Loof

I have two main languages of choice C++ and Python. When efficiency is a must, or for larger projects I like C++ and Visual Studio 2005. (haven't run into a scenario where C++ is to slow yet). For prototyping and playing around with algorithms or if speed is not a problem, I usually use Python, mainly because i love pythons syntax and the freedom to combine functional and imperative programing styles. In this scenario I simply use the text editor of the moment to code in (haven't found one I'm perfectly happy with yet, the favorites so far are the above mentioned Notepad++ and ConText or CrimsonEdit). As for versioning tool I have only tried CVS and SVN, but greatly prefer SVN since binary files aren't such a hassle when using it. Question to all of you that have used some of the commercial alternatives: How does using them differ from using SVN, and what features do those packages have that SVN lacks?

I'm pretty sure the game is not CPU bound. I tried running around a while with a CPU meter on my secondary screen and as long as I wasn't in combat the CPU was pretty close to 25% usage. And I only have a P4 2.8 GHZ. also noted that Windows + NWN2 seem to use around 800mb of ram. That being said memory speed could still be a factor, or cache utilization or disk access or about a million other things ...

Yes I know what context switching is. Taks was talking about a scenario where you had dual cores without shared caches I think. So a switch from one core to another would either result in alot of cache misses or having to copy the cache from one to the other. At least thats what I think he was talking about. What I asked you if you had a guess as to what could cause a single threaded process to perform noticeably slower on a dual-core then a single-core CPU. Note: with single threaded I'm refering to a process that has only one thread where any heavy computations take place... Just so you don't missunderstand me again

Ok then context switching jumping from one core to the next shouldn't be whats causing dual-cores to run slower then single-cores (as all games would then behave that way). You seem to be in the know of how lowlevel stuff works under windows. You don't happen to have educated guess as to what could cause such behavior? (that is if the lower performance even is because of single/dual core thingy...)

You mean there is no windows api that grants you any control of how your threads are managed? (Don't have much experience coding windows specific programs)

this really is not possible. code that is explicitly written for threads will, or should, take advantage of multi-core. otherwise, it will not, and will simply run on a single core. in other words, the instructions issued by the CPU do not know, nor care, that there is another CPU available. the issue, however, may be context switching. i do not know if winders will switch your process from one core to another, but that is a killer sometimes (linux will do that). tas <{POST_SNAPBACK}> Well there should be a way for the programmer to tell the OS how to handle context switching, otherwise we have a scenario that can give bad performance on multi-cores with nothing we can do about it. Assuming the option exists and it wasn't used (or was used badly) what we end up with is basically what I said. So I would call my statement a gross simplification and not incorrect =P

No there shouldn't be a downside to having two cores, only a potential upside. But computers as you probably knew are very complex machines, so it could be that they actualy run slower because the programmers wrote some code that is suboptimal on dual-cores but not on single-cores. Also about the patching to utilize both cores, dream on... Splitting a piece of code into several calculation heavy parts is a non trivial problem (hard as hell according to some programmers). So since Bioware didn't code for multi-cores, obsidian probably didn't have time to do it during development, and I have never heard of such a large change being done in a patch. @Kraftan: Yeah almost all modern games are GPU limited and not CPU limited, but there really isn't any law that says it has to be that way. If someone made a game tomorrow with very advanced AI, physics and lots of procedural content it's possible/probable that such a game would be CPU bound. In other words its a design choice...

Ok so you and mkreku disagree about the power of the charge kick,kick,..., sounds like I will have to try the game myself to find out =)

""" I think most reviewers just don't see the entire strength of the game. If you just keep running and screaming frontal at your enemie like this is Quake or Doom, you certainly don't see the whole aspects that the game offers you. """ I agree that this is probably not how the combat was suposed to be used. But a good combat system should make that tactic a sure death for the player. Having many options in your gameplay looses its point if there is a simple tactic that always results in victory, wouldn't you agree? PS: hypothetical argument as I haven't played DM yet...

huh me or mkreku??

Doesn't sound that fun... But most of your complaints seem to be rooted in level design or enemy design. If someone was mainly interested in the game so they could look at the new ideas for the combat system. Would you say that its a complete failure? Or are the basics of the system sound and just badly implemented?

hihihihi sounds like your problem is with ****y GM's and not with forced or self created characters. But it was a funny story =)

My recomendation would be to ignore C++ for now. Even if most games are written in it, it's not the best language to start learning in if you have never coded anything before (to many ways to do the same thing). I would instead recomend that picking an easy to work with higher level language (my personal favorite at the moment is python). It will be much easier to pick upp the basic princaples and concepts in such a language then it would in C/C++, and once one has programmed a while picking up a new language becomes a trivial problem that can be overcome in a week or two. C++ may be an exception to that timelimit since it so damn huge, but even if it takes longer then a few weeks to learn once you get down to learning C++ it will still go an order of magnitude faster or so then if you started with it.

I have a Audigy 2 ZS running on a WinXP sp2 installation. Haven't toutched the sound drivers since I installed the system, but I think I used the driver on the disc that came with the sound card. But I didn't install anything except the actual driver, so maybe the problems come from some of the bloated extra programs creative thinks you should have. Anyway, the bottom line being I'm pretty sure it's quite possible to have sp2 and an audigy card playing nicely together.

Actualy the 360 graphics card has features that wont be implemented on desktop cards untill next generation. So at release time consoles can definetly have nicer bells and whistles then a PC. But yeah in a year or so the PC will probably have passed the 360 on all points again (with the possible exeption of only having 2 cpu cores, but then those 2 are more advanced...).

Rhomal I must say that I disagree with you about open source not being cutting edge enough. I think that in many instances open source projects have an advantage in that trying new things is easier and so they become more open to inovation. Where I think open source does have a problem is in ease of use, documentation, and optimization. The reasons I think this is so are: That since open source projects eneraly don't have money to hire testers or do usage tests they are totaly reliant on voluntary user feedback, which might be slow comming in. This in combination with the fact the the people that develop a product are very familiar with it and therefor have a hard time seeing what needs explaining and which interfaces are hard to use if you don't know all the commands at the top of your head. Writing documentation is boring. If noone is paying or forcing you to do so its alot less likely to get done. Some kinds of optimization can also be boring and so suffer from the same reasons that documentation does. That being said I am pretty sceptical about statements such as kraftans about the superior performance on linux/osx. While I think that in some cases linux/osx is probably faster I don't buy it as a general rule. For one counter example have a look at this test of Doom3: http://www.anandtech.com/linux/showdoc.aspx?i=2241 But yeah in some ways windows sucks, but then linux and os x suck in other ways so I find this whole "my OS kicks your OS's ass" debate pretty tedious

Does "Videogame Design and Multimedia" have anything to do with programming or is it a design (with posibly a bit of scripting thrown in). If its mainly design I don't think maths and algorithsm will be much of a focus at all. For programming it depends alot on what kind, but some basic algebra and calcalus, plus linear algebra (and possibly some statistics, but thats mainly of theoretical use). For algorithms it's even hardet to say as there aren't as many easy to use classifications to fling about ;-)

A few fast coments about this and that: If I'm not wrong GIF is now free to use (although what kraftan said used to be true), have never heard that jpg had the same limits. One reason that I can think of to choose D3D over openGL is that since it has a big company pushing for it the driver suport is generaly better. But on the other hand D3D has a higher driver overhead from what I have heard (although one of my teachers said that MS will probably change to a more GL like solution with a state-machine for DX10) About the question of where its posible to have both. Of course you can have two different codepaths, It just takes a bit of extra work to write two versions of the renderer. But its not a question of writeing two whole game engines as some people seem to think (depends a bit on how good your engine design is I guess) I think there was something else I wanted to comment on but I seem to have forgotten it now, oh well.

One thing that I get the impression you have to look out for based on available programs here in sweden. Look out so you don't apply for a program called "game programming", that is 99% design and 1% programming introduction and 0% maths, if you want to be a programmer. Not that designers aren't needed its just that alot of the programs here with the word game in the description aren't what they seem to be.

Yeah i played in a alien game once, but I don't think we used the normal rules. We used phoenix comand rules, but i think that is the system that the alien rules come from so they are probably similar. The above rules also get my vote for most complex and probably for most realistic (haven't studied them closely enough to be sure), but they seemed to be pretty focused on tactical combat and not on much else so I dunno how realistic they are for anything else...

agrees with kraftan

Yeah I have started developing my own ruleset a few times, but dont think i have ever finished it all the way from scratch. I have heavily modified existing rulesets though. My coments on each of your points: Setting: Well not much to say here... Precentile sytem: They are fast and simple so if thats a main focus its a good choice. But personaly I prefer targetnumber systems instead of roll under systems partly because I feel that its easier to asign varying dificultys. Another thing I dislike about precentile rolls is that the random factor is linear as oposed to a curve, but thats probably because I want the standard deviation of my random factor to be small (although I dont mind the range of posible outcomes being large... hope that wasn't to statistical an explanation). Classless: Definetly the way to go in my opinion, so no comments. Levelless: Good choice. Experience being spendable whenever the character has "downtime" is good. How are you planing to hand out experience? Personaly I dislike the gain by useing skill system, even if it is teh most realistic way. The reason that I do is because I feel it encurages rolling for the sake of rolling. In other words in PnP players tend to use skillrolls to solve problems that could be roleplayed without dice because they want to increase their skills. In crpg's you instead get the hit the wall to become a great warrior problem. Damagelevels: If this is good or bad depends alot on what kind of feel you are aiming at for the setting. They generaly tend to result in deadlier more realistic feel, where even "warriors" think a time or two before atacking someone. So if thats what you are aiming for its a good choice. But I have never seen a damagelevel system translated to a computer game and I instinctivly feel that it might be a bit bothersome to get right (hp systems are easier to tweek I think). Attributes-skills-derivates: This is pretty much standard. Just one question: Do atributes affect skills or how easy they are to learn or are skills independent of atributes. If this is the case how do you represent differing levels of talent for a skill? Roof/waterline and varying difficulty: See coments about precentaile rolls. Weapon damage based on amunition I agree with. Personaly I prefer having weapon damage as a fixed number and varying this based on the random factor of the tohit roll instead of having radnom damage factor for the weapon/ammo. Action system: One alernative to AP or actions per round that i have seen is to lett initiative determin number of actions. Example: if your starting initiative was 22, and you fired yoru pistol that has an attack cost of 4 your next action takes place at 18 an dso on untill you reach 0 and can take no more actions that round. Which is basicaly a modified AP system where where everyone has simultanious turns , which I think you said you wanted... Haven't realy seen any other good alternatives to the problem. One downside of the above system is that it can be pretty slow untill all the players know it by heart. I have also never seen it in a computer game, but in principle I think it shoudl work. Hope atleast some of this was helpfull =)

There are no nice people, just ones that are better at hideing the monster inside =P

Not that I'm a developer... but I thought it was a nice and refreshing new idea to try. One of the big limitations of consoles in my opinion has always been the controllers and the positioning sesnsativity might make some types of games that used to be a big bother more doable (for instance first person and RTS games). Depending on how detailed and fast the positioning system is of course...

Genre arguments are meaningless and pointless. That being said I consider Star Wars to be a space opera which acording to me is a subgenre of sci-fi and therefore sci-fi. Also the argument about sci-fi haveing to be based on current science is in my opinion rather fishy as many works commonly considered sci-fi don't have that element and instead focus alot more on possible futures of social trends seen today. Take for example almost all cyberpunk/darkfuture litterature...