So it's already time to take a look back at the development of T.W.T.P.B. our frantic shoot em up game. I just can't believe that it's been more than 1.5 years ago since my create a game in 8h challenge and that the game is now finally finished. If you can say such a thing about a game. Going on 29 public iteration releases every 14 days. Feels a bit sad not to make a release this Friday...
These are my own thoughts and feelings about the development, what went right and what went wrong.
What should be repeated
Finish the game. I'm mighty proud that T.W.T.P.B. is finished. I believed in and saw my ideas through, did my best with them and delivered a final product. Finishing a project is hard, most of game development is transpiration, and a very little part is inspiration. If you start a project you should realize this... especially if you're an indie developer.
During the whole development I had a strong vision about what the game was about and the "lore" of the game world. This helped immensely in prioritizing, cutting features and simply making decisions about how stuff would work or look. Features that would have been "nice to have" could be cut since they simply did not fit the lore and my vision of the game. Even if much of this lore is not actually visible in the final product it helped keeping the product focused and "The game has a great, cohesive feel to it" as Michael Scarpelli at GameTunnel put it.
For once we have made a good looking game! Keeping the graphics simple, stylized and clean have been a success. I had the idea that I, as a programmer, should be able to produce all the graphics. Now I had great help from spjelk doing some very important things, like the ship model, but much of the game is actually "programmers art" and this can be pulled off simply because of the abstract style and focus on effects and a procedural look.
Some of the boss and level design is really good. I like how experience in playing the game matters a great deal and that knowing what to do and when to do it is paramount to success. It's not just "wave the wii-mote and watch that score go", this game is actually quite hard, and requires skill and practice. As it seems one of my strengths as a game designer is to be able to create that sense of flow. It is definitely present in both Time Breaker and in T.W.T.P.B.
On the technical side using a model-view-controller architecture alongside immediate mode gui and integrated game editors worked like a charm. One thing I tried out a bit was "immediate mode particle systems", that is particle systems that are reconstructed from scratch every frame and not a bunch of "active" particle objects that you see in common particle systems. This all worked our great!
This was also my first real project that used shaders to a greater extent. This was also a good thing, allowing for more per pixel control of how stuff gets rendered. By the way the shaders are free to use and abuse. I think the shader api is nice and mature enough to get used. One thing that irritates me is how textures are filtered and that you have no control on how the texture look up works. This makes it hard to use textures as lookup tables in some cases.
What should be avoided
Balancing the game and the abilities proved to be very hard. It's very easy to fall into the dominant strategy trap rendering all other abilites virtually useless. However, much is also up to how the player wants to play the game as such it's not a bad thing. Also since I lived with the game for such a long time some of the levels that I consider to be on the easy side are viewed as very hard. It's not easy to look at your game with fresh eyes. The community helped, but since there was mostly a few people involved that made their voices heard they also lived with the game for a long time.
My experiences of shoot em ups are mostly from the Amiga days, playing Xenon II, Sidewinder, Battle Squadron and the likes. I love shoot em ups, but I have not kept up with recent development in the genre. During development I tried to catch up playing more recent classics such as the excellent Ikaruga. But some decisions where made that could be considered bad design that has lived on to the final product. I should have taken more of a pause and studied the genre thoroughly before making decisions. This would certainly have helped with both marketing and getting the core shoot em up crowd on my side of things. You should simply play the games that stand out in your genre and know what is considered to be state of the art.
The keyboard and game-pad input took a lot of time and also are not that essential to the game as mouse is the premium way of controlling things. This time could probably have been put to better use.
Feel free to comment!
Submitted by hObbE
Fri, 12/19/2008 - 21:24