Open Source Software Experiences?

[Tigershark_BAS 0]

Has anyone out there been involved in contributing to an open-source community project to build a final product?

And yes...I am aware that addon making/mission making in teams is open-source in it's most basic form.

But what I mean is a fully fledged open-source project.

Can anyone share their experiences? Would anyone like to be involved in one?

I'd like to know what experiences people have had with this. Reason I ask is I'm seriously entertaining the idea of perhaps starting one up to make the ultimate modern land based warfare simulator. There certainly is plenty of talent here to tap into...question is....is there a desire to make one?

There are plenty of engines available to prototype any design ideas we come up with. As for the final engine we would need to see what happens. Definitely at this stage there are design documents, models and textures and GUI's to be created and tested before we even consider what engine to use for a final product.

Sound off and tell Unky Tiger what a dreaming fool he is and how he should join Hellfish at the "funny farm".

*starts making noises with his lips while drooling*

[raedor 8]

pretty good idea, this project.

but as far as i know "our" ofp community this project is going to die in polls about "armor values" and so on.

i can support you in 1 or 2 years, when i'm better in programming and such stuff

[Jezz 0]

me and kenji from rhs and another 5 guys from uni spent about a year wroking on a helicopter flightsim together never got finished though due to kenji moving to japan , but we did manage to sell some of the engine tech to company for money .

funnily alot of the of the models i know kenji made for this project got ported into ofp the rhs shilka being one of them

[Tigershark_BAS 0]

The problem I see at the moment is a lack of programmers.

We seem to have plenty of modellers and texture artists as well as people with basic programming skills to make the CPP files for addons for example....and scripting.....but I'm not sure we have the people with the skills to do physics and AI programming.

Do we?

[raedor 8]

The problem I see at the moment is a lack of programmers.

We seem to have plenty of modellers and texture artists as well as people with basic programming skills to make the CPP files for addons for example....and scripting.....but I'm not sure we have the people with the skills to do physics and AI programming.

Do we?

that's what i am learning currently

[Jezz 0]

me also but uni is pretty pyscho at the moment(doing a software engineering degree)

[Kaitnieks 0]

I got very good experience and bad experience with releasing open-source programs.

I'm a coder my self, and people "hire" me to work for them.

But in my free time I make programs and release them opensource to various communits/groups (I code in C++/Java/Delphi/Pascal).

How ever I think it has to do with age group because many people from the age 10/15 abused my programs and even stole it claiming it as their own. So maturity has some thing to do with this also.

PS: If you need any coders I can help you out if you wan't I got some spare time now and don't know what to do with it.

Sincerly,

Aivars Irmejis

[Tigershark_BAS 0]

Kait,

Could you make C++ GUI's from a Photoshop image?

Tiger

[D.murphy man 0]

This sounds like very intresting idea! kinda like a community open source OFP

Maybe with some nifty network codeing and low CPU demanding models we could have some MMO War sims going on, now that would be amazing.

Any way hope it takes off,

Murphy

[EiZei 0]

This sounds like very intresting idea! kinda like a community open source OFP

Maybe with some nifty network codeing and low CPU demanding models we could have some MMO War sims going on, now that would be amazing.

Any way hope it takes off,

Murphy

You are aware what kind of resources that requires? Freeciv seemed to take ages before it matured and it was hell lot less complex. Developing a massive multiplayer FPS that would not be crap requires enormous amount of effort and expertise.

[Pathy 0]

What i always wanted to do was make a realistic kind of C&C.

You know how the C&C series is like, well, pretty unrealistic, doesnt really lend itself to real life tactics and stuff. Well i had thought about making something where, for example, tanks can get knocked out in one hit, different zones of damage ect. I was thinking of making the main force British btw

I realised this is basically what the close combat series achieved, but i suppose it´d still be cool to do it for a modern era, with modern tactics, modern weapons, in 3d and on a larger scale. (bigger battlefields) ect

Are you thinking FPS or RTS? Either way, if you need a texture....

[hellfish6 7]

I've always kind of wanted to do just what Pathy suggests. A 3D battlefield sim that's part TacOps, part Combat Mission and part Close Combat, maybe with an integrated first-person sim built into it.

I'd just like to see, in real-time 3D, a battalion of tanks charging across central Germany, with an integrated platoon to brigade level chain of command (click on brigade commander's unit/TOC and give orders to the whole brigade - click on a platoon commander's tank and give orders to the platoon).

Not much to ask for, eh?

[Tigershark_BAS 0]

I was thinking FPS....the first real "simulator" FPS....after OFP of course

[Tigershark_BAS 0]

An example of the engine we could use to prototype:

Grass

Terrain and House

Long Distance Terrain

[Tigershark_BAS 0]

I've always kind of wanted to do just what Pathy suggests. A 3D battlefield sim that's part TacOps, part Combat Mission and part Close Combat, maybe with an integrated first-person sim built into it.

I'd just like to see, in real-time 3D, a battalion of tanks charging across central Germany, with an integrated platoon to brigade level chain of command (click on brigade commander's unit/TOC and give orders to the whole brigade - click on a platoon commander's tank and give orders to the platoon).

Not much to ask for, eh?

Kinda like the Total War series?

This was the first game I think to implement a turn based scenario with then an optional RTS element to dictate the outcome of the turn.

Except you propose a turn based/FPS hybrid.

Interesting....very interesting....I wonder if the load time between turns might drive you insane though.

Definitely an intriguing idea though. You could write that up and sell it Hellfish.

[Kaitnieks 0]

Kait,

Could you make C++ GUI's from a Photoshop image?

Tiger

If I got the image I can code the GUI yes.

[der bastler 0]

Suggestions:

Base

ClanLib: http://clanlib.org/

3D engines

Crystal Space: http://crystal.sourceforge.net/

Examples:

http://community.crystalspace3d.org/tiki-br....geId=87

http://community.crystalspace3d.org/tiki-br....geId=49

http://community.crystalspace3d.org/tiki-br....eId=124

Irrlicht Engine: http://irrlicht.sourceforge.net/

Examples:

http://irrlicht.sourceforge.net/images/sotw/024l.jpg

http://irrlicht.sourceforge.net/images/shots/057.jpg

http://irrlicht.sourceforge.net/images/shots/052.jpg

http://irrlicht.sourceforge.net/images/shots/047.jpg

http://irrlicht.sourceforge.net/images/shots/064.jpg

http://irrlicht.sourceforge.net/images/shots/055.jpg

[Pathy 0]

I've always kind of wanted to do just what Pathy suggests. A 3D battlefield sim that's part TacOps, part Combat Mission and part Close Combat, maybe with an integrated first-person sim built into it.

I'd just like to see, in real-time 3D, a battalion of tanks charging across central Germany, with an integrated platoon to brigade level chain of command (click on brigade commander's unit/TOC and give orders to the whole brigade - click on a platoon commander's tank and give orders to the platoon).

Not much to ask for, eh?

Kinda like the Total War series?

This was the first game I think to implement a turn based scenario with then an optional RTS element to dictate the outcome of the turn.

Except you propose a turn based/FPS hybrid.

Interesting....very interesting....I wonder if the load time between turns might drive you insane though.

Definitely an intriguing idea though. You could write that up and sell it Hellfish.

Thats exactly what i mean Hellfish, and dynamic too, so that sucess builds on sucess and the like. If youve ever played DID total Air War, like how that is dynamic (one of the best dynamic engines ive played with tbh). Of course close combat was always dynamic also....

And individual soldiers getting experience, personality ect. Everything mapped to the last detail.

I´d say real time more than turn based.

[soul_assassin 1723]

IMHO this can develop into a major project. And I think it can work out pretty nicely. Let me know if you need a 2d artist, i'd be glad to help out.

[hellfish6 7]

I've always kind of wanted to do just what Pathy suggests. A 3D battlefield sim that's part TacOps, part Combat Mission and part Close Combat, maybe with an integrated first-person sim built into it.

I'd just like to see, in real-time 3D, a battalion of tanks charging across central Germany, with an integrated platoon to brigade level chain of command (click on brigade commander's unit/TOC and give orders to the whole brigade - click on a platoon commander's tank and give orders to the platoon).

Not much to ask for, eh?

Kinda like the Total War series?

This was the first game I think to implement a turn based scenario with then an optional RTS element to dictate the outcome of the turn.

Except you propose a turn based/FPS hybrid.

Interesting....very interesting....I wonder if the load time between turns might drive you insane though.

Definitely an intriguing idea though. You could write that up and sell it Hellfish.

Thats exactly what i mean Hellfish, and dynamic too, so that sucess builds on sucess and the like. If youve ever played DID total Air War, like how that is dynamic (one of the best dynamic engines ive played with tbh). Of course close combat was always dynamic also....

And individual soldiers getting experience, personality ect. Everything mapped to the last detail.

I´d say real time more than turn based.  

I hadn't played that, but I've played the Apache/Commanche vs Havoc/Hokum series which had a pretty cool dynamic engine, as well as Falcon 4 and Silent Hunter III. I don't know about personalities of individual soldiers, but I think the personalities of individual leaders would be essential - and you should be able to foster them and assign them as needed like SHIII.

I really think dynamic campaigns/missions are the way to go for single player portions of games - they never play out the same way twice and are always interesting. I'm not a fan of the linear game play you see in 95% of games.

Whatever project that may come out of this, if any at all, I'd strongly suggest a dynamic campaign system - where finite forces fight over a finite battle space. If you're going to invade an island and the enemy has only one battalion available to them, with no hope of reinforcements, they should only have one battalion with only what that battalion has.

The closest I've seen to this - in any game - is the Chain of Command's work and, specifically, the dynamic campaign Bagel built using the Command Engine and the SEB Nam Pack 2 ages ago. I spent countless hours on that mission, but it never made it past 90-95% completion. (And kudos to the Dynamic Missions Alliance for leading the current revolution in OFP missions)

I think there's definately a talent pool available in the community to get something going.

[hellfish6 7]

I've always kind of wanted to do just what Pathy suggests. A 3D battlefield sim that's part TacOps, part Combat Mission and part Close Combat, maybe with an integrated first-person sim built into it.

I'd just like to see, in real-time 3D, a battalion of tanks charging across central Germany, with an integrated platoon to brigade level chain of command (click on brigade commander's unit/TOC and give orders to the whole brigade - click on a platoon commander's tank and give orders to the platoon).

Not much to ask for, eh?

Kinda like the Total War series?

This was the first game I think to implement a turn based scenario with then an optional RTS element to dictate the outcome of the turn.

Except you propose a turn based/FPS hybrid.

Interesting....very interesting....I wonder if the load time between turns might drive you insane though.

Definitely an intriguing idea though. You could write that up and sell it Hellfish.

Yeah, I guess it would be a lot like to Total War series now that I think about it.

The game I'm currently trying to sell to a developer is set in 1918-1939, like of like an Indiana Jones style RPG where you can be an adventurer, mercenary, government agent, academic or pretty much whatever you want to be. You can own land, every region on the earth will have landmarks (hotels, terrain features, plantations, factories, important buildings, cantinas, etc.) where missions can be assigned or take place.

You can buy and sell most everything. Most of the game would take place on a 2D world map (broken into regions - like Chicago, Paris, Lapland, Togo, etc - if you've ever played Victoria or Hearts of Iron you know what I mean) but adventures would take place in a 3D environment. I don't even like RPGs, but my design doc for this game makes me want to play one.

[D.murphy man 0]

I've always thought about getting into making a game with an egine my self many a time, closes ive gotten is OFP with limitless amounts of addons and the extreamly dynamic and easy to learn/use script egine we can practicle make intire new games.

Ive also flirted with a program called 'Game Maker' which had an easy scripting language (never got to grips with it though) and also i handy drop and drag interface for making simple 2d games and also exteamly complex 3d ones but i doubt this would be useful for a project of this scope. Its more for people with little or no programming skills who just want to make simple games.

[shinRaiden 0]

There's several components involved...

* Systems engine

* Graphics engine

* Art data

* Network system

If you're doing a 'pure' OSS project without commercial backing, that would result in even your project leads doing their work on the side of their RL. When you're working in that fashion, you can effectively shoot any scheduling or standardization desires right out the window. You can have coders working PT or in hobby-time, but the managers need to be more on top of things.

You're also going to need a management system. You're going to need a full dedicated server with some sort of controlled versioning system management software package.

There's also vision and development plans. Look at how long it took BIS to get OFP to market, and that was commercial full-time development. Look at how long they've continued to support it, and where they're taking it to the future.

Along with the oganizational plan you also need a product plan. What niche and standards are you planning to target? What features and weaknesses does your 'competition' have? What do you need to do, and what would you like to do?

Then there's IP issues. Don't know if you recall that just before Doom3 went Gold, John Carmack posted in his .plan about a cool new method he devised for calculating Z-buffer clipping on shadows. Well, Creative had their lawyers out trolling and discovered they previously had a patent on that obscure technique, and it didn't matter if Carmack independently developed it, id still had lots of money to sue for. They ended up extorting id into a fat EAX licensing deal that id hadn't planned on in exchange for 'forgiving' that oversight.

Art is another complicating factor. Sometimes it's handled under the same blanket OSS license, sometimes it's handled under a seperate restictive license to protect the artist's time and efforts. Art is not something that generally can be chopped into trivial components as a line or two of code can be, and I'm sure you're well aware of the complications where artists bring in 'lite' versions of their 'pro' products.

Let's look at the various components.

1 - Systems engine.

You've got to determine the permissable level of complexity and level of interaction. You could - theoretically - model and code every toggle switch on a flight panel. You would then have to make an actual hardware mockup to be able to use it. Conversely, something like NetPanzer is extremely simple - click on a location to move there, click on a unit to attack it. You have to determine your balance.

Then there's stuff like what we've seen in OFP functions. Some stuff is handled automatically by the engine, and some is scriptable. I doubt we as the community will ever stop arguing about what should be where, and what level of abstraction vs detail in the functions is correct or acceptable.

2 - Graphics engine

There's a lot of things here as well. You can tune the system for relatively short indoor distances like Quake, or you can tun for very long distances like in a flight simulator. Those decisions affect the performance targets for the rest of the system, as well as the detail specifications for the art data.

Added to that is how to go about making pixels. If you go with DirectX, you get a wealth of resources in that direction, at the cost of portability and true scalability. If you go with OpenGL, you can add in scalability to rediculous levels and readily port the project to OS platforms other than Windows platforms. The cost is a different way to write code, and all the other nuances of OpenGL.

Additionally there's the issues of the rest of the DirectX suite, including DirectPlay, DirectInput, DirectSound, DirectMusic, etc.

3 - Art Data

There's a whole ton of stuff that goes into this as well. You need to know what functionality is needed by the project so you can determine waht can be handled dydnamically in the engine and what needs data helpers.

Research, modeling, licensing, texturing, lightmapping, and scalability are all time and money and resource drains. If you want hundreds of thousands of objects rendered with thermal imaging on 100km view distances, obviously your baseline standard for model poly and texture detail may be significantly different than what you might otherwise prefer.

Obtaining resources can be complicated as well. For maps, you can either go 100% artificial or modeled on the real world to some extent. If you start using real world references, sooner or later you'll have some sort of limitation in data or licensing.

4 - Networking

There's a lot more to networking than just sending packets between a client and a server. If abstracted low enough, you could reasonably implement remote or slaved rendering, distributed servers, p2p unit syncronization, and a whole host of other little tidbits. What about integration with other programs? Will this run totally it it's own little box, or do you want to integrate it with other networked games or simulators?

If you choose to go the client-server route, there's the question of self-hosted vs 'official' host. If you're hoping to implement any sort of RPG'ish player profiles, or epic campaigns, that almost mandates controled and provided servers.

There's DirectPlay vs. Sockets again, IPv6, VoIP, NAT firewalling, chat comms, clan identification and management, and stuff that goes along with that.

---

Now that I've said my little rant, how about some pushing on the OFP2 requests wiki? I've spent almost no time on it over the past week, but I've already trimmed out about 2/3rds of the requests on the AI thread into simplified one-line items.

[PiNs_Da_Smoka 0]

Garage Games is a very good site. Which gives you the torque engine for $100. You are probably aware of this already, and i also didn't read all the posts, but it is pretty popular base for alot of projects. I used to be involved in 2 very pormising projects on here (as a modeler) before i joined the army. But like i said, just in case no one mentioned this, its a very good option, although you wont be starting from scratch, but with this engine you pretty much can do what you want and its almost like starting over. Anyways....

*drops 2 pennies on the ground*

-Edit-

Also, ever heard of DARKBasic?

*drops another 2 pennies*

[Acecombat 0]

*Picks up the pair of penny's pins dropped and shakes them in hand , thus waking up Tigershark and others from their wet dreams*