. you could do a meter to measure how suspicous your behaviour is, if it's low you could walk right upto a guard and he wouldn't attack you but might start questioning you, if it's high you got problems, disguises wont work as well or at all, your only option is to hide. if you were to make the ai smarter/more realistic in those two aspects you could introduce some hitman like gameplay elements. garret could dress up as a guard and walk in unobstructed however this may get him only part of the way for a guard couldn't get into a vault himself. you could also introduce more roleplaying to the game with dialog choices, persuasion, intimidation, bluffing, diplomacy all options the player has in discussion (noting that whether they work or not depends on what the character is trying to do). now that my rant is over, are you planning to exceed the boundary of thief 1/2/3 ai and gameplay elements? or is this factor still in discussion?

so with the lighting are you going to leave the maps totally unlit save for the obvious lighting from a torch or electric light? and what about day scenarios? i know traditionally thief has been set in darkness but it would be a very nice gameplay element to have to beat the daylight clock.

One thing i really liked was the concept artworks for thief 3, the painted and slightly cartoonish look to them. also in the game i enjoyed the break in some parts from trying to get mega graphics power and put some artistic style in (like the moonlight through the windows). I guess the way i've seen thief is more about the art and the 'hey that looks pretty cool' despite it not having half a billion poly's. What artistic stylisations (if any) are you wanting to apply to the base content?