Moribundus

Sweet I want to make a strategy game, that will be based on similar principles as Rebellion (fleets, agents, missions), so that it'll be much easier to make a MOD, that not only replaces units/textures etc., but also allows for customization at strategy level. I have just registered the project on SourceForge (I was surprised by the complexity of their services - bugtrackers, job ad-boards, etc.) and I'll post more info here once I get that design doc translated (it has over 15 pages )

Hmm, I guess it'd be polite of me to introduce myself too. *overcomes his laziness* Oh, hello there, then. I'm a huge fan of Rebellion - so many good ideas and background of the old sw movies, ah. I have been watching this site for some time, but have joined just recently. And since there's still no Rebellion 2, I'm going to try to do something about it (currently finishing and translating basic part of my design document). PS: English is not my native language (actually I'm czech), so please, don't beat me with sticks if I say something stupid.

What about allowing player to accelerate time when he's in sector and no action is going on? Eg. if the player would be waiting for enemy's fleet (perhaps trying to ambush). Otherwise 1/24

Let's say that you can move ships through hyperspace individually, unless the ship is part of some fleet. In that case, you must move the entire fleet (or moving the ship moves the whole fleet), eg. fleet and its ships. Or let all ships be part of some fleet - no matter how large the fleet is, and allow "unassigned" designation for fleets (so it won't mess up alliance stats), or don't count one-ship fleets into stats Maybe all ships could belong to those "unassigned" fleets right after their creation. The building process might be able to bring serious chaos into your stats too. I doubt anyone makes 10 ships at the same time at one planet and assignes them to one fleet, instead it's more natural to build 10 ships at 10 planets and create 10 one-ship fleets (at least until those ships reach the real fleets and join them). How would this get along with your AI code?

Why shouldn't alliances contain both ships and fleets? Alliances would contain individual ships and fleets would work only as some kind of containers. This way it would be possible to manipulate with whole fleets and leave some ships independent at the same time. Hmm?

Battles were perfect in rome. I always ended up with routing the entire enemy army early in the battle and then slaughtering the rest of it's units on the run with my cavalry. Sometimes it was 10 to 1 victory. Battle AI certainly was worse that the auto-resolve algorithm Btw. backstabbing allies were quite common in the Middle Ages. However the AI diplomacy was still lacking (at least in rome) - protectorates were useless, since the protected nation would sooner or later turn hostile on you and attack your unguarded towns when you least expect it. And sometimes the AI won't make peace with you unless you met their ridiculous demands (even if they were in no position to make any). Kamikaze barbarians

Try VegaStrike - http://vegastrike.sourceforge.net. It's space trading/shooting game like freelancer/elite/privateer. It has some nice mods, too.

Hehe, yeah, the poll is missing something... Don't forget that you don't render any complicated scenes (but on the other hand, there isn't much to render in space, is there?). I'd personally go the opensource C++/OpenGL way, as bud mentioned. Besides, there's always chance that the project would attract attention of other opensource programmers and the team would grow (and I bet there are people who would love to help you make a new linux game).

Hell yeah, all total war titles are great games and even greater source of inspiration. I haven't played Medieval II yet, but I loved the minor factions in Rome (and first datadisk).

You cannot skip the "read bytecode instruction -> translate it -> send it to cpu" phase, which is done run time, so there's no way you can entirely eliminate the performance penalty. On the other hand your code will work on all platforms (just like Java), unless you use platform-specific extension, like managed DirectX (which you probably will). So my point was that i don't see any real advantage in using .NET made modules over common dlls/modules. And Evaders, the new design looks great!

Seems like best free solution for playing old games on Windows Vista. VPC also emulates hardware, so there's no problem with drivers (eg. for newer soundcards).

Sweet Well, looking for this information in memory would be like looking for a needle in a haystack, if it would be possible in the first place. I haven't played EAW so much too, so I don't remember if it shows some kind of stats concerning surviving ships at the end of combat. And even if it does, it would be still very hard to intercept.

The idea is nice, it's everyone's dream for sure, but we will never have the source code. We don't even have the source code of lucasarts oldest adventures. And as darthtuff mentioned, you probably won't be able to get any output out of EAW. So we would be probably better off with making our own game I'll add my two cents here. I'm currently planning to make a 4x space strategy myself (and hopefully with the help of my friends). I want it to be based mainly on diplomacy/warfare and it'll use C++/OpenGL. But it'll need some free time first

Wine, definitely (http://appdb.winehq.org/ check WineHQ to see which apps are reported to run). And there's also Cedega (http://www.transgaming.com/index.php?module=ContentExpress&file=index&func=display&ceid=29), which emulates even directx9 and works guite well, but is not free. Also some commercial games have their native linux versions (neverwinter nights, unreal tournaments, dooms, quakes and maybe others) and some of these games run even faster than on windows platforms . And imho with accelerated window managers (beryl/compiz), linux kicks Vista's butt on all fronts (almost).

Hi, I just managed to make Rebellion run under wine yesterday and I must say it works flawlessly (almost). I play it in windowed mode, it's better than changing my desktop resolution every time. You just have to run winecfg and in the graphics panel check the "Emulate virtual desktop option". Videos are playing, the character stat are there, just perfect. Oh, the only buggy thing I've noticed so far is that I can't skip the briefing when starting new game. I'm running Arch Linux with wine version 0.9.29. Btw. aren't the starting stat's random? Mad78: Try installing Microsoft Virtual PC 2004 with win98/95 (or XP ). Microsoft released it as freeware, you can get it directly from their site http://www.microsoft.com/downloads/details.aspx?FamilyId=6D58729D-DFA8-40BF-AFAF-20BCB7F01CD1&displaylang=en. But Vistas aren't listed as supported system (and micro.. is developing VPC 2007) so it's not guaranteed to work. Otherwise you would need another emulator (eg. VMWare).

Lol, FWIW neither was C++ True. But I meant that those languages/development environments were made for easier and more efficient creation of desktop and database applications. And that's where their strengths lie. he I must admit that I have never worked with .NET, but I don't see how it should connect developers using different languages. You still can't mix more languages together (into one code), can you? So it's nearly the same as using DLL libraries, but with slight performance penalty. Correct me, if I'm wrong. However .NET if used right can be powerful tool. Eg. RunUO shards (open source Ultima Online servers made in C#.NET) can recompile all user scripts (also C#) into bytecode and then use them during the startup of the server. So the scripts run faster than interpreted scripts and because almost whole server is made this way (except network layer i guess), it's also extremely modable. You just have to know what you want from it. More CPU expensive AI would only be better (at least for Rebellion) But otherwise I agree with you.

OpenGL vs. DirectX...end result the same. I believe OpenGL app may be faster in the end i think and it's platform independent. And it's a little chaotic with all these extensions, but just a little Nvidia also released and OpenGL shader debugger if i remember correctly. But they might as well have released the same (and hardware benchmarks) for DirectX. Programming languages? Forget about Visual Basic and Delphi, they weren't even designed for 3d games in the first place. As Nordwindranger mentioned .NET uses managed code (i believe), therefore won't be as fast and efficient as the non-managed languages. C# may be also somewhat slower than C++ (i'm not saying that C# and .NET is bad, I just don't think it's the best choice for making games). And also I guess that free 3d engine is also written in C++ (also consider it's license, if it's GPL or something like that, then you have to release the source of your program). As for the development environment. Visual Studio very good, you can also use Eclipse or NetBeans (both were designed for Java, but support other programming languages through plugins). Books are surely good and you can find many of them in ebooks on various sites for free. And some links to good sites dedicated to game development. Check out their articles and tutorials. http://www.gamedev.net http://www.devmaster.net http://www.flipcode.com - this used to be one of the best sites, it's only an archive of old articles now, but still very useful http://www.codesampler.com - some other tutorials I hope at least those links will be of some use to you

Hi, just few another random thoughts about your laser headache. Consider different model. We place turbolasers on "mount points" on the hull of the ship. Each MP has it's arc and power. Arc is determined by MP's placement, power by the total number of turbolasers this MP is representing. Mount point fires on an enemy ship when it's in range and in arc, probably fires at one ship until it's down (makes no sense for a turret to spin left-right until the gunner gets all dizzy ). Now, a small example... Lets take an ISD Vivisectionist and place five MPs on it. Fore, fore-port, aft-port, fore-starboard, aft-starboard (or maybe even top/bottom, but it still can reduce the "halo" effect of lasers to reasonable levels) with appropriate arcs and powers 20/10/10/10/10. If this ISD meets an enemy corvette and approaches it from front, it can attack with all three fore MP's, with a total power of 40. If that corvette would attack our ISD from side, only that side's MP's could respond and the power devoted to the destruction of this little corvette would be only 20 (30 with fore). However if the ISD would be attacked by two corvettes from both sides, it wouldn't notice the difference, since both sides would be evenly tearing the corvettes to pieces. Next, laser recharge rate (keeping separate recharge counters for each MP sounds like a good idea) would not depend on the number of turbolasers the ship has, but instead on it's reactor type (or class) and it's status (damaged, overheating, etc.) This also brings new possibilities for a ships's commander. First, ships position would be important because of firing arcs, and second, commander would be able to issue orders to take power from aft turbolasers and use it to boost reactor, thus allowing fore lasers to fire for a longer time without recharging. And so on... Well, just some ideas I don't know how far are you with coding, but the video looks very promising. Keep it up and happy coding.