Seeking fresh recruits -- game development

[keinmann]

Good afternoon, everyone...

I am seeking some new "recruits" for a serious project of mine. So I figured I'd give the Total War community a try, as I might be lucky enough to find the right guy/girl here. Before I go any further, I must assure you that this is not one of those run of the mill, fly-by-night "let's make a game" internet projects. This is a serious thing which has been in the inception/prototyping stages for a couple of years, and we do actually know what we're doing. So you want to know what exactly we are doing, right? And you want to help? lol...

I'm actually developing a game engine, at this stage, and not an actual game (though that's what the engine is for). But this engine is quite different from others. It is to be, almost entirely, written in C#; and is primarily targeting the DirectX 10, 10.1 and 11 technologies. However, I do hope to port it to OpenGL/Mono, to open up the possibilities of true cross-platform deployment; provided I can afford the necessary talent. Over the past few years I have developed several working prototypes of this engine. I have quite a bit of solid, perfected code from these prototype versions which will be used (everything from high-precision timers to dynamic lighting to advanced ballistic physics). And I have finally decided to begin work on an official, commercial version of the engine. I've literally just started a new Visual Studio solution and have begun importing my best chunks of code from past prototypes and tying up loose ends; [re]constructing the foundation of the engine. So this gives me a unique opportunity to seek new "recruits" whom I can tightly integrate into the project from the start. Let me give you a better technical/architectural picture of the engine...

This engine provides an abstract interface for game and simulation development which removes the complexities of dealing directly with low-level programming and the local hardware of end-user systems. The engine itself handles the "boilerplate" work of finding graphics cards and monitors, polling system specs/caps and selecting compatible display modes, shader models, yadda-yadda... It also abstracts away the complexities of dealing with multiple DirectX versions; and in its finished state, it will provide an interface with which the same game can run on virtually any rendering/media backend (be it DirectX, OpenGL, a custom software renderer, etc). The engine architecture is modular and component-based, making it both light-weight and extensible; but more importantly, the weak-coupling and low interdependency of engine sub-systems means that individual components can be easily rewritten or replaced (for example, to switch to an OpenGL backend or Miles Sound System) without breaking everything else.

As of now, I am predominantly targeting the latest DirectX SDK through the use of the SlimDX wrapper; which has become quite stable/robust and is suitable for the task. However, I plan to extend the capabilities and feature sets to the greatest extent possible. A complete wrapper for nVidia PhysX is one of the more important things on the agenda. I may also decide to support some other things, such as the CUDA libraries. The DirectX 11 renderer will indeed support multi-threaded rendering and all of the other great technologies that make 11 so awesome. But all of this is the bottom-end, low-level architecture of the engine... The actual development interface exposed to client code is a whole different story!

In short, the engine will provide all of the "bread and butter" features other commercial engines provide. However, there will be a greater focus on physics realism (including ballistics and flight physics), high-quality lighting and shaders, performance and stability and more. There's far more to it than that, but it's more than I have time (or space) to adequately explain...

I need more programmers. Ideally, I would like to find other people with the same (or greater) experience and passion for C# and game/graphics/physics programming. However, as I said before, there is a unique opportunity here for me to help train newbies and aspiring programmers. All you really need is a brain, a willingness to learn and put in effort and a passion for video games and simulations. This project will eventually shift gears into actual game development when the engine is ready; and I am fortunate to have the help of a good friend who is an amazing 3D artist and game designer (he actually has a degree in game design, unlike me). But this is not limited strictly to programmers. We do need 3D modelers, texture artists, designers and everything else. So if your talents lie outside the programming realm you are still to be welcomed with open arms.

NOTE: At this stage I am not officially hiring people to work on a salary. But if you prove yourself capable and reliable (and want the job) an official, paid position (and stock ownership) in the company can indeed be your fate.

If you have any questions, fire away...

Regards,

[GrnEyedDvl]

Hello again Keinmann.

First, I want to wish you luck on the project. I am thinking you will need it. Its ambitious to say the least.


And I am curious about a few things, such as why the Visual Studio choice and more specifically why C# as opposed to C++?

Also, there are plenty of open sourced engines out there, such as Delta and Ogre and many others. Why do you think building a new engine is the best way to go? I have played with a couple of those, and in my opinion they are quite good even though I am truly far from a professional programmer.

A few things you mentioned hint at a first person style of engine. I dont expect you to give too much away here, but a few more details would be nice.

I am also curious about what your infrastructure looks like so far, as far as the team and assets and most importantly how its all managed, or will be managed. You said this is not your typical "lets make a game" projects, and the main difference between those types of operations and a successful operation is not programming talent (though that is important), its management and the ability to follow through from conception to market.

I have no idea how many true programmers there are on TWC, though I can name at least three besides yourself. I hope you find some good people to turn this into reality with you. If you manage to get off the ground and running with this, then keep me informed (privately or publicly, your choice) once in a while of your progress. I may have some input that could be valuable to you.

At the very least, if you get a game released we can review it and give it a subforum much like we have done with Mount & Blade and Minecraft, which will give you some exposure. I and most members of this forum like indie game developers, and while I dont want to turn the site into an advertising platform via the forums, once you have an actual release we can support it.

[keinmann]

And I am curious about a few things, such as why the Visual Studio choice and more specifically why C# as opposed to C++?

The first question is easy... if you're writing C#, Visual Studio is the best available IDE. And I'd argue that if you know how to fully use it, it's virtually an unbeatable IDE period. Why C#? I think the better question, in light of C#/.NET's technological advances in recent years, is why not?

C# is now the "official" language of Windows development, finally ousting the dated Win32 API. Much of Windows 7 itself is written in C#, as a matter of fact. And the advantages it offers are pretty clear. It's now every bit as fast (sometimes faster) than native C/C++ applications. And the CLR is simply a better memory manager than you or I can ever hope to be unless we write a complete memory manager of our own (and even then I'd bet Microsoft's coordinated team of hundreds of experts still win). Video games, by their very nature, are incredibly memory/resource intensive. When the delicate balancing act of memory management is left in the hands of humans memory leaks are going to be inevitable... and it often becomes so complicated and difficult to pinpoint that these leaks end up in commercial game releases. Skyrim, for example... CTDs like crazy! Pretty much every game has serious memory management flaws and has a tendency to CTD or freeze at times (some more often than others). And the worst type of memory leak isn't the kind that crash the game, it's the kind that hide, undetected, and suck away performance (often going completely unfixed). While C# doesn't, by any means, totally eliminate memory leaks, it cuts down their frequency and severity 100 fold. Theoretically, this provides the potential for not only much greater stability but actually increasing speeds drastically. Memory footprints are smaller, resources get properly disposed of, no hanging/null/invalid pointers, etc. It's really only in application launch/startup speed that native languages can burn C#, as the initial JIT process adds some milliseconds in there (though NGEN can help with that!).

Furthermore, the game industry is still completely hung up on C and C++ (just like they're still stuck on 3DS Max, despite it being outclassed by Maya), and no one has yet to write a very good C# engine... despite the technology being right in front of our faces.

Also, there are plenty of open sourced engines out there, such as Delta and Ogre and many others. Why do you think building a new engine is the best way to go? I have played with a couple of those, and in my opinion they are quite good even though I am truly far from a professional programmer.

Yes, there are. But the people who developed those engines had to ask themselves the same question. The answer I'm sure they arrived at was that they wanted an engine tailor-made to do what they wanted it to do, rather than someone else's vision of a great game engine.

Most game engines suffer greatly in some of the areas I deem the most important, based on what type of software I intend to develop with the engine. While there are plenty of great physics APIs out there, very few of them focus on things like realistic flight physics, ballistics, etc. Very few have dynamic damage models for humans, vehicles, buildings and objects. A lot of engines these days have tried to marginalize the role of programmers and make game development a job for designers with fancy tools with fancy buttons; but it typically fails to ever work properly and gives all the games a "cookie cutter" look (all look far too similar). An even greater number of them are far too abstracted/simplified, offering very little exposure to the metal (hardware), and are too reliant upon scripting (which gets slow). On top of that, most engines are far too bloated and eat up too much memory; not to mention bugs and memory leaks which can be almost invisible from outside the "black box". And finally, we must also consider the costs of commercial use! Some of the really high-end engines which are only available to certain people upon inquiry want up to 35% of your revenues... not your profits, but your revenues!

A few things you mentioned hint at a first person style of engine. I dont expect you to give too much away here, but a few more details would be nice.

Well, not exactly a fps-type engine, though it will be, in its finished state, capable of creating a FPS with indoor and outdoor scenes and seamless, open-world maps. Switching between first-person and, say, isometric mode (as seen in NTW) is as simple as selecting a different camera class or extending/writing a new one. Let me know what other details you'd like to know about and I'll try to explain the best I can.

I am also curious about what your infrastructure looks like so far, as far as the team and assets and most importantly how its all managed, or will be managed. You said this is not your typical "lets make a game" projects, and the main difference between those types of operations and a successful operation is not programming talent (though that is important), its management and the ability to follow through from conception to market.

A very good question and some very good points. I'd rather not get too explicit about things like financial assets (as funding will be coming out of my own pockets initially), but there are some things I'm able to tell you. I'm going to transition from running this project personally to founding a new limited liability corporation or a type-C corporation in the near future, in which I will initially own all or most of the stock. Some of the shares are going to be split up amongst the company executives/founders and employees. Once the ledgers are all balanced out and we've seen to all of the legal requirements I am going to seek outside venture capital; and I already have a few personal contacts and leads for this. We may also look into "kickstart" funding options, at least on a supplementary basis.

This area is one of the main ways in which this is going to be very different from an amateur "let's make a game" project. I'm actually probably a much better businessman, executive and financial officer than I am a programmer. I actually make my living as a day trader (professional trader of stocks, commodities/futures, currency and other financial instruments). My friends and family often call me "the financial wizard" because of this and my ability to consistently earn astonishingly high returns in short periods of time. For example, in my first year of trading I turned a mere $1000 into over $16,500 in the first 5 months, and then used that to net a year-end profit in excess of $57k. Now I'm working with much larger sums of capital and seeing much bigger profits. Just two months ago I had a single week in which I had a net profit of just under $10k. So I know how to both manage the business and manage its cash assets; we can generate profits before a game ever sees the light of day, provided we have the capital.

I have no idea how many true programmers there are on TWC, though I can name at least three besides yourself. I hope you find some good people to turn this into reality with you. If you manage to get off the ground and running with this, then keep me informed (privately or publicly, your choice) once in a while of your progress. I may have some input that could be valuable to you.

Definitely! Sounds great!

At the very least, if you get a game released we can review it and give it a subforum much like we have done with Mount & Blade and Minecraft, which will give you some exposure. I and most members of this forum like indie game developers, and while I dont want to turn the site into an advertising platform via the forums, once you have an actual release we can support it.

And so does this! But keep in mind I'm not working on a game just yet. Just the engine for now!

[Bolkonsky]

Hello again Keinmann.
And I am curious about a few things, such as why the Visual Studio choice and more specifically why C# as opposed to C++?

Also, there are plenty of open sourced engines out there, such as Delta and Ogre and many others. Why do you think building a new engine is the best way to go? I have played with a couple of those, and in my opinion they are quite good even though I am truly far from a professional programmer.

There is actually some decent C# engines as well, including an Ogre 3d port - Axiom.

[keinmann]

There is actually some decent C# engines as well, including an Ogre 3d port - Axiom.

I have the source for both of them on the desktop (recent checkouts have been there in some form for over 2 years), right at my fingertips. So I'm able to look at what's been done by other managed engines to learn from their successes (and failures).

Im not a fully formed programmer per se (still in uni), but I do forward that the idea to use C# is the best one, though I would go for openGL/CL now rather than latter. There is a growing mac gamers community (me included) that would benefit from that.

Rest assured that I plan to support OpenGL asap. But there are several things I must answer/consider before I can hope to successfully implement an OLG backend. First of all, I need people who know OpenGL! I know enough, at this point, to initialize an OGL game window and render some primitives... but not enough to write a high-performance renderer that not only does everything a game developer needs it to but shares a common parent interface with all of the other backends. If I have to learn OpenGL from start to finish and do it myself then all that time I spent didn't even get the work done -- just got me to where I could start lol. Secondly, I have to decide on either an existing OpenGL wrapper (e.g., Open Tao Framework), which could be of lesser quality than SlimDX and require modification, or writing a completely new one, which could take too long. Money makes the world go 'round, and time is money.

Anything that I can help on tips, just ask me. I cant help you full time, since Im in another time consuming project, and it appears that my uni has gone out of strike.

I sincerely appreciate this kind offer, and hope I can take advantage (in a good/mutually-beneficial way, I mean) of your skills. What sort of things do you know and what are you good at? I'm not demanding anyone work full-time on this yet. It would be insane for me to expect such dedication without paying you or otherwise compensating you with the same regularity with which you work. You will, of course, be named in the credits, get free software licenses and possibly other (even cash/stock) 'gifts' depending upon your level of input and what promises/agreements we made.

No, what I'm really looking for is part-time help. And I'm hoping some of those part-timers end up becoming full-time, paid employees on a salary. I'm looking for new and old talent to form the backbone of a development team and company, rather than being forced to bring in a bunch of complete strangers further down the road. Really, I want good friends... people I can trust... and to have their loyalty and dedication when serious money becomes involved. As I said before, the only requirements right now are an ability and willingness to learn, a passion for games/sims and some integrity as an individual.

[Bolkonsky]

I have the source for both of them on the desktop (recent checkouts have been there in some form for over 2 years), right at my fingertips. So I'm able to look at what's been done by other managed engines to learn from their successes (and failures).

It might be worth looking into XNA. It literally took me 10 minutes to set up an XNA engine that can convert heightmaps to 3d with shaders and everything. XNA is incredibly powerful for something of it's simplicity.

[keinmann]

It might be worth looking into XNA. It literally took me 10 minutes to set up an XNA engine that can convert heightmaps to 3d with shaders and everything. XNA is incredibly powerful for something of it's simplicity.

I've used XNA a lot in my years of programming. XNA is nice... easy to use, reduced boiler-plate coding, solid, cohesive framework, excellent stability... Problem with XNA is that it's built upon the extremely dated DirectX 9 API and there are no 64-bit XNA binaries. I tried convincing some folks I know at Microsoft to build some 64-bit binaries but they refused, claiming that no one "needed" them (enough). But it's the DirectX 9 back-end that's the biggest limitation of XNA. DirectX 9 came out in 2002, so it's over a decade old now (the last installment, 9c, is from 2004). As it might be 4-7 years before a game built with my engine sees the light of day, that would make DirectX 9, the technology XNA is built on, nearly 20 years old. That's why I'm not even supporting 9, and focusing only on D3D 10, 11 and beyond...

A game development project from TWC is something I've pondered for a long time. I'm sure there is the programming knowledge here somewhere that if combined could produce something decent. And there is no doubt this place is crammed full of artists and modellers and the likes.

The talent is certainly here. There's a lot of talent elsewhere on the internet as well. The reason most of the rag-tag "let's make a sweet game" projects fail isn't really a lack of talent (though a lot of them lack a real programmer)... they fail at organization and finance. Most of them have never heard of things like the iterative development cycle, never read a book on software architecture, don't know even the basic components/structure of a game program or engine and have no idea how to run a company and its finances. They just discuss a good game idea and start making models and textures... some get as far as loading some models and rendering them, but they all end up hitting a brick wall... everyone gets stumped, and wonders "How do we put this thing together?" I know, because I engaged myself in a few such projects in my late teen years.

But I'm "going for it" now... I've worked as a part of real development companies and learned the ropes from them. And the last few years of my new job, day trading, which I was forced into by the recession and lack of tech jobs, has been nothing short of "business finance bootcamp". I never would have thought when I was 18 that one day I'd be critiquing CEO performance, pouring over financial statements, watching conference call webcasts and following Federal Reserve minutes with eager anticipation, but here I am, lol. I have my own small business now, which I'm in the process of organizing into a mid-size conglomerate. And I want this to be a part of it. The trick is to make this branch of business self-sustaining, so it doesn't leech funds out of my financial/investment subsidiary. But I think things will be soon on the right track.

I've started very many game projects, finished very few haha. We see a ton of "lets make a game" threads on TWC, but just maybe this time it's started by someone with at least a hope of getting it somewhere lol.

Boy, have I been there too!

I can't help as I simply don't have the time, as you will find with most students we're usually already on-board one or more of these projects ourselves. Good luck, and I hope no matter the outcome you keep us updated!

I understand. But I appreciate the interest all the same!

[karamazovmm]

Im not a fully formed programmer per se (still in uni), but I do forward that the idea to use C# is the best one, though I would go for openGL/CL now rather than latter. There is a growing mac gamers community (me included) that would benefit from that.

Anything that I can help on tips, just ask me. I cant help you full time, since Im in another time consuming project, and it appears that my uni has gone out of strike.

[keinmann]

BTW, since GrnEyedDvl wanted a bit more detail (I guess concerning the architecture/design) of the engine I have something I can share. I made a diagram which demonstrates the architectural structure of the proposed multi-platform rendering system. Here is that diagram:



However, keep in mind that this is not the real hierarchy of the engine. It's simply a conceptual/abstract diagram to demonstrate the general idea. The real thing is not that simple. For instance, the D3D10 and D3D11 renderers share a common parent and also share access to the DXGI layer, which isn't even shown in the diagram. And as I haven't started on OpenGL support yet, there's no way to know how the real OpenGL subsystem will look.

[karamazovmm]

Im currently dedicating myself to database systems, which is indeed my project atm. I never coded for games, just for minor apps, that didnt need multi access systems, and several layers as well.

languages that I know C, C++, C#, Java, its in order of what I know best to what I know worst, Im currently trying to learn ruby.

[keinmann]

Im currently dedicating myself to database systems, which is indeed my project atm. I never coded for games, just for minor apps, that didnt need multi access systems, and several layers as well.

languages that I know C, C++, C#, Java, its in order of what I know best to what I know worst, Im currently trying to learn ruby.

Me and Java have never gotten along... haha

Do you have any interest in game development, just out of curiosity?

[karamazovmm]

Me and Java have never gotten along... haha

Do you have any interest in game development, just out of curiosity?

Its actually something that never sparked any interest in me

I do play games to distract myself, not as often as I did, still more than I would like

But learning things is something that I like, Im in college since 2004, there has been 2 degrees and Im going for the 3rd, probably will take a break and go for masters after this one (college/uni works differently here than in the US, its 4-6 years, and youve got yourself a bachelor degree)

[.Mitch.]

A game development project from TWC is something I've pondered for a long time. I'm sure there is the programming knowledge here somewhere that if combined could produce something decent. And there is no doubt this place is crammed full of artists and modellers and the likes.

C# is by far my favourite language and XNA is quite good if you are starting off in the world of game creation and want to make something quick.

I've started very many game projects, finished very few haha. We see a ton of "lets make a game" threads on TWC, but just maybe this time it's started by someone with at least a hope of getting it somewhere lol.

I can't help as I simply don't have the time, as you will find with most students we're usually already on-board one or more of these projects ourselves. Good luck, and I hope no matter the outcome you keep us updated!

[keinmann]

Here's a bit of an update on things...

The new engine solution has grown to over 30 C# classes and interfaces in the last two days, when I literally started the new project structure. A few have been imported from past prototypes but a lot has been written from scratch. Each file typically runs from a couple hundred to a couple thousand lines of code... so that's a lot of coding I've squeezed in during the last two days! The project does build (compiles into a DLL) without any compiler errors, and the only warnings being emitted are ones I hard-coded into the source so I'll be able to go back and change a few things I want to improve upon.

Most of of the work I've done so far focuses on the HAI (Hardware Abstraction Interface) portion of the engine, for managing different types of hardware/systems, graphics settings, etc... The DXGI (DirectX Graphics Infrastructure) layer is more or less complete. It incorporates some very powerful classes that expose information about the local machine's capabilities, specifications and features. I stubbed out the same classes for OpenGL but they have no internal implementations. Among others, there is a "GraphicsCard" class and "Display" class which encapsulate physical graphics cards and monitors installed on a system and allow communicating with/polling them. There are lots of useful classes like this for finding out everything from the amount of RAM installed to the number of CPU cores to the optimal/native display mode of multiple monitors.

Now I've moved on to the actual implementation of rendering back-ends. I've just finished the vital graphics settings and DirectX10 initialization code, and now I'm working on a seamless way to tie this and other rendering back-ends together through a unified graphics manager class. Hopefully I'll have that hashed out in the next few hours and I can then implement the game window base classes. Then the engine will be able to launch, initialize and start pushing tris and pixels (rendering, that is).

EDIT: Also wrote a type of asynchronous task manager queue which handles batches of assigned tasks on a separate thread (yes, it's a multi-threaded engine).

[Bolkonsky]

Sounds good, seems like you're well on your way to a solid pipeline. Take a screenshot of your first triangle.

Any thoughts on making the engine moddable? A simple .ini file that tells the engine what folders to check for extensions, and load, would be a simple yet highly moddable option. Do you have an file format plans, too, for resources?

Dammit, you're getting me in the mood to try my hand with XNA again.

[keinmann]

Sounds good, seems like you're well on your way to a solid pipeline. Take a screenshot of your first triangle.

I will do that.

I was hoping that first tri would be tonight, but I'm out of commission with an absolutely MASSIVE toothache from the pits of hell. I'm drooling Oragel everywhere and eating oxycodone, and still I get no relief... very deep, deep pain. So I just hope I can eventually get a good night's sleep, and will be praying I wake up pain/swelling free to continue working on it...

Any thoughts on making the engine moddable? A simple .ini file that tells the engine what folders to check for extensions, and load, would be a simple yet highly moddable option. Do you have an file format plans, too, for resources?

I do intend to be as mod-friendly as possible without enabling blatant cheating/hacking. I don't want to make any promises yet, as getting this engine working properly is the most important thing, but I'd like to eventually give it the ability to load user's own C# assemblies against the engine to modify its behavior.

Dammit, you're getting me in the mood to try my hand with XNA again.

They have a surprisingly well-made racing 3D racing game demo that came with the last XNA installment I downloaded. It was so good that I'm sure they've kept it as part of the package to show what XNA can do. I bet it would be fun to extend/mod.

[.Mitch.]

Are you using an SVN or somesuch? I guess this is a big ask, especially as you seem to plan this in a business sense and I myself personally hate sharing code , but I'd love to take a look at your code and if nothing else check in on its development from time to time.

[keinmann]

Are you using an SVN or somesuch? I guess this is a big ask, especially as you seem to plan this in a business sense and I myself personally hate sharing code , but I'd love to take a look at your code and if nothing else check in on its development from time to time.

SVN is ready to go and will be used (it's a necessity when working with multiple people), but as of now the engine exists only on my PC. I won't use a full-blown SVN server until there indeed are other programmers needing to work on this.

This isn't going to be an open-source project though. However, I wouldn't mind tossing you a copy of the "bin" folder (binaries) when I start making some test builds. That'll give you a better idea of what the engine can do (and let you see it in action) as opposed to months of reading code.

Its actually something that never sparked any interest in me

I do play games to distract myself, not as often as I did, still more than I would like

But learning things is something that I like, Im in college since 2004, there has been 2 degrees and Im going for the 3rd, probably will take a break and go for masters after this one (college/uni works differently here than in the US, its 4-6 years, and youve got yourself a bachelor degree)

I see. Where exactly are you from?

[karamazovmm]

I see. Where exactly are you from?

brazil

[Musthavename]

Are you using an SVN or somesuch?

In case you aren't currently, I can't second this enough. Even if you do everything yourself, it's just so useful to have a project under source control. The most important thing with a project of this scope is that you plan it out as much as possible beforehand. If you're also planning on taking other people on to help, get some common coding standards and ensure everything is documented. It will really help in the long run, and you'll need rigorous documentation if you want this engine to be useful to anyone else, so it's far easier to do it as you go.

Frankly I'd say you should be doing this in a native language, but you're pretty convinced otherwise, so I won't try to argue the point (that said, I program C++ for a day job, so take of that what you will). I'd also be as desperate as possible to get away from anything even bordering on Microsoft specific. I can't fault C# too much as a language (or at least as a managed one), but even with Mono's existence I'd rather not pick up a language owned by Microsoft for fun.