[ButchDean]

Might sticky this depending on response but I am curious, as many probably are, as to what makes programming games particularly challenging for you?

As those who have attempted to make games are aware, you must learn a programming language of some form or another, then learn the associated SDKs and other tools like DirectX for instance. This seems to cause many to give up at the first hurdle even before beginning to write the game itself. Secondly, serious game projects take an eternity to complete and sometimes this is the reason why some give in.

So, what particular challenges does game programming present for you? Also, what is your planned strategy for overcoming these issues?

[Uroborus]

I think you touched upon one of the most difficult aspects: Time. I personally have never completed a game project, and this fact is a symptom of what I believe to be the most difficult part of game coding: Complexity. When I say this, I am not only referring to the coding, which, from a software engineering standpoint, is by far one of the smallest obstacles. And I am not even referring to the difficulty in learning the SDKs, languages, etc. Todays game production industry is analogous to the movie industry in the number of people required (outside of routine business maintenence for the development shops) to make the game itself. Just look at the credits for any commercial game product, from mobile games to full fledged high end titles. For a developer (or small group of developers) trying to break into the industry, I think that, because there are a lot of tasks to be handled specific to the game design itself, complexity of the project blows most attempts out of the water. Games aside, as far as commercial development goes in general, failure rate estimates on from-scratch projects of all kinds range from @30% to @60% depending on the source, and these include trained teams with budgets working full time. All of this is daunting, especially for start-up projects.

More specifically, another difficult and oft overlooked problem is art design. Programmers know other programmers, and people involved in various aspects of technology who are not programmers. Many IT people in general that lean on the visually creative side of things head toward web development or even print media. Modern competitive games need excellent 3D (or even 2D) work that can be completed to specifications on time and on budget. It sure seems that quality artists working in the games field are by far fewer than programmers and designers. Several friends and colleagues who have attempted projects in the past have been stopped cold by this one.

Good post, thought-provoking question!

Uroborus

[chinchang]

The most difficult aspect of programming a game is to organize the code as the game gets bigger and bigger so as to make it extensible and easy to maintain.

Another thing which is kind of a challenge is to program the game specific characteristics. Like almost all platformers have a jump function for the character. But even the simple thing like jump has to match the game theme. It can be a jump of a sumo or of a ninja This is very important while implementing the gameplay and overall feel.

Last thing I face difficulty with is the AI. It still remains a weak point of my games.

[westforduk]

This is quite an interesting link about finishing games, which is probably the hardest part, getting enough interest and motivation and sticking with a single concept and finishing it to a high standard.

http://makegames.tum...inishing-a-game

[xclite]

As Uroboros said, gathering the skills required to make it happen is the most challenging part. I know I'm no artist.

[ButchDean]

Some excellent points made here. Will address each of them properly later when I can give them all of my attention.

[atraub]

Level editors man... I can make my sprite run around the screen and do some cool looking stuff, but I can't create a level editor to save my life. I can conceptualize it and sketch it out but I just can't seem to pull it together. Without any sort of world to put my character in, I don't want to implement any sort of physics. In the case of my last game attempt, Revenge, the entire project was shelved due to my inability to make a level editor. Ultra Mega Man (my avatar) was going to be a central character. I sometimes wonder if he'll ever get the on-screen debut he deserves.

You're bringin me down ButchDean.

This post has been edited by atraub: 03 February 2011 - 07:58 AM

[Kilorn]

I think one of the most challenging aspects of game development is creating something that is going to be fun and not get incredibly stale or repetitive after a few hours of playing. This is always an ongoing struggle. On one hand, you want to make a game that's complex enough to keep it interesting for an extended amount of time, but not too complex as to require a 2 year learning curve before you can begin to understand how the game works. Obviously that was a huge exaggeration, but it gets the point across. The bottom line is, if a game isn't fun, then nobody is going to want to play it and if you can't keep it interesting, then people will play it for a very short period of time and get bored with it rather quickly.

[Fib]

If we extend this discussion to the most difficult part of creating a game, then I would say the overall design. The whole point of a playing a game is to have fun. I have played way too many games that have amazing graphics, superb AI, smooth gameplay, but are simply not fun.

The most difficult part about programming a game depends upon if you're by yourself, or have a team.

If you're by yourself, I think the most difficult part is learning all the skills necessary to program a game, and staying motivated. You have to become fluent in a programming language, be familiar with a graphics API, sound API, advanced math, optimization techniques so you're game doesn't run slow, making sure you're game will run on all kinds of different graphics cards, code organization so you can actually understand your own code, ect. The list could go on and on.

I think there are different difficulties to face if you're leading/apart of a team. But I've never done either so I can't really say.

I'm very interested in reading what ButchDean has to say because I know he's been working in the games industry for a while.

[Kilorn]

One of the hardest things about working on a team, is making sure your teammates actually pull their weight and do the job that they're supposed to do.

[stayscrisp]

I think resisting the urge to put everything you think of into the game rather than sticking with polishing some core parts is my main problem. It's fine when I'm just prototyping but I have to resist the urge to do loads when I put a project together.

Also making sure to keep your code nice and OO (if you code that way) and not chuck things in just because its less work and gets you a quick fix, it almost always comes back to bite you.

[Kilorn]

Most people refer to that as "feature creep", and I've been guilty of letting it get to me as well. It's far too easy to think of new things to make the game better, but it's a lot harder to push a deadline and justify a feature that may or may not add a lot to the game in the long run.

[Mila]

Before we begin, a disclaimer: I have never actually finished writing a game. But then again, is a game ever really finished? Philosophical questions aside, I have been diligently working on one particular game since September 2007, with no end date in sight.

My biggest problem with any large application, be it a game or a very large, involved piece of software is not simply time, but rather time management. When approaching the problem of developing a game, you generally want to jump right into the "fun" stuff: shooting things, graphics, and other mainstay gameplay concepts.

The problem is, you can't. Before you can start creating monsters and super-cool weapons, you need a platform to build them on. Before you can build your game world, you need to define it. Before you can define it, you need to build utilities. And so on. When you start a project with this great vision of a 3D first person shooter on a jungle planet with vicious Ice-Age-like wildlife and hostile aliens, you really don't want to waste your time on figuring out the best dungeon/jungle/building/whatever generating algorithms with cellular automata.

I've found that the best approach is a semi-Agile development one, complete with user stories and iterations, even if you're both the client and the contractor. This way I can always guarantee something is happening, and by taking "small bites" as it were out of the bigger problem progress is always happening.

Of course, seeing how I've been working on my own game by myself for almost four years, I might not be the best person to listen to.

One final tip. PLT Scheme is very very bad for writing games. In case anybody was really really bored and wondering.

This post has been edited by Mila: 03 February 2011 - 11:37 AM

[ButchDean]

Well, there is certainly one resounding theme that I can see from all these replies, and that is not necessarily taking on projects that are too big for you to complete (although this can be true for many cases), but in trying to complete the game project in as few steps as possible making each step overwhelming.

It is important to not that game programming is not easy even for the pros, because once you master one aspect there is always an enhancement to that aspect further down the line in creating a more advanced game that you must learn to get your head around more or less starting at square one. There is a saying that I recall hearing quite a while back that programming is 10% skill and 90% determination - and this is more than true for game programming. I personally have seen many otherwise promising and very clever programmers steer well clear of games because they immediately have a good idea of what's involved.

Rather than quote each of you I'll address what you've posted individually.

@Uroborus - Firstly, welcome to DIC! From the perspective of the work that you've attempted it is clear that you have simply chosen projects that are way too big for you, hence the reason why you've never completed a game. Kilorn recently made the excellent point on another thread somewhere that it is better to have a number of smaller completed game projects then larger incomplete ones. I actually learned this the hard way. After completing my first 3D game I found that I was getting into that habit, now, with a new found determination I'm applying my experience to sticking to a single project that will stretch me, but I know that I can finish. Don't be obsessed with the stats regarding staff and failed vs completed projects because you don't have those challenges resting on you with your projects. Just focus on getting the job done.

@chinchang - Every game developer strives for perfection, but obtaining that level of perfection takes a lot of experience. Experience also teaches you sometimes perfection cannot be achieved and you need to work around it. Don't stress over things like the specifics of AI, just make the game feel fun however you can even if it doesn't play exactly to how you intended. Are you aware how many times a games design is tweaked after implementation because the intended play doesn't feel right and need to be changed? Every time. It's not unique to you or anyone else. Regarding getting your code to fit together like a jigsaw, this is why we have coding standards! Coding standards standardize module interfaces (as well as many other aspects of code creation) to help avoid issue like those you describe. Also Google 'Technical Design Document'. A TDD is very specific about how things will be implemented in the code with full explanation. Implementation of the specifics of the TDD adhere to the coding standards.

@westforduk - I personally find it best not to read such material. Okay, that person couldn't find the motivation or people to complete their project, but that isn't necessarily game development's falt - it happens across all fields in software development.

@xclite - You wouldn't believe how bad an artist I am. There are free resources on the internet where you can find reasonable assets to put in your game. Also, the number one rule for any project is to only rely on yourself as other can quit at any time. If you can't find people select a project that you can manage on your own. Better to complete one game than no game at all!

(To be continued for the remaining so far...)

[xclite]

ButchDean, on 03 February 2011 - 03:50 PM, said:

@xclite - You wouldn't believe how bad an artist I am. There are free resources on the internet where you can find reasonable assets to put in your game. Also, the number one rule for any project is to only rely on yourself as other can quit at any time. If you can't find people select a project that you can manage on your own. Better to complete one game than no game at all!

(To be continued for the remaining so far...)

I know you can find sprite sets online - almost makes doing a cool 2d game within the realm of my skill haha.