What did you learn?

Member
Posts: 268
Joined: 2005.04
Post: #31
Quote:Originally posted by Camacho
I hope this is a joke. If not, coming from a winner, and a game that did well in the contest, I am extremely dissappointed.


Of course it's a joke.

But if a bunch of this year's winners mysteriously "disappear" sometime between now and next year, I'm certainly not going to be involved.

Grin <- See? Smiley. It's a joke.
Quote this message in a reply
Member
Posts: 74
Joined: 2003.08
Post: #32
I learned that I had a project of too much complexity. The testers/gamers couldn't figure out what was happening in Pantheon, and I did not invest enough time in relation to this complexity (aka didn't get finished). On the other hand I had a very complex game, not only in relation to the user interface but also in code and internal structure. And I'm kinda proud that i was able to handle this kind of project all on my own. I also think that this would have not been possible in this way with a c like language (as ppl in the chat know i don't like these). Next time i will make something simpler, and also use more time on the projecting phase, nothing beats a good thought out concept.
Quote this message in a reply
Founder
Posts: 1,138
Joined: 2002.04
Post: #33
Time to raise this thread back from the dead. Some of you learned some things in 2003. Did you apply that to 2004? And what did you learn in 2004? (Besides that people love to hate bugs.) Before answering, be sure to read all the comments from 2003.

Carlos A. Camacho,
Founder
iDevGames
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #34
if you have polish, then nothing else really matters
Quote this message in a reply
Sage
Posts: 1,232
Joined: 2002.10
Post: #35
Except *cough* gameplay.
Quote this message in a reply
Member
Posts: 336
Joined: 2004.07
Post: #36
1) Learn to use the contest to gauge your own strengths, weaknesses, and limitations as a developer. Personally, I found that I can easily project my visions of gameplay into the game and throw in a dash of personal style to boot. My weakness has to do with my artistic talent, and the fact that I don't have any.

Holmes Wrote:If you don't have an Artist, get one!

Amen. Leading into:

2) Team up. Have at least one programmer and one artist. Ideally, have a musician and some web development skills amongst you as well. Being a lone wolf sucks. Even if you CAN do everything, you won't be able to do it as fast as two people. Plus, it's burnout protection. Motivation is contagious. So team up!

3) For uDevGames 2005, the winning entry will feature you fighting off drunken zombie insects in a nanoscopic mining spaceship in Ancient Greece. While you're doing this you have to use your set of giant pincers to dig through the snow and find all the gold nuggets and send them in a horse & cart to Bristol, before the whole freakin' place fills up with water. You heard it here first! Ninja

Justin Ficarrotta
http://www.justinfic.com
"It is better to be The Man than to work for The Man." - Alexander Seropian
Quote this message in a reply
Moderator
Posts: 1,560
Joined: 2003.10
Post: #37
Pick a project that's appropriate for 3 months, not one that should take 6 or more.

- Alex Diener
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #38
JustinFic Wrote:3) For uDevGames 2005, the winning entry will feature you fighting off drunken zombie insects in a nanoscopic mining spaceship in Ancient Greece. While you're doing this you have to use your set of giant pincers to dig through the snow and find all the gold nuggets and send them in a horse & cart to Bristol, before the whole freakin' place fills up with water. You heard it here first! Ninja
I'm getting right on it!
Quote this message in a reply
Moderator
Posts: 770
Joined: 2003.04
Post: #39
Gameplay. Gameplay. Gameplay. At least 0.09 more!
Also:
* Add features sooner, instead of waiting for bug reports before enabling them
* In-game instructions!
* Bullet-time isn't fun for everyone... at least the way I implemented it...
* SDL_mixer has built-in OGG support... I shouldn't have spent an entire week Blush trying to build a custom version of OpenAL to play it when it was already available in the sound API I was using to begin with!
* Blender export scripts work fine until you need them.
* Texturing models is HARD without the proper tools.

Finally, gameplay can mean "simple". If it doesn't, a tutorial level is a must...
Quote this message in a reply
Moderator
Posts: 508
Joined: 2002.09
Post: #40
diordna Wrote:2. Don't expect people to read the Read Me, even if the title of aforementioned file contains a threat

So very true, I forget how many times I have received mails from people using my apps where the only replay I gave was: "please read the Read Me file"

As for what I've learned, even though I didn't participate (except for the music stuff):

1. Just enter the contest, even when I have no idea of a game in the first month
2. Be more active on the IRC channel
3. and like it was said many times: plan!

Next year, I'm entering the contest so be affraid Rasp

"When you dream, there are no rules..."
Quote this message in a reply
Member
Posts: 78
Joined: 2002.06
Post: #41
Quote:if you have polish, then nothing else really matters
Cynically I have to agree with you, people seem to vote 5-irritation instead of 0+enjoyment. Not even gameplay will save you.

Quote:In-game instructions!
Spot on.


Simple is good, and some types of games (really good games) like (the ones that require more work) Industrial Revolution can't compete with the simple and polished in *voting*.

As for my tips
1. Don't use ODE
2. Don't bother with OpenAL
3. Test Test and Test
4. Keep it simple and polished
Quote this message in a reply
Member
Posts: 184
Joined: 2004.07
Post: #42
PowerMacX Wrote:* Bullet-time isn't fun for everyone... at least the way I implemented it...
* SDL_mixer has built-in OGG support... I shouldn't have spent an entire week Blush trying to build a custom version of OpenAL to play it when it was already available in the sound API I was using to begin with!
* Texturing models is HARD without the proper tools.

A couple notes here- I felt there was not any point for bullet-time in your game, since you didn't have to do anything particularly intricate to avoid enemies. Since it didn't help you any, it just wasted your time (this is similar to bullet-time in Prince of Persia/Sands of Time- my friends and I played through the game and almost never used the feature in combat, just for dodging traps, since intricate movement was required. In Max Payne you need it for jumping corners, as you'd again need intricate movements to avoid all the incoming enemy shots. Bullets moved a lot faster in Max Payne than Okugai, as well, if I recall correctly.)

SDL_mixer does support OGG, but SDL_mixer functions very poorly and eats up a lot of CPU. It does work great if you just want something very simple and fast to get up (I think I implemented Audio in my game in about an hour, tops.)

Texturing is probably easiest if you have something to give you a texture atlas. The way I ended up going is just linear-mapping different parts of my models, which probably didn't look as good as it could, but was certain sufficient.

willThimbleby Wrote:As for my tips
1. Don't use ODE
2. Don't bother with OpenAL
3. Test Test and Test
4. Keep it simple and polished

I have to agree with this for the most part, but I think if you have some experience with ODE or OpenAL you probably can use them without a hitch. I used ODE for the first time in this project, and it worked great for me. Your problem was that you made all cars boxes, which hurt the dynamics. OpenAL I heard a lot of people had success with.

I have to echo what a lot of people are saying- if you spend a lot of time going for rich, intricate gameplay, you are leaving yourself open for losing on polish, which in turn probably determines a weight on the rest of your categories (it's a lot easier to be superficial about judging, so a lot of people will only notice the polish, and not give much thought to much else.) This is just if you want to win, though, but I think there are other goals of the contest, such as finally getting to try and make a game that you've been itching to.
Quote this message in a reply
Member
Posts: 78
Joined: 2002.06
Post: #43
I do have experience with both of these, and BitRacer uses fairly involved ODE. My problem was that ODE forced me to make all cars boxes. The other geometries aren't good enough for fast collisions.

My issue with OpenAL was that AFAIK (on the Mac) it provides nothing more than 3D spatial sound (no doppler etc). I can do that myself much more easily, without the hassle of OpenAL.
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #44
the only reason I used openal, was that I already had it in the engine... so why not just use it?

though in the future I might consider FMOD
Quote this message in a reply
Moderator
Posts: 770
Joined: 2003.04
Post: #45
phydeaux Wrote:A couple notes here- I felt there was not any point for bullet-time in your game, since you didn't have to do anything particularly intricate to avoid enemies. Since it didn't help you any, it just wasted your time (this is similar to bullet-time in Prince of Persia/Sands of Time- my friends and I played through the game and almost never used the feature in combat, just for dodging traps, since intricate movement was required. In Max Payne you need it for jumping corners, as you'd again need intricate movements to avoid all the incoming enemy shots. Bullets moved a lot faster in Max Payne than Okugai, as well, if I recall correctly.)

I know, I was only going to implement bullet-time when you were about to die, so you could "transfer your spirit to another body", but couldn't finish it on time. Still, I found it useful when trying to get to the ship with low health, especially on level 3.

phydeaux Wrote:SDL_mixer does support OGG, but SDL_mixer functions very poorly and eats up a lot of CPU. It does work great if you just want something very simple and fast to get up (I think I implemented Audio in my game in about an hour, tops.)

I made a very simple C++ wrapper for OpenAL, which I used on my CMG entry wanna-be "Failed", but the 3D effects were very disappointing to say the least. As for SDL_mixer eating a lot of CPU, you're probably right, but the game was mostly GPU bound, so I could spare a few CPU cicles Wink

phydeaux Wrote:Texturing is probably easiest if you have something to give you a texture atlas. The way I ended up going is just linear-mapping different parts of my models, which probably didn't look as good as it could, but was certain sufficient.

My solution: sphere mapping Rasp

phydeaux Wrote:I have to echo what a lot of people are saying- if you spend a lot of time going for rich, intricate gameplay, you are leaving yourself open for losing on polish, which in turn probably determines a weight on the rest of your categories (it's a lot easier to be superficial about judging, so a lot of people will only notice the polish, and not give much thought to much else.) This is just if you want to win, though, but I think there are other goals of the contest, such as finally getting to try and make a game that you've been itching to.

I just saw the polish scores... I want to see that 3rd decimal! Cry
I guess I should have added a background picture to my .dmg LOL
Quote this message in a reply
Post Reply