Game Ideas ?

As you say it's very easy to go overboard with ideas and kind of cancel your idea before you get to really try it. My tips are:

Go with some basic gameplay that is tried and tested and in a genre you like. Your idea of a match 3 puzzle mechanic sounds solid to me. Implement the basics as fast as possible and get a feeling for the gameplay. Then iterate and add more!

Sure hUb could be an option Some things to consider:

Use standard api's and languages. Something that produces a .exe file and requires no fancy extra installs. hUb only takes care of DirectX so if your game plans on using something else make sure your game handles this gracefully.

Do not modify game data, that is don't modify any files that are installed with the game. These files will be overwritten the next time the game is updated.

Go for iterative development and be prepared for feedback and update loops. This is a must!

Your goal should be to charge some money for your game. This will make your more professional in your development and I also think it is important to value ones own hard work.

That's what I can think of from the top of my head... feel free to ask any questions.

My reason for choosing Java and OpenGL is that I would like this game to run on Linux (Ubuntu) without much hassle. I've always been interested in this since it makes sense to me to keep things as portable as possible.

On the other hand I've never been a huge fan of downloading JRE/JVM-stuff myself. It's almost as horrible to me as being forced to download .NET to run an open source game, because I think that sort of defeats the point. But there is a price to pay for portability, and I think Java handles this fairly well.

I do plan on developing iteratively, and I also plan on using TDD since this is something I've begun believing firmly in.

On the don't modify game data bit. I'm not entirely certain on how far you mean this goes. Things I'd like to save would be settings, like resolution/volume etc. and possibly saved games (although I'm not sure I'd need this in a puzzler), or high-scores.

I was actually thinking I'd let it be free to download, since I'm worried it will be a hassle to have it any other way. But if it isn't that difficult I might charge a small amount for it. =)

______________________________________________________________
There's no kill like an overkill!

Its about time you do a game!!!!

I can only agree with johno and hObbE, the feedback loop from iterative development is really great, it helps me keep focus on the things that really matter(gameplay) instead of digging myself a technical hole...

I think it is important to get basic game play and controls up and running before anything else.

And of course refactor your code, remove things that stinks. Test driven development is really great since you get the security you need to do the changes you want. I use TDD for technical parts of my code in Hero and it really saves time.

Dont worry about music, or graphical resources it can always be bought later (check out soundrangers.com)

Releasing (a non-perfect exe) can be scary, but with hUb you can always do an update when someone finds something you don't want to be exposed. When we started releasing I thought it was really hard to take in some of the "negative" feedback, but fixing those things and put out a new release on hUb without those bug/problems is really great.

Now code like the wind and get me something to test!

Daniel

Example on test driven development.

I rewrote my light map generator using a new algorithm, if I would not use TDD I would need to start the game all the time, and draw a new set of levels to test if I get the shadows right... I would not know if I introduced any new bugs until I verified it by watching the result... and you know LM-generation can take a lot of time...

So instead, before I did the change I created a small test, tracing a few obvious rays, examples that I could get the correct answer by doing the math in head.
I implemented the new algorithm and ran the tests, and corrected the bugs without needing to wait for ray tracing results. When everything looked peachy I created a test level and let the ray-tracer work. If I found something from the manual test, I added a few more test cases that that bug and then fixed the bugs.

So the tests give me the assurance I need to do larger changes in the code, they work exceptionally well when you work with mathematical stuff. And I need not to worry to reintroduce bugs since my tests catch them...

I've got limited experience in using TDD, something I need to change!

On the few things I've used TDD I've mostly created an automated test from a manual one, and then worked to solve some bug. I guess this is somewhat the "wrong" way as you should make the test(s) first.

I also have a hard time keeping my tests in sync with the code. Probably since I use primitive commenting to disable tests when debugging...

I really need to get more into TDD... though my time coding has been very limited as of late

@dntoll it's "of course", not "of cause"

I barely knew about TDD before I began working (I blame you hObbE! j/k). Now it feels quite natural to me, I don't know what tools you use to develop in, but I really like how jUnit works, and it integrates very neatly into eclipse. Another tool we use is EclEmma, a code coverage tool, which can be quite an effective combination with jUnit.

Basically you write a test case for a function you need, run it and see that it fails, then code the function and run the test again and watch it go green (or red, then you code again until it becomes green). It's a great way to stay focused on what you need instead of doing things you think you might need some day. And it's just wonderful for refactoring and such things, as you can easily see if you have broken anything.

But back on topic. I'm still interested in using Java for my game, I'm looking into different ways of deploying it over hUb. I found something called JSmooth, but it is an old project that seems dead. I'm also considering using a webstart approach. There are other options as well that I will be exploring. Hopefully I can post tips and hints for the next person that wants to try out some Java via hUb.

I've also been thinking about the gameplay. I want it to be easy to understand and to start enjoying, but difficult to master. So far all my ideas seem unintuitive and awkward.. I think I'll have a look at some puzzle games on youtube this weekend and see if something there can point me in the right direction.

______________________________________________________________
There's no kill like an overkill!

You are correct that it would make hUb unnecessary for my game. But then again webstart has its drawbacks too, and I would like my game to feel like it belongs in hUb. I'm just not sure how yet.. ;P

______________________________________________________________
There's no kill like an overkill!

These two puzzlers have quite a different type of player in mind. So I think about who your player is...

My personal opinion is that a more classic match three is definitely the safer bet (i.e. more casual audience), especially if you are to make something polished.

You should also take a look at Bejeweled Twist.

Bejeweled Twist has the advantage of not running into the dreaded match three syndrome "no more moves". No more moves is the single most boring aspect of match threes as it totally breaks the flow of the game and it's also quite hard for the player to do something about it, some games consider it a game over state (which I think is unfair) and some like CPQ resorts to various shuffles of the game board. I think this is something to look out for.

Check out the most popular match-3 games

Your ideas regarding webstart sound cool, we could definitely try to make hUb more compatible to different tech. But first we need to see something up and running

I was thinking of having a game mode where you have action points, and each move will cost one point. If you run out of points the game ends. And if you create a match of 3 you get one point back (so plus minus zero). Then you get more points for larger combos. This way I don't have to worry about no more moves left, also I think this is a game play style unique to my game, although I can't be sure. There is a similar version based on time in Bejeweled 2, but I don't enjoy time constraints in games, thus my version was born.

But you are right, I should get started with the actual implementation. =) I'll check out the other match 3 games after work.

Yosumin is close to the style I am looking for, at least when it comes to artwork. The game itself seems interesting, but I don't think it is very accessible to the larger crowd. Thanks for the tip though!

______________________________________________________________
There's no kill like an overkill!

I tired of it quickly as well, gameplay-wise. Maybe that's because my brain can't seem to see those kinds of things fast enough for me...

Also, a word to the wise is to keep the flash-gfx transitions down to a minimum, or at least allow the user to click past them. Yosumin failed here as well.

/johno
"you can't stop the change"

Summer holiday is approaching, just one more week to go then I have 3 weeks to geek around with whatever I want. I'm hoping to get far enough on my game to put it up on hUb before my vacation is over.

I've figured out the basic gameplay elements somewhat, and I have found the solution to having a java game and a few other more technical things. (I think)

I'm still lacking in a style for the game, artwise and music wise. I should probably stick with what I know though, it will probably be something synth/bitpopy for the music and an artstyle similar to portal/mirrors edge etc, clean and crisp. The main reason for this is that I want to make a game that will not scare away people.

Future projects might have a different undertone though, but I won't bother planning for that now. =)

I've also thought up a working title for the game until I can figure out its proper name. Project Jazz. (It has nothing to do with Jazz though, in any way)

______________________________________________________________
There's no kill like an overkill!