next »

[Dacke]

I'm making a very portable framework iPhone, 360 C++ support and other platforms  Linux, Mac OS X Windows.

everything very carefully. Using Doxygen, regular comments, classes header and source files, namespaces directories on the hard drive, working up from the bottom (classes Image, Window, View, Tileset, AnimatedImage, Sound encapsulated Vector and String collision redefining blocks/ground/slopes

basic framework, actual game,level editor, draw levels on paper tilesets building levels in the editor,Qt,

To me it sounds like your project is almost doomed to fail
But perhaps you have a herculean ability to stick with it. Please report back!
Good luck anyway!

I sometimes start with the juicy bits and then lose steam.

That's what I do.

I do the fun bit first, then never quite get around to doing the dull bit. Because it's dull!

Hey, that's no fair. You left out the part where I said I felt this was a good thing.
If I feel the project is promising enough I do keep working on it. If I want it to reach the public eye it becomes pretty fun to add the "boring stuff".

I have way to many interesting things to explore in programming. I don't want to get stuck with a giant project just because I have invested a fortune in plumbing.

[oahda]

Woah. You totally butchered my post.
That was a very confusing read.

[Dacke]

Woah. You totally butchered my post.
That a very confusing read.

Hehe. Yeah
I just wanted to point out that your project sounded incredibly ambitious. Which is impressive. But may be hard to get done.

But I do hope all the best for you!

[Triplefox]

Hands up if you started your last game project by programming the title screen?

  yo I do title screens first(sometimes)

I actually have a good reason for doing so, which is that I want to establish game flow early on. Even if the title screen is placeholder in nature I like bringing it at least a little bit towards my vision for the final game and make sure that it doesn't get grating after 1000s of loads during development. But I only tend to do that when I'm pushing towards a finished product and not just experimenting.

The more coding I do, the greater the amount of boring stuff I see, and the more I want to get that out of the way and then savor the tiny nuggets of awesome that are done in an hour or two. I would consider almost everything about the engine I'm working on to be boring, as I had already sketched out the ideas for the most exciting parts months ago, and now I'm just hammering them down to work out the biggest flaws.

The worst part is, I don't even consider it to be "tight" code that I'm writing anymore. I used to be a huge stickler about that and would do multiple rewrites to approach the most perfect solution I could conceive of, and then I discovered how much I really needed to get the job done. Now most of the time I'll write a note like "this is a naive implementation" or "I really don't like the structure here" and then never find myself going back because it turns out to be a good enough black box in the end, and in the meantime I get to deliver value by actually finishing games.

[Schtee]

The worst part is, I don't even consider it to be "tight" code that I'm writing anymore. I used to be a huge stickler about that and would do multiple rewrites to approach the most perfect solution I could conceive of, and then I discovered how much I really needed to get the job done. Now most of the time I'll write a note like "this is a naive implementation" or "I really don't like the structure here" and then never find myself going back because it turns out to be a good enough black box in the end, and in the meantime I get to deliver value by actually finishing games.

I am doing this more and more...and I'm getting more and more done. I hope to go back and make it all less crap, but I guess if that doesn't happen, it doesn't need to!

[oahda]

Woah. You totally butchered my post.
That was a very confusing read.

Hehe. Yeah
I just wanted to point out that your project sounded incredibly ambitious. Which is impressive. But may be hard to get done.

But I do hope all the best for you!

Nah, the project itself shouldn't be too ambitious.
It is actually the result of moving down two times from projects of much larger scale, including moving down one dimension.
I just make sure that everything is very well commented and well structured, so that I do not get lost if I have to return to certain parts of the code a long time after I wrote it.

[Klaim]

Hands up if you started your last game project by programming the title screen?

I've implemented the (fake) "virtual operating system" that is the base of the background of my game first -- that is in fact the title screen and the main menue and a part of the game interface. 

After having done that and a lot of code "around", last year I stoped every feature developpement but the feature "in the middle" (the "domain" part that they talk about in the other article). It was a really better idea than making the plumbering. Too bad I had to pause the project this year, will be back on it at the end of the year once I've finished the first version of another non-game project.

Anyway, I at least learnt a lot by installing TRAC, learning Linux environnement, setting project workflow in TRAC, setting tickets, installing SVN, working on build scripts, integrating a lot of libraries, working on glue code, having a lot of graphic tests, coding some GUI system... before even starting the game representation in code.

Today I'd start directly with that, the game representation in code (that is the domain language in the case of a game). That way it's really easier to build on it.

What keep me motivated for my game is that nobody (even after 4 years of dev and chat about it) did something like that before, so I'm still a pioneer in what I want to achieve.

[Blindsight]

Starting with the key elements is always good IMO, at least for the first rendition of the software. I often find myself re-writing code in order to accommodate for more/better 'plumbing'.

I also find myself writing creators for whatever dataset before the main functionality so I know how it is going to be structured etc.

[TobiasW]

I read this and was like "Is this me? "

http://coderoom.wordpress.com/2010/05/18/start-in-the-middle/

You, Sir, are evil. I wanted to go to sleep two hours ago, but then I unsuspectingly clicked on your innocent looking link and have been following links from there ever since. Worse than Wikipedia, and close to TVTropes...

[bateleur]

Hands up if you started your last game project by programming the title screen?

Amusingly, I started a new project this week by programming the title screen.

Well OK, not literally true. I started it by programming a reusable framework for switching between functionally unrelated screens... and then I wrote the title screen as an easy way to test it!

[oahda]

I read this and was like "Is this me? "

http://coderoom.wordpress.com/2010/05/18/start-in-the-middle/

You, Sir, are evil. I wanted to go to sleep two hours ago, but then I unsuspectingly clicked on your innocent looking link and have been following links from there ever since. Worse than Wikipedia, and close to TVTropes...

What did you click? After reading this, I opened up the article again, but I couldn't find anything interesting to click.