GAME ENGINES AND OTHERS: PART 1 OF 3

There goes the weekend. I hope you have enjoyed it. Now it's time to get back to reality ... Mondays = kill-joy ... For some.

Remember last Friday's post? I had commented something about game engines, right? Ok, this is the first part of a small series of posts regarding game engines and other components like: game frameworks, game-oriented languages and game-oriented middleware.

As soon as you try to join the game industry you will run over a bunch of software components which may help you build a videogame from scratch, to some extent. Are all of these components a "game engine"? Quick response: "No". Why not? To answer that, let's first take some due notes of the following concepts:

1. Game Engine: a software component that provides the core functionality that will help you implement "the whole" programming areas of the videogame.
   
2. Game Framework: a suit of robust software and tools that offers a set of rapid and standard solutions to common program requirements and problems, all related to videogames.
3. Game-Oriented Language: a language specifically built to let you program videogames with ease -or at least, (a bit) easier/faster than by using "general-purpose" languages.
   
4. Game-Oriented Middleware: a software component that provides the core functionality that will help you implement only "a part" of the whole programming chain of the videogame.
   

The purpose of all of the above-mentioned type of "helpers" is to simplify the development and programming of videogames and, in general, to provide some sort of "re-usability".

By doing so, they usually bring at least one implementation layer between your final game and "low-level" libraries (like "DirectX" or "OpenAL"); generally speaking, "low-level" libraries provide access to computer/console hardware through a set of basic functions: take for instance "Application Programming Interfaces" (API) like "Direct3D" or "OpenGL", which can be considered as software abstractions of the "Graphic Processing Unit" of videocards (GPU), thus letting you handle and build "rendering systems".

Please, bear in mind that those 4 concepts are just one general and simple way of categorize the software components that you may find out there. Meaning, you could bring more categories to the field, rename or re-arrange the provided ones, but the important thing is that you realize that when you start to move down "the logic tree" trying to narrow general/base concepts, specificity may be found, and in turn, new categories may appear.

For instance, "Middleware" could sound too misleading for some since the word it-self wouldn't necessarily imply that it deals with just a part of the game-programming chain (contrary to "the whole"), but it'd imply that it brings at least one layer between your game and a low-level API (thus, some could deem "1" and "2" above as more general middleware).

Again, it's up to you to build your basic list of categories and place each component within them either with a great amount of certainty (for "obvious" choices) or "at a rough guess" (for "not-that-obvious" ones).

1. Game Engines.

If you remember one of my previous posts, when you want to program a video-game you will have to deal with lots of main logic areas: Rendering Pipeline, AI, Physics, Scripting, etc.

Game engines come at a hand to such purposes. They bring programmer-friendly solutions, methods, and techniques which will let you face the task of implementing each logic area according to the program requirements of your videogame.

In other words, a game engine will provide you with the proper software components so that you can program the core aspects of your game in a rapid and reliable way. This doesn't mean that you will press some button and, say, the physics of your video-game will be "auto-magically" programmed for you. Following the physics' example, it means that you'll have a group of pre-built functions that will help you set the proper physics' behaviors for your video-game, but how you assemble the puzzle is your Company's task, you team's task or your task.

What kind of game engines can be found? Well, basically there are 3 type of engines: (a) general-purpose engines, (b) game-type-specific engines, and (c) game-bound engines.

(a) General-purpose engine: offers a set of functions so that you can develop and program any type of video-game.

(b) Game-type-specific engine: offers a set of features so that you can develop and program a certain type of video-game. For instance, a "First-Person-Shooter" game (FPS).

(c) Game-bound engine: created for a particular "title" -or series. Its use is normally considered as "one-time" even though sequels of the game may appear, which use a modified version of the base engine.

Most of the (currently available) game engines were developed to be use with C++ language, which is the standard language for game development (alias, the language used by 99% of each and every "Pro"). They are usually deployed as a set of "dynamic-link libraries" (DLLs), or add-ons to the most-used IDEs (like Visual C++, Dev-C++, etc.), so that they can be easily referenced, integrated and used in your game's implementation.

Sometimes, it's curious and instructive to see how these engines have evolved during years. One may think that it all starts with a general-purpose engine and then everything moves in a top-down direction. Makes sense. However, one may go wrong sometimes ...

For instance, GarageGames' "Torque Game Engine" was originally designed for the game "Tribes 2". After that, the engine went public so that developers could use it to build "Tribe-like" videogames for a minimum price which varied as a function of the kind of licenses offered.

Since then, the engine has been modified in such manner that today is one of the most used multi-purpose cross-platform game engines among "indies". What is more, it can be considered a game framework since it offers a whole set of tools that help developers leverage the production pipeline.

Therefore, GarageGames' TGE moved along the whole range of types of engines, starting from "(c)" and upto "(a)".

2. Game Frameworks.

"A suit of robust software and tools that offers a set of rapid and standard solutions to common program requirements and problems, all related to video-games". Meaning?

Game frameworks usually go beyond than game engines do, in the sense that they not only provide the libraries you need to carry on programming tasks with ease but also bring a set of tools into play which help you develop parts of your game in a graphical manner.

By the use of "Graphical User Interfaces" (GUI), the developer enjoys a more controlled experience over areas of programming and design, generally resulting in a more comfortable interaction with game APIs. Although it cannot be always assured that desired results are reached thorough the only use of GUIs, some frameworks also offer the possibility of either using scripts to do a fine tunning of the final output and/or extend the game logic, or accessing the final source code before compilation.

Some years ago these frameworks only consisted of either an add-on to existing languages (the game engine) plus a game-level editor to create new levels for your game. Now-a-days, they have evolved to more complete systems formed by several tools which offer more features so as to produce more accurate results with less lines of source code manually inputted by developers.

Levels construction, pre-calculation of lights and shadows, creation and visualization of meshes and animations, setting of game characters and their respective behaviors, general rules for the game (like collisions and physics), integration of cut-scenes, shaders, music and sound, ... these are just a few elements in the list of features offered by top-most game frameworks.

3. Game-Oriented Languages.

What is the difference between a general-purpose language and a game-oriented one? Simple. With the former you can usually program any kind of software -limitations may appear depending on the language. Take C# for instance, with it you can implement from graphically-appealing applications to console command-type programs. With the latter, only video-game related software can be produced -some exceptions do exist, where multimedia applications can also be created.

Does this mean that I can only use game-oriented languages to create a videogame? No, you can also use general-purpose languages -like C++, C#, Visual Basic and Delphi- by plugging in add-ons, class libraries, game engines and or game-oriented middleware, where available.

Why using then a game-oriented language? That is a matter of assessment, resources and preference. These kind of languages offer their own syntax, commands and structure. Not all of them follow the Object-Oriented Programming Paradigm (OOP) and some of them sometimes consist of a wrapper of a base API (like DirectX) with some moddifications and additions to ease the user experience.

What is the difference with "scripting"? A script is interpretated at runtime. Game-oriented languages usually provide a compiler that converts the source code into machine code at compilation time. Also, many of these languages provide an "Integrated Devleopment Environment (IDE) which consists at least of an official source-code editor.

Should I choose an OOP-tailored language? Again, that is a matter of preference. Some may still find some comfort on using non-OOP languages, but believe me, once you start to understand OOP and appreciate the advantages of OOP languages, you'll use OOP and you won't want to turn back the page again.

4. Game-Oriented Middleware.

Contrary to the criteria that leads a component to be considered as a "Game Engine", this kind of middleware specializes in a certain "logic area of incumbence" of the game-programming chain. For instance, AI.

There is a new growing wave of enterprises that specialize in the production of middleware components that one can "hook" onto their game projects. One of the main advantages of this components is the strong set of features they provide plus a high grade of reliability. One of the "cons" is simple, the more number of components you use in your game project, the more difficult will be for you to assemble the puzzle and follow the evolution of each component. Moreover, performance hit should be monitor closely since using a great number of external components does not necessarily mean that performance will increase, a priori.

The main areas covered by game-oriented middleware are rendering, physics and artificial intelligence. But new specific solutions to general problems are appearing from time to time.

Again, like "game engines" do, game-oriented middleware is generally targeted to C++ market. Some wrappers may exist for other languages/technologies, though, generally provided by third-party enthusiasts. Therefore, if used in a project, these wrappers should be always used with caution, having analized the whole spectrum of options.

Ok, I have read carefully the 4 points mentioned above, but what should I consider so as to choose the proper game components for my game project? Good question: it will be covered in next posts, but here is a short list: price, licenses (royalties, terms and conditions, etc.), support, bugs, average period between updates, performance, features, target platforms, type of deployment/integration, type of files supported (for images, meshes, animations, etc.) and so on.

"I've read your interesting post but I cannot wait for the next 'part' of the series, so where can I find more information about game engines and such?". Impatient, uh!? Ok, you can then pay a visit to "DevMaster.Net" site, where you will find a very extensive and detailed database of engines -most of them ranked by users.

Soooo ... with these words we are reaching the end of this first part about game engines and others. I'm not sure so far but I guess this will turn out to be a three-part series. ETA for "Part 2"? Maybe next Monday ... but as usually said "it will be released when it is done" ... ;)

[We will continue to discuss game engines on later posts, focusing on some of the existing/upcoming engines for .NET]

THERE AIN'T NO TIMES LIKE THE OLD TIMES

Today I was thinking of posting some comments about some of the engines out there, but then I said "naaa! It's Friday. Let's take it easy and leave that for Monday".

So, what's for today? Doing some search for new posts in a couple forums I found some references to "old" games and I couldn't help remembering the old machines, old games, old sweet days. Snif! Snif! So let's go retro ...

Going back to the mid 80's I remember receiving my first computer ever: a ZX Spectrum Plus ... aahhh! My first love. Responsible for my passion on videogames. Titles like "Pssst", "Atic-Atac" and "Game Over" come very fast to my mind. Those were the days when you had to wait like 5 minutes or so until the game loaded into memory (48k ... wow!) while you prayed not to receive that nasty error message that appeared when the cassette player emmitted a wrong sound. Remember that? I bet you do. Then without loosing your hope, you used to rewind the tape, press "play" so that everything started again.

Then I switched to a "Commodore 128D" with a 5 1/4 drive. A whole new world for me. Just the fact of not having to use a tape was terriffic for me. Games like "Dig Dug" and "Bruce Lee" helped me tolerate those boring afternoons after high school when there was no one to call, no football matches, not even homework.

Then came an "Amiga 2000" (without hardrive I regret -too expensive back in those days) but again with an unknown hardware for me: a 3 1/2 floppy drive (I know what you're thinking: "what a challenge!"). This was a big change regarding visual quality: the graphics, colors, the cut-scenes. And also regarding sounds and music: my favorite? "Shadow of the Beast" (that guitar track that played when the game was over was really enjoyable). Games like "Prince of Persia" (with those impressive movements of the main character), "Lotus Espirit Turbo Challenge" and "The Blues Brothers" (a very funny platform game) caught my attention for hours and hours.

Finally, I bought my first PC computer: a notebook "Presario 1080". After that I realized something "I love notebooks". I have to admit that desktops are in general more powerful and you can get a better desktop for the same price or less than a notebook (especially if you buy the part separately and assemble the machine by your-self), but notebooks are quite more handy. With time I started to value more "handiness and comfort" than "hardware power" (yeah, I know, to some extent, since it has a direct relation to your needs as well as your purchasing power). We all know PC games but the first I played was "Pod" (since it came with my notebook), then "Primal Rage", "Alone in the Dark", the "FIFA" and "Star Wars" series (Rebel Assault, Tie Fighter, etc.), among others.

Now-a-days, I've got a 64-bit-processor desktop with the power to run smoothly each and every newest PC game in the market (like one of my favorites "Prince of Persia: The Sands of Time") but I still enjoy playing classic retro games, every now and then.

Well, that's all guys. Have a nice weekend!

[By the way, which are/were your favorites "retro" games?]

MANAGED DIRECTX AND WINDOWS XP PRO 64-BITS

As most of you may know, there is no MDX support for x64 platform yet. So, in order to run any MDX program the platform target should be set to "x86" (that is, 32 bit).

For users of WinXP Pro x64 that want to run Managed DirectX code with Visual Sudio 2005 (also applicable to the XNA framework) -any version, take due note that there are two possible ways to avoid the nasty "Bad image format" exception.

As mentioned in Benjamin Nitschke's blog, the first one, is by modifying the "csproj" file with the notepad. Just locate and open the file and then add the following line:

<PlatformTarget>x86</PlatformTarget>

(if the "node" already exists, just replace it with the above-mentioned one).

The second one, is by using VS IDE it-self. Following this link for further instructions.

Stay tuned!

SAY AGAIN ... GAME DEVELOPMENT?

Now that you have done some deep research on OOP paradigm and C# programming we can approach a bit further to the main topic of this site. Shall we? (not there yet but getting closer, be patience)

Well, the main three questions that you should ask your-self before jumping into deep waters are:

1. Which is the story-line of my game?
2. Will I design and program the game engine by my-self?
3. How the h#$& am I going to finance this project?

And yes, this is a short list.

1. Every game -even the simplest of all out there- has a plot, a story, a script; something that tells the player what the game is about, what he/she might do during game-play, the characters/roles that do exist, etc.

If you think it thoroughly, you may end up finding a bunch of similarities with filming: you prepare the set, you place the characters, you establish the rules, you are the director. The main difference is that game development goes some steps further since everything happens real-time and not in a (fully) deterministic way (meaning, every time the game is played things always differ sometime, somewhere, somehow).

But the point is: your game has to count on a script. At least some sketches of a plot (general idea), spots (places in the universe of the game), characters (who's the good, the bad and the ugly), missions ("take the flag"), stages (level 1, ..., level n) and any other objects/subjects of relevance (like key dialogs and possible cut-scenes, if any).

Lot of games have soared because they lacked of a (good) script. Technically speaking, these games may have introduced advances in the fields of game programming, superb graphic effects (GFXs) and astonishing sound effects (SFXs), but gamers flamed it -or even the game never shipped, because of a bad-designed, repetitive (monotonous sequel) or non-existing story-line.

It is always educational to read post-mortem articles of abandon-ware and finished products. Also comments of gamers about published games -AAA or not; you will be surprised to find that sometimes (rarely) a game succeeds even though GFXs are not that good.

2. Before starting to program the game itself there is always a "Game Engine" that lays beneath it. So, what is this "engine"? Think of it as the logic structure that will support the implementation of your game.

You will need to handle lots of aspects like, among others:

• Loading/unloading assets to/from memory: meshes, levels, sounds, etc..
• Rendering process: fixed pipeline vs. Shaders, lighting and shadows (static vs. Dynamic), post processing effects.
  
• General mechanics (the so called "Physics"): collisions, gravity, rigid bodies, dynamic action, reaction and interaction of objects (cars, spaceships, .
• Artificial Intelligence ("AI"): finite state machines, neural networks, flocking, etc.
• User Input/Output: mouse, keyboard, joysticks, (force-feedback) game-pads.
• GFXs: particles, billboards, water, clouds, skies, backgrounds in general, day/night cycles, lens flares.
• SFXs: mono/stereo sounds, and 3D sounds.
• Music: start/finish loop, midi or pre-recorded?
• Story: cut-scenes management.
• Main loop process: one thread vs. Multiple threads.
• Gameplay: single-player vs. multi-player.

Ufff! I'm just tired of simply mention it, mate. You can build a generic engine (it can be used on other game projects of different types) or one specific to your (type of) project (it can be only used to produce, say, racing games or sequels of your game). Then comes the game programming: type of game, menu of the game, levels of the game, etc.

So there are two stages of implementation: (I) the game engine, and (II) the game.

The good news: you may not need to program a game engine and concentrate on the game itself. The bad news: even so, you will still have to deal with the above-mentioned areas. It's important that you know in advance that an engine will not handle all things for you, but provide you with "pre-made" implementations that you could use in order to handle those things. So where's the catch?

There are plenty of (full or partial middle-ware) engines out there you can integrate to your projects and use as a base for your games. The problem is that you have to integrate them properly, like assembling a puzzle. And this is not always that easy. As we will see in future articles, not all of the engines available in the market:

• Handle all of the areas of a game as you wish (so you may end up bringing together multiple "partial" engines/components or implemeting by your-self the ones that there is lack of),
  
• Target the .NET technology (you may end up using a third-party wrapper or creating your-own), and
  
• Are free and/or royalty free (very important point).
  

Therefore, it's a trade-off:

1. If you decide to design and implement (all of some of the aspects of) the game engine by your-self you could get trapped in something that either is out of your league, takes so much time that when your finish it you have not time/money left to implement the game or the standard technology has changed in the industry, or you decide to quite your original project and enter the game-engine buisness (that is, you offer your engine to the market in different flavors and licenses as many Companies do), and
2. If you decide to use third-party engines -since there has been lately a boost in the number of engines available, you will have to choose carefully the one suitable for your project after a lot of testing: performance, bugs, support, how easy it can be integrated to your project (programming language, technology, etc.) and last but not least, price plus terms and conditions. Switching engines in a middle/advanced stage of the development process of a project is always very expensive in terms of time, effort and money. The nearer your project's deadlines are, the worse the price your project will pay ...
   

As any trade-off, a bad decision could result in "no shortcuts" for you and your team.

3. Tough one. Maybe you have resolved both two above with superb ease, but let's face it: do you have the needed resources to back up your idea from scratch to completion? Or do you know someone willing to finance your "Know-how"?

Game development is not a game. Period.

If you are (just about to get) in, you'd better realize that in order to succeed -either as a "pro" or as an "indie"- you will have to work full time on the project, or you will be eventually getting out of the industry with your tail between your legs.

And by "full time" I mean that you cannot share your time with other jobs if you expect to finish and ship your creation this century. Exceptions exist but it's a general rule.

Thus, the problem is: you have an idea, you believe in your idea, you have played tons and tons of videogames so you know what gamers want ... ok, basically, you feel in your gutts that your idea will rock ... now, "who is with me?" (you see that your shadow start to run away from you ... sad).

You need to feed your-self and -in many cases- your family. Plain and simple. So it is hazardous to quit a job to pursue a dream. So, unless you abandon your idea of being and "indie" and start looking for a job in the industry (or the worst-case scenario for you: getting back to your previous job totally defeated), you better start looking for ways of raising money to support your project. What in fact you should have been done/assessed before jumping to the swimming pool (aways look for water first).

There are many ways of accomplish this task, like:

1. You have a job for years so as to save the money you will need (or that you think you will) to go on "solo" later.
2. You start a joint venture with other people with the same goal (team), sharing risks, costs and profit, proportionally.
3. You managed to find someone that know someone else that knows how to contact a final someone else who -maybe and almost remotely- could think of invest some capital in your project and are in the talks with this visionary ... ejem! ... businessman. In general, the percentage of your final proceedings in this situation will be as low as it could get (you knew that, of course). Risk-reward law.

There is some sites with interesting figures on their projects' costs and profits.

Dissapointed? Don't. Just bear in mind that nothing is as strightforward as it looks at first sight. Being and "indie" is always complicated. You don't have access to the same quantity and quality of resources that big companies have. It doesn't rain money and you have to eat. Time won't wait for you. A year and a half has gone, you are still not even close to a beta and the next season of "Lost" is coming ...

However, in spite of all of the above, you got the strenght to overcome these assymetries, right? That is one of the main things that will help you achieve your goals: your conviction. A needed but not sufficient condition. You will have to combine it with other factors like, among others: your capacity to solve problems efficiently, your speed to accommodate to changes in the industry (like techniques, technology, game-quality standards and so on), your skills as a developer/programmer, your wisdom to deal with potential investors and customers, your means to promote your game worldwide, the moral support of your family and close friends, and your growing collection of four-leaf clovers!

It always start with a dream ... now, let's check on the results of those six number I picked for yesterday's Loto ...

[Next time we will be talking about some of the .NET-based game engines out there so that you can go and check them]

LET'S GET STARTED

OK, this is how the deal goes ... I will be creating a new entry in my blog from time to time "after hours" (when my eyes want to pop out of my head) ... So with the exception of today, you will find no more that one entry per day (and I promise I won't add a new entry every day, I have a life, you know) ... And yes! I'm not an English-native speaker so please forget any grammar and spell errors "I certainlywill make" (I know, I know, just kidding: "I will certainly make").

Having said that, we can now move onto the first topic.

If you read my previous entry you will see that I claim that sometimes resources (in any form of documentation, white-paper, book, source code, etc.) in some way assumes that you already know certain concepts, techniques, programming languages ...

Thus, if my blog will mainly deal with game development/programming with C#, what should you know first to properly follow and understand those resources? Think! Think! ... Tic-tac, tic-tac, ... Yes, you're right ... C# (pronounced "C-sharp" ).

1) What is this so-called "C-Sharp" language? In short, it's Microsoft's flagship language.

The long answer, it is a high-level language based on MS framework that deals for you with, among other things:

• Memory management (say 'goodbye' to pointers and function pointers and say 'hello' to the garbage collector delegates and so on, but of course, you can still use pointers with C# if you want, not to mention with C++.NET).
• Translation of source code to MS Intermediate Language or "MSIL" (now, executables and dlls are no longer machine code after "compilation" and no matter which .NET language you choose, you can reference and use assemblies created with other .NET languages with ease).
• Proper execution of assemblies on clients' machines (this will make sure that an MSIL assembly is "natively" compiled to machine code on the machine that executes/calls it in order to avoid the problems of running programs pre-compiled on third-parties' machines).

Regarding the language structure, talking in a dirty way (you are over 18, right?!), C# is a sort of mixture of aheem ... va! (sorry for that but my flu is killing me) and C++. Let's say that it lays between those two. Sort of ...

2) Being C# developed by MS, does this mean that C#-based assemblies will only run on PC computers with Windows OS? No, there is an open-source project called "Mono" which provides the necessary software that let you develop and run .NET applications on a multi-platform environment (Linux, Solaris, Mac OS X, Unix and even Windows).

You can use the "Mono-Develop IDE" to compile and run any assemblies you create for .NET through Mono's framework.

Off course, you can use Visual Studio 2005 to develop on Windows platforms, either on the commercial flavors or the free ones (called "Express Editions"). You can also use an open-source, and yet powerful, IDE called "Sharp Develop".

3) What makes this C# language a value asset for programming? Glad you ask.

Friendly, portable, powerful, fast, with lots of features that help you program ... And as you will see, as long as approach to game-related topics, that it not only a language suitable for applications (for which it is mostly used nowadays) but also for games.

Coming from Java? Not a problem. You will quickly find your-self programming with ease.

Coming from C++? Not a problem either. You will feel a great amount of relief when you notice that you can stop dealing with pointers. Speed? Almost the same under certain circumstances and growing (it will eventually match C++). Differences? Heritage is only allowed on a single basis. Delegates replace function pointers. "Generics" resembles "Templates". Etc., etc.

Coming from any other sort of basic and non-basic language? You will find that it is not hard to learn, even if it is the first time that you get your hands on Object Oriented Programming ("OOP").

4) Which versions of the MS framework are currently out there? Good question.

You have three main versions:

1. .NET Framework 1.x: the one that started everything. Many applications still use it.
2. .NET Framework 2.x: it has introduced one of the greatest features you will find: "Generics" (some friend!).
3. .NET Framework 3.x: formerly called "WinFX". Currently, it's under a beta stage and will ship within "Windows Vista" next year. It will change the way you use the components that form the Framework it-self (from now on referred as "Foundations") . For instance, the "Window Presentation Foundation" will replace GDI (everything will be based on DirectX), and will introduce "XML for Applications" or XAML (pronnounced "Zamel") as a way to create GUI's.

Please take due note that from .NET Framework 2.x and on, you will also find installation versions for x86, x64 and IA64 machines (i.e.: if you have a x64 Win OS and an AMD 64-bit processor, go ahead and install the version for x64 computers).

5) Where can I find books, tutorials and any other resources to get my hands on C# programming? Interested, uh?

For books, always visit vendors like "Amazon" and "Barnes and Noble", so to speak. Use the search for references like C#, .NET Framework, and so on.

And for tutorials, you can visit places like "The Code Project", "C# Corner" and "GotDotNet".

But always visit the main source: "MS Developers Network" and "MS Forums".

Well, to finish off today's entry (... yes, please ...) I could spend a lot of weeks and even months to talk about C#, but it is not the point here since you can see you shall find plenty of resources out there. I just wanted to give you a brief explanation of its main characteristics as well as to point you the right places to find resources.

If you aren't a newbie then it's most likely that you already knew all of the above-mentioned information, but if you are a newbie I hope it helps ... :)

Just remember, the use of C# is growing faster and faster among "indies" (we will talk about this in later posts) and I believe it's a matter of time until the game industry adopts it as a standard. And believe me, it will ... ;)