crakkie

I googled around and found a few. http://www.centralpictures.com/ce/tp/paper.pdf An academic paper on building procedural cities. http://unreal.epicgames.com/Fire/AnimatingTextures.htm Unreal's procedural textures http://www.planetside.co.uk/terragen/galle...=8&ind=1&group= A somewhat impressive procedural landscape design program http://mrl.nyu.edu/~perlin/doc/hypertexture/ Noise, from Perlin http://www.gamasutra.com/features/20010302/oneil_01.htm Procedural planet generation http://accad.osu.edu/~smay/al.pdf A procedural modeling and animation language http://liris.cnrs.fr/docs/RR-2004-008.pdf Putting cracks in your models, procedurally of course http://www.devx.com/Intel/Article/20182 Procedural content creation overview Initiative! That's what I need more of. Hope this helps, nik.

How does 'reduced processing' mean better looking textures? The textures in this game are as good as most games out right now. These bricks, for example, though they are a little too specular. Things like faces and paintings on the walls would be difficult to model with this system, but for wall and floor textures, this is great. The most interesting thing about this game is that the frikkin level geometry, models, and animations are procedural (and they have a demo that has procedural voicing). The sound and music aren't as hard to imagine (the sounds sound like fm). If one can create level and model geometry this well algorithmically, then you can possibly create random levels (with mind-numbingly complex heuristics) that would look this good. Random towns you would encounter on a map would be different every time you played a game. In a galactic exploration game you could find a new world every time you landed, like you would if you were actually exploring the galaxy. Hell, the entire game might be different every time. This would be a lot of work, yes, and much more research needs to be done on this subject. Occasionally your generated medieval town might be a big cube with dodecahedron buildings and your generated aliens might be blue puddles that only talk about cheese, but if it actually worked reliably...that would cool. This is something I've always wanted to do in a game ( my dream job -- random-procedural content generation, or so I call it ), ever since I played Planet's Edge. Nice to see it's being done so well without me. Do any Dev's know about research on content generation? Any papers or articles? I'd appreciate it.

I think that's the point. Dummy terminals offload the work to the server. Dynamic websites do the same (like this forum) and only make the client worry about displaying the end result. The XAML/Avalon/Longhorn thing is to have network applications run on the client. The .NET framwork already defines most of the functionality and Avalon defines the interface, so the applications are very thin. This means the applications run in realtime, transferring only the data needed to interface with the server (as opposed to html/js/css apps where you need to send a request and get a page back each time you want to transfer something). This is very cool. The only framework like this on linux is J2EE or E17, which is perpetually unfinished. Yes, it would have been unevil of them to make it an open standard, but at least they aren't sueing Mono out of existance. For now. MUAHa. Ha. ha. Also, Longhorn/.NET uses C++, Managed C++. You get all the speed benefits of C# with all the elegant clarity of C++!! ( << sarcasm ) ugg. What was this thread about again? Oh, books. The C++ Primer by Lippman is good. Thorough, like Stroustrop's, but reads much easier, in my opinion.

Well, they did make their own 3D engine for Jefferson/Van Buren. As to how fast it ran, how flexible it was, etc., I can't attest (maybe no one ever will). It sure was pretty, though--from the 4 screenshots I've seen. Not sure if Lionheart used an in-house engine.

Haskell is a screwy functional language and Plankalkul was the first high-level programming language, spec'd by a nazi scientist and not implemented until the early 70's (by which time it was obsolete). Plan - Kalkul : 'programmed calculus' or something to that effect (if anyone knows German...) Smart ass, I was being As for UML, I know there are some programs to turn your UML model into emtpy methods (VS.NET Enter. Archi. does this, as well a a couple of open source tools), saving you the trouble of writing them down yourself. Seriously though, I think the point of most Software Engineering is just to make managerial types nod. You explain something over their heads with vaguely familiar and natural-sounding words, easing them into a steady, smiling nod.

Everything is a competition Sounds like a challenge to me. I'm learning Forth as we speak. Then I'll learn B, and Haskell, and AspectJ, and Plankalkul. Then I'll show you.

Wow, 2 people who know forth on the same forum. As for me, Pascal, Delphi, C, C++, Java, Perl, sh, ML (the little I remember from college), HTML/XML/Make if we're counting that. I feel awfully boring. Any advice on how to get a development job, anyone? I've been stuck in IT for the last year, but I have a CS degree and a math minor. Need job. Am willing to work for less than you if you don't live in India or China or Russia or Indonesia.

'UT depends on the skill of the model I'm using' 'Your're wrong because Baldur's Gate requires the use of one's hands' Truly, your insights smash my logicks to bits.

They already have a standard for audio/video systems, I mentioned it already. It's called Simple Directmedia Library. It's a cross-platform library for abstracting the audio and video systems, just like directx (except for the cross-platform part). It's what Bioware used to make the linux port of NWN. It uses openGL for 3D, so you would be running into the same low-level problems FrankK was describing. And thanks for the reply, FrankK

I'm not trying to make RPG's slow enough to accomidate those of us with only two fingers. Well before you start playing UT. The difference between letting the character perform actions and taking control of performing these actions yourself is the latter is an action game. Sorry for the technicalities, but this is simply what an action game is: you control the action of the character and you must you use your dexterity, timing and reaction. If you make an RPG-like game where all the skills are controlled by your dexterity, it's an action game (Deus Ex, Morrowind). In all the games you mention (except Morrowind) you do have the option to pause at the end of each round, negating your need for quick pause-button reflexes (and turning it into a crappy hack of turn-based). I don't want to get into a "What's a real RPG" discussion here, we're OT anyway. Start a new thead if you want to keep on this idea.

Like I said, the market share's growing and it's almost as popular as Macs. Epic released the Linux demo of UT2004 the day after the Windows version. Anyway, the dev's don't seem to be answering. No such luck from Obsidian, eh?

I've read that in the movie, Pilate doesn't want to kill Jesus, but Jewish clerics convince him to. Several times, I've read this. Yes, the Jewish priests did want him dead. So did the Romans. He was advocating politicaly dangerous things. No one knows how the decision to have him crucified was reached. "The Passion" has decided to blame it on the Jewish clergy. Those friendly Roman fellas wouldn't have crucified some guy just for proclaiming himself the son of god and king of the jews. Oh, wait, they crucified people for stealing bread... EDIT : and in response to the original topic, yes I'm probably going to go see it. If it pulls in enough dough, Mel might make a sequel. Jesus 2 : Easter's Over, Baby

If they are doing KotOR2, then they'll probably just do like Bioware and use the Star Wars RPG system (d20). I don't know why people use a system as "die" oriented as d20 for computer games. 2nd edition rules were much easier to understand, they were just tedious as all hell when doing them by hand. Shadowrun's unprogrammable "d6" system is one reason it hasn't had a good CRPG treatment in the last decade.

I never suggested that convoluted simulations be put in place for everything. Some things work just fine as is and complicating the simulation further would be neither noticeable to the player or necessary. But what's the excuse for not having projectiles or fireballs or whatever that move according to real world physics and bounce and ricochet accordingly? What's the excuse for using die rolls to check if the PC is silently hidden in shadows when computers can calculate line-of-sight and light sources and how sounds moves through an environment? Why break time up into rounds if it could possibly be avoided? Basically, all I'm saying is that I think CRPG developers shouldn't be afraid to pick and choose from the ruleset, ditching rules where another algorithm could produce better results. That, and I'd really really like to see a good CRPG in true, full 3d with all of those neat things that true, full 3d worlds can do in modern video games. Sorry if my example was crappy, but it still sounds like you want to replace the RPG rules with simulations of the environment. Having algorithms check for hiding in shadows, critical hits, how much damage was done, where bullets ricochet, whether someone was knocked over, and how well they hear is extremely difficult if you don't use random values. My convoluted example was pointing out that if you want to see how much damage was actually done to someone, you would need a fairly complex model to test that. Or you could use random values (or static ones) to gauge the damage. In that case, you're basically using RPG rules. The same goes for the rest of the checks. If these systems have a solid threshold, they'll seem less realistic. i.e. "I know that if a guard comes within 2' of me he'll see me, no matter how dark it is." Also, how would these skills increase over time (a rather big RPG thing) if you're just using the mechanics of the engine? If you're bouncing fireballs around a corner, it sounds like you're playing an action game. RPG deal with tactics, they always have. Skill with a mouse and gauging angles of deflection properly aren't really an important part of playing a character. Your CHARACTER has the skills, you just tell him how to use them.

Unfortunately, convincing physics and representational hit detection still can't simulate a reality, which is what rpg rules try to do. You could have every section of a model's body cordoned off to check where the sword struck, which organs where hit and how badly, check how the force of the blow would unbalance the model, check if an artery was hit, if it put this person into shock, so on and so forth. All of this built into the game mechanics. Or you could use RPG rules. 1-6 damage, critical hits, stunning, knockback and such. These are well defined rules that are already pretty much balanced and just need to be put into place. Don't get me wrong: I'd love to see the former, but I don't think it will happen anytime soon. It seems like a pretty mammoth project in itself. Maybe a good idea for a game-resource company, like Havoc. Reality Game Systems Inc. All I need are a bunch of people who know how to do it and a reason for them to give me some of the money. Any takers?

There's been a lot of work making Linux a useful gaming platform. Nvidia and ATI have written opengl drivers for linux. SDL does most of what DirectX can do (except 3D). Civ2 and NWN used SDL to replace DirectX in the linux ports. A lot of games work on linux, but they're emulated and run slow and shoddily. Loki tried to push Linux gaming way too soon and folded. With a growing market share (3%, Macs in comparison are around 5%) is it a good time to start treating Linux like Macs as far as porting viability? Does writing a game for PC+consoles make this more difficult? (guess I'm asking dev's here)

I realize (as Feargus said) that Linux will never be more popular than Windows. If, however, it were to have 5-10% market share at some point, would it be worth making ports for it, as Bioware did with NWN?

I heard the flogging scene goes on for more than 10 minutes. And it probably will come to a city near you... it's playing on ~2800 screens (almost as much as The Last Samurai and Master and Commander) the jeebus phreaks are going nuts, I'm sure here's one, she thinks god made this movie

No, you're not the only one. When movies or novels don't give me closure, it's closer to real life(which rarely affords those kind of endings). Take Fallout for instance: in the multiple endings, past events and choices manifested themselves in ways that were certainly not inexorably moral. By and large they were indifferent or cruel and left you feeling cheated and disallusioned. I love this, of course I'm an "emotional masochist" my girlfriend keeps telling me.

Labadieville, Louisiana, USA. A sprawling metropolis covering dozens of square feet.

It's Q - & - A, a question and answer session between the developers and their cats. Cats must answer correctly with one 'meow' for no, and two for yes. When one of the felines misses more that 3 questions, it is put down. It's a vital part of any game's development, and no game can be truely 'good' without it. Example : TOEE. No cats were killed during its production.

...err, I suppose the original question was actually asking how one would model a game consisting of quests and events which, while linear in and of themselves, are not necessarily guaranteed to happen in any set order. I've seen pix of Rockstar's studio during GTA3's production, and they have a large board with a lot of post-its in a tree-like structure mapping out all the quests in the game. I believe they mapped out the stories and plot arcs based on the locations in which they occured, as opposed to the point in the game where they might happen. I imagine you could organize the plots/subplots according to where they happen and what their prerequisites are. If you kept good records of a quest/events prerequisites (isEvil, killedThisGuysBrother, haveBeenHereBefore) you could keep locations heavily layered with quests (Like Tarant) rather well organized, I imagine, even if they came back to the same location several times. I'm trying to write a text-rpg myself. I use openoffice/draw to model the locations' quests. Just ovals w/ brief descriptions branching out from rectanular locations. As far as the main story arc's paths, you could just use a tree to show where different choices diverge, branch out, end, or converge. I've only seen story-organizing programs, which look like project managers, for linear (pardon the expression, Gormir) storytelling.

They use Java at my school for the introductionary imperative programming course (?!) "Don't worry about that 'public', 'class' or 'object' stuff, just write it in and do the bubble sort." I didn't use C or C++ until my senior year (for databases and OpenGL). Even my frikin Operating Systems class was in Java! But a CS degree is usually just what it says: a science degree. Theory, fundamentals, and math. Writing software is engineering (or art, depending on who you ask), not science. Big schools' CS programs have a separate 'Software Engineering' or 'Computer Engineering' degree that delve more into this.

Pagan was Ultima 8. It was actually a bit of a departure from the previous games in the series as you were on a separate planet with none of your companions. That and it was THE worst (surpassing U9 by far) Ultima game ever conceived. It was like diablo, but even less of an RPG and with godaweful controls. Oh god the memory, stab it out of my eyes

I'm only a wannabe game programmer... I do know you will need some well-rounded programming skills (most dev houses move their programmers around to different duties, so I hear), but specializing in one aspect is a good idea. Have at least basic knowledge in all of these things: Graphics - 2D and 3D(D3D and OGL - OpenGL is NOT dead) C & C++ UI's (toolkits and from scratch) AI (A*, minmax, rigging up something for the bad guys to choose their attacks) plain old application programming (main loop, event and exception handling, I/O, etc.) some good software engineering skills (UML & whatever the hell else people use; mine are terrible) Also, Gamedev and Gamasutra have some good info for people starting out. Gamedev has a good little primer on the basics and requirements of game programming. They suggest writing tetris with a high score since it has a game loop, event handling, simple graphics and rules, and I/O (for the high score file). Write it. Realize what you wrote sucks. Rewrite it. It should at that point. Then you have a computer game (albeit tetris) that you wrote. Write something else. Try to get a development job doing any kind of development in any language. Write a chess game or text RPG in your spare time. Run out of spare time. Turn pale. Lose you friends. Stop eating. Waste away. That's where I'm at so that's where my advice ends. If any Obsidian programmers are out there, I'm sure they can say more.