Page 2 of 2

Posted: Fri Nov 12, 2004 3:46 pm
by Subhuman
Mismatch wrote:Indulging in those activities however, makes you a cockogobbling, anal receiving, rimjob aquiring, donkeyfisting, mom raping(which isnt that bad), ball tugging pee-face.
So....everyone who played Thief is a mom-raping pee face? Niiiice.

Posted: Fri Nov 12, 2004 3:53 pm
by Blargh
ARE YOU BETTER OFF ALONE !? ARE YOU !? ALONE !? YOU THERE ! YOU !? ALONE !? BETTER OFF ?! WHAT ?

Ah yes, those were the days. They were.

AND THAT'S HOW THE WEST WAS WON ! :drunk:

Posted: Fri Nov 12, 2004 3:54 pm
by Mismatch
Subhuman wrote:
Mismatch wrote:Indulging in those activities however, makes you a cockogobbling, anal receiving, rimjob aquiring, donkeyfisting, mom raping(which isnt that bad), ball tugging pee-face.
So....everyone who played Thief is a mom-raping pee face? Niiiice.
grand innit.

Posted: Fri Nov 12, 2004 4:03 pm
by Blargh
You fucking pikey. :drunk:

Posted: Sun Nov 14, 2004 7:10 am
by Subhuman
Pikey blimey crikey. Pip-pip cheerio and a spot o' tea, guvnah!

Turn-based sucks.

Posted: Sun Nov 14, 2004 12:48 pm
by Blargh
By Jove ! A DIFFERENT AVATAR !?!? Jesusfuckingcockchrist !! Is this possible ? Surely not ! You're a real merry old soul, Subhuman. Yes. :drunk:

Posted: Sun Nov 14, 2004 11:37 pm
by Subhuman
I don't like changing my avatars every Tuesday like some of you. I get confused when people do that.

Hey, cool, Statistics is up and running.

Posted: Thu Jan 06, 2005 12:03 am
by RAK
A sports car isn't better than an SUV.
Yes it is. I've come up with a list of reasons why SUVs suck.

1: They are extremely inefficient in their fuel use. There's the hippy crap out of the way.
2: They are extremely heavy and unwieldy.
3: They have awful acceleration (The Hummer H1 Alpha seems to have an acceleration rate of 0-60 in 13.5 seconds, equivalent to a normal car, and that's with a 6.5L turbo-diesel).
4: They have awful acceleration for their engine size. Put that 6.5L engine in an ordinary car, and you might see 450hp or more. As it is, the Hummer H1 has only 200hp.
5: They can't corner well because of their weight.
6: They are used to knock down small children(I'm only messing).
7: They are driven by soccer moms. That in itself makes them uncool.

As I say: [Child's voice, on the point of tears] "Mommy, mommy, PLEASE go faster!"

Posted: Sun Jan 30, 2005 1:44 pm
by Mismatch
RAK wrote:blahblahblah
you are pointless.


Well, back to the real issue. about TB, I've posted this before, and I am certain that it'd work. Many ppl say that TB wouldnt wrok without hexes and a 2d engine. This is utter shit. It'd work.

In any 3g game, you have coordinates, (x, y, z). Let every character, when in combat, use theese coordinates to calcullate the length of a vector from when he is to where he wants to go (this is fairly easy), take the length of the resulting vector Þ and just divide it with a decided value (the value should be decided by how long you are allowed to move for 1AP), and the result will be the AP cost of your move.
Cost for other things such as shooting and chit can be any fixed value, as in FO1 and FO2.

Posted: Sun Jan 30, 2005 11:05 pm
by Sovy Kurosei
I prefer the hexes model, when concerning movement. It better emulates people taking up space in their zone as opposed to using collission models which still appear to be wonky at times. There was one game that used vectors in determining AP costs, and that game was called PTOII, a game by Koei. What ended up happening was that I could abuse the system to make my fleets go further simply by taking advantage of how the calculations rounded out to. You could increase the amount of APs you have and make AP costs round up to get around that, but you will still wind up with the underlying problem of pixel hunting in order to get the most movement for the least AP costs. I guess you could make something as tedious like that pointless by jacking up AP points up to a 100 or some other large number.

Also, when using hexes, I can see an improvement in AI performance (for your party and for the opponents you may be facing) because you are reducing the choices that the AI can make, or having to rely on heuristics which can bug up badly at times.

Posted: Sun Jan 30, 2005 11:23 pm
by Mismatch
I see, valid points.
Hexes'd prolly be better, but much more of a headache to implement I'd wager.

However, AI shouldnt be affected that much. The AI quality is prolly more of a result of design rather that if you use hexes or not.

are there any more options?

Posted: Mon Jan 31, 2005 12:33 am
by Sovy Kurosei
It would be more difficult, especially since every modern game engine relies on vectors. Likely what will happen, if you wanted to include hexes, is to emulate it into the game, making a map then layering the hexes over as opposed to Fallout 1 & 2 where the design was based on hexes.

As for the AI, right now in real-time combat most baddies simply charge right towards you, or do something completely rudimentary like take cover behind the closest obstacle, even if it is a barrel of explosives. On the other hand, with a TB system the computer has the necessary resource (time) to calculate the move of one character. I've been thinking that the pathfinding would be more reliable with hexes and (I hope) faster. And that is another thing that I am wondering, how will the computer be able to handle obstacles and pathfinding reasonably well in a "vector" system like you described without it ending up being a simple rush towards the player character or the enemies doing some rather strange and erretic path finding in order to get to you. You have to balance between the AI taking up more resources for better results.

But this is 2005, not 1995, so maybe computers can get away with the vector approach quite well on today's machines.

Posted: Mon Jan 31, 2005 10:27 am
by Mismatch
As for the AI, right now in real-time combat...
true, but lets not speak about real-time. I was merely saying that AI shouldnt be affected if you choose between turnbased with hexes or vectors.
nd that is another thing that I am wondering, how will the computer be able to handle obstacles and pathfinding reasonably well in a "vector" system like you described without it ending up being a simple rush towards the player character or the enemies doing some rather strange and erretic path finding in order to get to you. You have to balance between the AI taking up more resources for better results.
Offcourse this would depend alot of how things are implemented by bethesda. But say an enemy stands X feet away, with ooze between you and him (In FO2 sulik ALWAYS ran into the ooze), I suppose that for non combat purposes, some sortta check would have to be done at each 'tick' to determine whether you or someone else is standing in the ooze. To make such a check you would either let the ooze be an object with a ooze flag set, or the ooze'd be a ststic part of the enviroment with just a certain ground type. Naturally I'd preffer if the ooze was an object, prefferably placed in a container, call this container 'hazardousGround'.
Theen, a simple call to the hazardousground container could be made:
hazardousGround->getClosest(x_Coord, y_coord, z_coord);
/* this'd return your closest hazardous ground in preffered forn
mebbe as a 2d plane in a 3d space.
then you could just calculate whether your closest path will intersect with this plane or not, a rather simple calculation.Pros with this is that you only need to consider a limited area, and thius it is more efficient. The drawback is that you only consider the closest obstacle. Another way would be to simply call a plotfunction with your position and where you wanna go, this algorithm could return the first move you need to do in order to get where you wanna go. Only the first since you'd need to call this once per turn. Mmore innefficient yes, but as we said, efficiecy is not an issue while dealing with TB, I can imagine a calling sequence summit like this pseudocode: */
move(plotRoute( getPlayerPos(), getMyPos()));
/*where move takes a possible move as data, getPlayewrPos returns the position of the humanplayer who is the enemy for the agent calling this function, and getMyPos returns the agents position*/
there are prolly 100s of more ways, and many mnore efficient, this was merely an example.
cheers.

Posted: Tue Feb 08, 2005 12:07 pm
by Plank
An answer to the real-time (not that I really want real-time) combat question would be to introduce a Soldat kind of way of shooting. It's like
Baldurs Gate real-time except you hold down the button to keep your gun firing, or repeatedly press it if it's a pistol or small rifle or whatever. Same goes with first-person.

However, with real-time in any perspective, the actual length of combat would need to be extended by giving enemies lots of health, or lowering the damage of bullets, so that the fight would actually last a decent amount of time and be fun and tactical (just like in F1 and F2), as opposed to say, Counter-Strike, where it's basically one shot one kill.

Enemy AI has to be good for this too work. They can't just stand there and shoot at you, they need to run and leap around, hide behind corners, throw grenades out, and do basically anything to survive. We're talking about whats likely to be a very advanced engine so it'd be possible.

lol

Posted: Fri Feb 11, 2005 11:33 pm
by popscythe
Do you guys remember morrowind? We're talking about the company that made that "Game". The one where as far as I could tell, the point was to walk down a never ending path full of really big worms. If you read the guide at this point because you are tired of the path o' squirmyness, you have torn the spacebar off your keyboard and replaced it with the drinking bird like homer simpson, because everyone knows that jumping forever will eventually make your character be able to fling himself seventy three feet straight up in the air. But that's not the point. the point is that first person "rpg" games are flawed in that AI is... not real. The enemies of every game are scripted to do certain stuff, and for the most part in morrowind you could just stick out from behind a rock slightly, or get on top of something, or elsewise avoid the guards and then shoot them with arrows for all eternity until they fall over and you get all their stuff, then do that again in every town along the wormathon racetrack. So if fallout is like that, where the basis of getting cool items is shooting someone in the arm for half an hour because they weren't programmed good enough... Give me hexes and save yourself the trouble. See... improvement isn't improvement if you take an already good thing and make it stupid, but more glittery. Why do they have to change it like the seem (from press statements) hooked on doing? Or is the ability to make this a game that your average xbox junkie is going to buy going to corrupt the devs so bad that we get Fable again? FUCK. The very thought makes me want to take 16 super stimpacks and then rest until morning.

Posted: Sat Feb 12, 2005 11:44 am
by Mismatch
well, the trouble isnt that noone wants hexes, I suppose most of us do.
however, I see a lesser probablility of hexes being implemented in a game using a 3d engine. and so a discussion about different solutions to the AP problem came to be.

one thing should be mentioned though when speaking about ai. the enehmehs should take cover more often, step out of cover, shoot, step into cover again.
ah, and lobbing grenades'd be nice, not just throwing.

Posted: Tue Mar 01, 2005 5:42 am
by Spunky
I'm not necessarily arguing for or against turn-based here, but I don't see how TB has any impact on exploiting programming limitations, at least the ones mentioned here. Certainly not using Fallout as any kind of model for exploit-free programming :) The enemies in Fallout pull pretty much the same stunts of any FPS AI: they'll stand and shoot until dead, and if they lose line-of-sight they bee-line right for your position. And I don't understand complaining about exploits in a single-player RPG anyway. It's not the game developer's responsiblity that you ROLE PLAY your character properly.




In my opinion, the value of turn-based combat in Fallout is more about reproducing the feel of a P&P role-playing game, and the pacing is just as important as the detail. When you're playing a live RPG with friends around a table, everyone including the NPCs act in turns and the GM describes every action sequentially. All attention is brought to individual movements, and you can sort of piece the events together into a continuous 'cinematic' moment with your imagination.

Personally I think that is why many primitive CRPGs of the 80's can still be so much fun even next to today's examples, at least for the imaginative/fantasizing types.

Computers are arguably able to represent an equivalent level of atmosphere and detail on the fly, but many people probably feel "robbed" of the imaginative aspect of P&P. On the other hand, it potentially allows the "digital GM" game designers to share their own cinematic vision with people, making CRPGs more of an interactive artistic experience.

One approach that's come to my mind is to have all actions declared in turn, let the computer take time to resolve and somehow sequentially script them out, then play everything back at once in real-time. This would of course take a tremendous amount of work on the programming end.



As far as working out APs in a real-time system, what problems would remain if all distance-related calculations were rounded off to a predetermined interval? This way your character would still move in set increments of distance equivalent to one "hex." For instance, if you set each step at 2 feet, clicking on the ground 10.3 feet away would move the character exactly 10 feet and cost 5 AP.

If it is a point-and-click movement model the cursor would indicate AP cost like current Fallout games, whereis with keyboard-controlled movement a single key-press would move the characters the minimum distance (ie, 2 feet @ 1AP). When the character moves that distance, the system checks if the key is still down and if so the character continues moving, etc.

For representing sequence, everyone's AP could just start at zero, and you don't get any until those ahead in sequence spend a 'turn' of AP.

Posted: Fri Apr 22, 2005 12:42 pm
by Mismatch
In my opinion, the value of turn-based combat in Fallout is more about reproducing the feel of a P&P role-playing game, and the pacing is just as important as the detail. When you're playing a live RPG with friends around a table, everyone including the NPCs act in turns and the GM describes every action sequentially. All attention is brought to individual movements, and you can sort of piece the events together into a continuous 'cinematic' moment with your imagination.
Second that.
One of my favourite parts with the FO turnbased is the aiming, ever since playing the fallouts Ive always missed the ability to aim properly in RPG's.

Someone posted a link to a really interesting article about this subject (Which bethsoft should read), but i cant bloody find it.
(If anyone know which article Im talkin about, please do post the link.)