Big Game Projects

[/\/\isanThr0p]

Hey Everyone
well I have seen quite a lot of people who wanted to start a bigger project here. Many failed though. I would like to hear your opinions on bigger game projects and especially tips for it. Where are the biggest problems? (looking to arbiter here...)
Would you recommend to make heavy use of classes?
...

[Ambivalentiowa]

The biggest problem is keeping up your self-motivation. After a while it gets difficult to keep working on the big project when you could be doing smaller finished programs.

[/\/\isanThr0p]

That's true. It's also one of the biggest problems when working in teams and someone doesn't work right anymore...

anything about the coding part?

[Ambivalentiowa]

Nope, sorry but i've lost my motivation.

[riis]

As you probably know, the complexity of larger projects increases enormously. So far, I've worked (in my free time) on two types of projects: 1) small projects, which turned out to be no real challenges and 2) big projects, which became too complex to handle (or which needed too much time).
If you'll ever start a new big project, try to cut down complexity as much as possible. Classes might help, but probably a good design will help you much more.

[/\/\isanThr0p]

that far I have only worked on smaller projects. The biggest thing I ever made was a 3 player jump and run game. It was planned to be four player but I ran out of input devices since I had no joystick.
It seems like there is really no information on programming bigger projects. I do not want to run into the situation you talked about that you have to give up after archieving and working quite a bit.

[Zaei]

Groups are best for large projects. If you are the only one working on a project it is VERY easy to just let it go, and never finish. But, if you have 4 or 5 people working on the project together, its a whole lot more difficult to let your team mates down. If you are making a game, pick something that you REALLY WANT to play, because that also helps motivation.

Organization is really imporatant for big projects. It is really important to separate all of the parts of the project into reusable chunks (this is where VB classes come in handy). On the flip side, again, if you are writing a game, VB classes add a ton of overhead, which can make things difficult.

Z.

[/\/\isanThr0p]

do classes really help in VB aren't they missing functionality?

[Zaei]

There are some things that VB6 classes cannot do(several forms of inheritance) that make programming easier. Other then that, all VB objects are IDispatch objects, which adds overhead to everything you do with them (Your function calls look more like a scripting language internally).

Z.

[Ambivalentiowa]

Since i've never used them in my programming, but have heard lots about them, what are classes?

[Zaei]

A class is like a cookie cutter. You instantiate objects from classes. Each object contains some data, and has methods that act on that data. Think about a form. It has Methods like Move, and Hide, and Data like Top and Left, all in one thing. There is a Form class, and each of your forms is instantiated from that class, and given properites that make it unique (position, size, etc).

You can define a Player class, that has a position, and a speed, and a Method called Move, which would add the speed variable to the position variable.

Z.

[Ambivalentiowa]

Ah, thank you very much. This will help me in my future game endevours.

[Arbiter]

Misanthrop,

Sorry, I've only just noticed this thread. I did reply when you asked me though in another thread, dunno if you saw it or not....

It went as follows...

Originally posted by the VB God

Misanthrop,

Replying to your question asked in that other thread...

There are a few problems to look out for when the project starts getting big. Some are technical, others are mental and psychological.

Off the top of my head....

Technical:
(*) Ensuring you've made regular and comprehensive backups so when you implement a change and it cocks it all up you can take a step back without losing too much work.
(*) Good framework plan. Planning the game at the start reaps massive (unseen) benefits towards the end. Behemoth and I keep wanting to implement changes to the core way the game works, and changing the framework 2yrs after it's been written is extremely hard.

Mental:
(*) Keeping your head around all the code. Sentience is now over 33000 lines of code and it becomes somewhat difficult to remember where it all goes, what it all does and how it all talks to each other.

Psychological:
(*) Keeping motivated to actually finish it. Sentience has been going 2.5 - 3 yrs now and there have been a couple of times when Behemoth and I have nearly packed it inthrough lack of motivation. It's hard to get back into the swing of things after that. You need to ensure the project is something you want to do, and are something you're going to finish.
(*) On a project where you're not the only person, ensuring everyone else keeps up the rate of work is also an issue. If they're paid, it's easy to do as you have leverage - but if they're volunteers or a friend, it's rather difficult.

I've just woke up so that will do for now...

There you go.

[/\/\isanThr0p]

thanks for all the help.
Since I am not planning to start a big project right away I am not thinking about what I really want right now. I know I couldn't spend enough time so I am not gonna start right now since that would already kill my motivation...
My main problem is to keep the programming clean. I had tons of "sideeffects" in all of my programms that got little bigger. I think I improved my programming habits a little, but than again I doubt I would be ready for something real big yet..
please keep the comments coming!!!

[Ambivalentiowa]

Originally posted by Arbiter
Mental:
(*) Keeping your head around all the code. Sentience is now over 33000 lines of code and it becomes somewhat difficult to remember where it all goes, what it all does and how it all talks to each other.

Isn't that what commenting is for?

[Arbiter]

Sentience has a grand total of about 8 comments in the entire thing.

It's now over 34500 lines of code. I've added about 1500 lines in the last fortnight. And that doesn't include the lines I manage to remove when I go on an efficiency clampdown and try to speed up functions.

You try keeping up that speed and commenting at as well. There's just no chance.

Especially when you consider that I work full time as a Business Analyst for British Gas (which rarely involves less than about 40-50 hours a week) and it becomes even harder to manage.

Something has to give. And, as I'm the only person who looks at the code, I decided to drop commenting.

Life's full of compromises...

[Arbiter]

Originally posted by /\/\isanThr0p
thanks for all the help.
Since I am not planning to start a big project right away I am not thinking about what I really want right now. I know I couldn't spend enough time so I am not gonna start right now since that would already kill my motivation...
My main problem is to keep the programming clean. I had tons of "sideeffects" in all of my programms that got little bigger. I think I improved my programming habits a little, but than again I doubt I would be ready for something real big yet..
please keep the comments coming!!!

I know what you mean about programming clean. When I started on Sentience I'd only just started using VB. I'd used a normal Basic on an old Atari when I was younger, but that was years ago. I learnt VB as I wrote Sentience. Ambitious, but fun.

I knew I'd end up rewriting loads of it as I learned the best ways of doing things, so I built the framework to be as modular as possible. If I later learn a better way of doing something, I can just go back to the section that deals with it, scrap it and start again and wouldn't affect the rest of the program.

It's payed massive dividend in the last year where Behemoth and I have changed our minds about how we want things to work...

[Zaei]

Something I just remembered =).

Friends are usually a bad idea. Its REALLY REALLY easy for someone to say that they will help you, and just never do it, when they are comfortable with you.

Z.

[Ambivalentiowa]

Originally posted by Arbiter

Something has to give. And, as I'm the only person who looks at the code, I decided to drop commenting.

Life's full of compromises...

You could send me the code, and then i comment it and send it back to you.

[Arbiter]

You'd honestly go through 34500 lines of code and try and comment it?

It's a nice offer and I thank you for your kindness but it does have one subtle flaw.....

If someone who hasn't seen it before can look at the code and understand it enough to comment it, then surely the comments are redundant... ?

[Ambivalentiowa]

Well, i could still try.

[/\/\isanThr0p]

Ambivalentiowa got to much time on his hands
and: If it was hard to write it should also be hard to read

and Zaei what do you mean with the friends thing?

[Zaei]

When you work with a group of friends, they are generally more unreliable then people who actually want to work on a project.

Z.

[Ambivalentiowa]

Originally posted by /\/\isanThr0p
Ambivalentiowa got too much time on his hands

I don't have too much time on my hands, i just like to try and help others.

[/\/\isanThr0p]

alright then I dont want to discourage you from being nice

[[DBH]Longbow]

hy, i think in big projects you need a good team, motivated!
best ist if you work with some friends, you can talk with them every day ...!

another "friendship-project" (will be big than ever)

www.cooldata.org -> thegörs

[/\/\isanThr0p]

have you read Zaei's post about friends on the project?
well I think if you really know the person and you know that he has the same motivation as you do it's probably best to have someone you already know and you have more contact to than just some guy on the net...

[Zaei]

True, but make sure =).

Z.